CN116486374A - Risky obstacle determination method, self-driving vehicle, electronic device and medium - Google Patents

Abstract

The disclosure provides a method and device for determining risky obstacles, an automatic driving vehicle, electronic equipment, and a storage medium, and relates to the field of computer technology, especially to the field of automatic driving. The specific implementation plan is: determine an obstacle whose existence is uncertain; perform a first-level collision detection on the obstacle to obtain a first-level collision detection result; when the first-level collision detection result is determined to represent the collision risk between the self-driving vehicle and the obstacle, perform a second-level collision detection on the obstacle to obtain a second-level collision detection result; The above solutions improve the ability to deal with the risk of collision with obstacles, especially in the case of multi-sensor fusion, which effectively improves the safety of autonomous vehicles.

Description

Risky obstacle determination method, self-driving vehicle, electronic device and medium

technical field

The present disclosure relates to the field of computer technology, in particular to the field of automatic driving, and specifically relates to a method and device for determining risky obstacles, an automatic driving vehicle, electronic equipment, a storage medium, and a program product.

Background technique

During the driving process of an autonomous vehicle, the problem of collision with obstacles may occur. Therefore, the self-driving vehicle usually judges whether there is a risk of collision between the self-driving vehicle and the obstacle based on its own motion information and the movement information of the obstacle, and re-plans the driving trajectory of the self-driving vehicle so that the self-driving vehicle can drive safely.

Contents of the invention

The present disclosure provides a risk obstacle determination method, device, automatic driving vehicle, electronic equipment, storage medium and program product.

According to an aspect of the present disclosure, a risky obstacle determination method is provided, including:

Identify obstacles of uncertain existence;

Perform a first-level collision detection on the above obstacles to obtain the first-level collision detection result;

When it is determined that the above-mentioned first-level collision detection result is used to represent the risk of collision between the self-driving vehicle and the above-mentioned obstacle, perform a second-level collision detection on the above-mentioned obstacle to obtain a second-level collision detection result; and

In a case where it is determined that the above-mentioned secondary collision detection result is used to indicate that there is a collision risk between the above-mentioned automatic driving vehicle and the above-mentioned obstacle, it is determined that the above-mentioned obstacle is a target risk obstacle.

According to another aspect of the present disclosure, a risky obstacle determination device is provided, including:

an obstacle determination module, configured to determine obstacles whose existence is uncertain;

A first-level detection module, configured to perform a first-level collision detection on the above-mentioned obstacles, and obtain a first-level collision detection result;

A second-level detection module, configured to perform a second-level collision detection on the above-mentioned obstacle to obtain a second-level collision detection result when it is determined that the above-mentioned first-level collision detection result is used to represent a collision risk between the self-driving vehicle and the above-mentioned obstacle; and

The risk determination module is configured to determine that the above-mentioned obstacle is a target risk obstacle when it is determined that the above-mentioned secondary collision detection result is used to indicate that there is a collision risk between the above-mentioned automatic driving vehicle and the above-mentioned obstacle.

According to another aspect of the present disclosure, an electronic device is provided, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the method of the present disclosure.

According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions, wherein the above-mentioned computer instructions are used to make the above-mentioned computer execute the method according to the present disclosure.

According to another aspect of the present disclosure, there is provided a computer program product, including a computer program, which implements the method of the present disclosure when executed by a processor.

According to another aspect of the present disclosure, an automatic driving vehicle is provided, including the electronic device of the present disclosure.

It should be understood that what is described in this section is not intended to identify key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will be readily understood through the following description.

Description of drawings

The accompanying drawings are used to better understand the present solution, and do not constitute a limitation to the present disclosure. in:

Fig. 1A schematically shows a schematic diagram of a scene where a risky obstacle determination method and device can be applied according to an embodiment of the present disclosure;

FIG. 1B schematically shows a structural diagram of an autonomous vehicle according to an embodiment of the present disclosure;

Fig. 2 schematically shows a flow chart of a risky obstacle determination method according to an embodiment of the present disclosure;

Fig. 3 schematically shows a flow chart of determining a target area according to an embodiment of the present disclosure;

Fig. 4 schematically shows a schematic diagram of a first-level collision detection according to an embodiment of the present disclosure;

Fig. 5 schematically shows a schematic diagram of secondary collision detection according to an embodiment of the present disclosure;

Fig. 6 schematically shows a flow chart of a risky obstacle determination method according to another embodiment of the present disclosure;

Fig. 7 schematically shows a block diagram of a risky obstacle determination device according to an embodiment of the present disclosure; and

Fig. 8 schematically shows a block diagram of an electronic device suitable for implementing a risky obstacle determination method according to an embodiment of the present disclosure.

Detailed ways

Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and they should be regarded as exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

In the technical solution of this disclosure, the collection, storage, use, processing, transmission, provision, disclosure, and application of the user's personal information involved are all in compliance with relevant laws and regulations, necessary confidentiality measures have been taken, and they do not violate public order and good customs.

In the technical solution of the present disclosure, before acquiring or collecting the user's personal information, the user's authorization or consent is obtained.

In the open road environment, in order to enable the autonomous driving vehicle to drive safely and comfortably, multiple sensors of different types are usually configured, and the perception data of multiple sensors are fused to realize the perception of the surrounding environment of the autonomous driving vehicle. Obstacles faced by the self-driving vehicle during driving are thus determined. However, in special weather such as night, rain and fog, or in the face of irregular obstacles or obstacles being blocked, it will cause missed or false detection of obstacles, which will affect the accuracy and precision of obstacle detection. Therefore, for obstacles with uncertain existence, such as those with unstable or abnormal existence, they can be reported to the cloud server for monitoring, and the self-driving vehicles can be remotely controlled to reduce the occurrence of safety accidents.

At present, the improvement of the detection range of the sensor makes the surrounding environment perceived by the self-driving vehicle usually panoramic, and relatively more obstacles with uncertain existence are detected, which leads to an increase in the frequency of reporting, thereby affecting the remote processing ability of obstacles with collision risks.

In view of this, an embodiment of the present disclosure provides a method for determining a risky obstacle, including: determining an obstacle whose existence is uncertain; performing a first-level collision detection on the obstacle to obtain a first-level collision detection result; when the first-level collision detection result is determined to indicate the collision risk between the autonomous vehicle and the obstacle, performing a second-level collision detection on the obstacle to obtain a second-level collision detection result;

Fig. 1A schematically shows a schematic diagram of a scene where the method and device for determining risky obstacles according to an embodiment of the present disclosure can be applied.

It should be noted that FIG. 1A is only an example of the system architecture to which the embodiments of the present disclosure can be applied, so as to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios. For example, in another embodiment, the exemplary system architecture to which the method and apparatus for determining risky obstacles may include a terminal device, but the terminal device may implement the method and apparatus for determining risky obstacles provided by the embodiments of the present disclosure without interacting with the server.

As shown in FIG. 1A , a system architecture 100 according to this embodiment may include an autonomous vehicle 101 , a network 102 and a server 103 . Autonomous vehicle 101 may be communicatively coupled to one or more servers 103 via network 102 . Network 102 may be any type of network, such as a wired or wireless local area network (LAN), a wide area network (WAN) such as the Internet, a cellular network, a satellite network, or a combination thereof. The server 103 may be any type of server or server cluster, for example, a cloud server, an application server, a backend server or a combination thereof. The server may be a data analysis server, a content server, a traffic information server, a map and point of interest (MPOI) server, or a location server, etc. For example, the server 103 receives obstacle information transmitted from the autonomous vehicle 101 , and remotely controls the driving or obstacle avoidance of the autonomous vehicle 101 based on the obstacle information.

Autonomous vehicle 101 may refer to a vehicle configured to operate in an autonomous driving mode. But it is not limited to this. Self-driving vehicles can also operate in manual mode, in fully autonomous driving mode, or in partially autonomous driving mode.

It should be understood that the number of self-driving vehicles, networks and servers in Figure 1A is illustrative only. There can be any number of autonomous vehicles, networks, and servers depending on implementation needs.

FIG. 1B schematically shows a structural diagram of an autonomous vehicle according to an embodiment of the present disclosure. As shown in FIG. 1B , the self-driving vehicle 101 may include: a vehicle-mounted terminal 1011 , a perception module 1012 , a collision detection module 1013 and a path planning module 1014 . The self-driving vehicle 101 may also include common components included in ordinary vehicles, such as: engine, wheels, steering wheel, transmission and so on. Commonly used components can be controlled by the on-board terminal and the vehicle control module using various communication commands, such as acceleration commands, deceleration commands, steering commands, and braking commands.

The various modules in autonomous vehicle 101 may be communicatively coupled to each other via interconnects, buses, networks, or combinations thereof. For example, may be communicatively coupled to each other via a Controller Area Network (CAN) bus. The CAN bus is a vehicle bus standard designed to allow microcontrollers and devices to communicate with each other in applications without a host.

The vehicle terminal 1011 may include, but is not limited to, one or more cameras, a global positioning system (GPS) unit, an inertial measurement unit (IMU), a radar unit, a speed sensing unit, an image recognition unit, and a light detection and ranging (LIDAR) unit. The GPS unit may include a transceiver operable to provide information about the location of the autonomous vehicle. The IMU unit can sense changes in position and orientation of the autonomous vehicle based on inertial acceleration. A radar unit may represent a system that utilizes radio signals to sense obstacles within the surroundings of an autonomous vehicle. In addition to sensing obstacles, the radar unit may additionally sense the velocity and/or heading of the obstacle. LIDAR units use laser light to sense obstacles in the environment of an autonomous vehicle. A LIDAR unit may include, among other components, one or more laser sources, a laser scanner, and one or more detectors. Cameras may include one or more devices used to capture images of the environment around the autonomous vehicle. The cameras may be still cameras and/or video cameras. The camera may be mechanically movable, for example, by mounting the camera on a rotating or tilting platform.

The vehicle terminal 1011 may also include other sensors, such as sonar sensors, infrared sensors, steering sensors, speed sensors, accelerator sensors, brake sensors, and audio sensors (eg, microphones). Audio sensors can be configured to pick up sound from the environment around the autonomous vehicle. The steering sensor may be configured to sense the steering angle of the steering wheel, the wheels of the autonomous vehicle, or a combination thereof. The accelerator sensor and the brake sensor respectively sense the accelerator position and the brake position of the self-driving vehicle. In some cases, the accelerator and brake sensors may be integrated into an integrated accelerator/brake sensor.

The self-driving vehicle information, obstacle information, map information of the self-driving vehicle 101 itself, and perception data of the surrounding environment such as signal lamps or signs can be acquired through the vehicle-mounted terminal 1011 .

The sensing module 1012 may receive multiple sensing data of different types from the vehicle terminal 1011, such as image data, point cloud data, and the like. Multiple sensing data can be fused to obtain target area and obstacle information. It is also possible to determine obstacles whose existence is uncertain based on obstacle information and target areas.

The collision detection module 1013 may receive the obstacle information of the uncertain obstacle sent by the perception module 1012, and determine whether there is a collision risk between the obstacle and the autonomous vehicle based on the obstacle information of the uncertain obstacle. When it is determined that there is a risk of collision, the obstacle information is reported to the cloud server. If it is determined that there is no collision risk, the obstacle information is transmitted to the path planning module 1014 .

The path planning module 1014 may receive the obstacle information of the identified obstacle sent by the sensing module 1012, perform path planning based on the obstacle information of the identified obstacle, and obtain the planned path information.

It should be noted that the self-driving vehicle 101 may also include a vehicle control module and a wireless communication module.

Vehicle control modules may include, but are not limited to, steering units, accelerator units (also known as accelerator units), and brake units. The vehicle control module can receive the planned route information of the route planning module, and control the steering unit, accelerator unit and braking unit based on the planned route information. The steering unit is used to adjust the direction or heading of the autonomous vehicle. The accelerator unit is used to control the speed of the electric motor or engine, and thus control the speed and acceleration of the self-driving vehicle. The braking unit decelerates the autonomous vehicle by providing friction to decelerate the autonomous vehicle's wheels or tires.

The wireless communication module allows communication between the autonomous vehicle and external modules such as devices, sensors, other vehicles, etc. For example, the wireless communication module may communicate wirelessly directly with one or more devices, or communicate wirelessly via a communication network, eg, communicate with a server through the network. The wireless communication module may use any cellular communication network or wireless local area network (WLAN), for example using WiFi, to communicate with another component or module. The user interface module may be part of peripheral devices implemented within the autonomous vehicle, including, for example, a keypad, touch screen display, microphone and speaker, and the like.

It should be noted that the sequence number of each operation in the following methods is only used as a representation of the operation for description, and should not be regarded as indicating the execution order of the respective operations. The methods do not need to be performed in the exact order presented, unless explicitly stated otherwise.

Fig. 2 schematically shows a flow chart of a risky obstacle determination method according to an embodiment of the present disclosure.

As shown in FIG. 2, the method includes operations S210-S240.

In operation S210, an obstacle whose existence is uncertain is determined.

In operation S220, a first-level collision detection is performed on the obstacle to obtain a first-level collision detection result.

In operation S230, when it is determined that the first-level collision detection result is used to represent the existence of a collision risk between the autonomous vehicle and the obstacle, perform a second-level collision detection on the obstacle to obtain a second-level collision detection result.

In operation S240, if it is determined that the secondary collision detection result is used to indicate that there is a collision risk between the autonomous driving vehicle and the obstacle, it is determined that the obstacle is a target risk obstacle.

According to an embodiment of the present disclosure, an obstacle with uncertain existence represents an obstacle with unstable or abnormal existence, such as a car with a door open, a strange-shaped car, a semi-trailer, or a car pulling cargo. For obstacles whose existence is uncertain, it is necessary to further determine whether there is a risk of collision between the autonomous vehicle and the obstacle. If there is a risk of collision, the obstacle information can be reported to the cloud server. In order for the cloud server to receive obstacle information, it can remotely control the self-driving vehicle to avoid obstacles, so that the self-driving vehicle can drive safely and comfortably. For obstacles whose existence is determined, such as a driving self-driving vehicle, the automatic driving system can directly plan the driving path of the self-driving vehicle based on the obstacle information of the obstacle, so as to control the self-driving vehicle to avoid obstacles so that the self-driving vehicle can drive safely.

According to an embodiment of the present disclosure, after determining an obstacle whose existence is uncertain, the autonomous vehicle may not directly report to the cloud server. Instead, there is a first-level collision detection for obstacles. Performing a first-level collision detection on an obstacle, and obtaining a first-level collision detection result may include: determining respective speeds of the self-driving vehicle and the obstacle. Based on the respective velocities of the autonomous vehicle and the obstacle, a first-level collision detection result is determined. But it is not limited to this. Level 1 collision detection results can also be determined based on the respective acceleration and velocity of the autonomous vehicle and the obstacle.

According to an embodiment of the present disclosure, in a case where it is determined that the result of the primary collision detection is used to indicate that there is a risk of collision between the autonomous vehicle and the obstacle, a secondary collision detection is performed on the obstacle. Based on the distance between the self-driving vehicle and the obstacle, it is determined whether there is a risk of collision between the two. If the distance is less than the threshold, it means that the obstacle is close to the self-driving vehicle and a collision may occur. Therefore, the obstacle is determined as the target risk obstacle, the obstacle information of the target risk obstacle is stored in a list, and the list is reported to the cloud server.

According to an embodiment of the present disclosure, when the primary collision detection result or the secondary collision detection result is determined to indicate that there is no collision risk between the autonomous vehicle and the obstacle, the obstacle may be determined to be a non-target risk obstacle. The obstacle information of the obstacle can be transmitted to the path planning layer of the autonomous vehicle. The obstacle avoidance process is performed by the path planning layer based on the obstacle information.

According to the embodiments of the present disclosure, two-stage collision risk detection is performed for obstacles with uncertain existence, and reporting is performed when the obstacle is determined to be a target risk obstacle, thereby reducing the frequency of reporting, reducing the data processing capacity of the cloud server, and further improving the remote processing capability for obstacles, effectively improving the safety of autonomous vehicles.

According to an embodiment of the present disclosure, for operation S210 shown in FIG. 2 , determining an obstacle whose existence is uncertain may include the following operations.

For example, based on the planned route information, the target area is determined. Identify initial obstacles whose existence is uncertain. If it is determined that the initial obstacle is located within the target area, the obstacle is determined based on the initial obstacle.

According to an embodiment of the present disclosure, the target area is determined based on the planned path information. The target area can represent the area where the autonomous vehicle may be at risk of collision. The initial obstacles include obstacles of uncertain existence within the target area, and obstacles of uncertain existence on or outside the target area. If it is determined that the initial obstacle is located in the target area, the initial obstacle may be used as an obstacle. Obstacles that need to be detected for collision are thus determined. Through the initial screening of the initial obstacles, the initial obstacles outside the target area and on the target area are filtered out, and the initial obstacles in the target area are left. The initial obstacles are used as obstacles for subsequent first-level collision detection and second-level collision detection. To a certain extent, the amount of data processing is reduced, and the detection efficiency is effectively improved.

According to an embodiment of the present disclosure, determining the target area based on the planned path information may include the following operations.

For example, based on the planned route information, the target route point is determined. Perceived road boundary information, specified road boundary information, and predetermined safety boundary information are determined based on the target waypoint. Perceived road boundary information is obtained from perceptual data. It is specified that road boundary information is obtained from map data. The target boundary information is determined from the perceived road boundary information, the specified road boundary information and the predetermined safety boundary information. Based on the target boundary information, the target area is determined.

According to an embodiment of the present disclosure, determining the target route point based on the planned route information may include an operation of: determining the route point from the planned route information. The path distance between the path point and the current track point of the autonomous vehicle is greater than the obstacle detection distance threshold. Determine the target waypoint based on the current trajectory point and waypoint of the autonomous vehicle.

According to an embodiment of the present disclosure, determining the target waypoint based on the current track point and the waypoint of the automatic driving vehicle may include: performing interpolation processing on the waypoint and the current track point of the automatic driving vehicle to obtain the target waypoint. Since the distance between the waypoints in the planned path information and the current track point of the autonomous vehicle is usually greater than the obstacle detection distance threshold, it is interpolated first to make the waypoints dense, so that the distance between the waypoints is similar to or smaller than the obstacle detection distance threshold. In this way, the division of the target area can be made accurate and effective through the target waypoint.

According to other embodiments of the present disclosure, when the number of waypoints in the planned route information is relatively dense, for example, the distance between the multiple waypoints and the current trajectory point of the autonomous vehicle is smaller than the obstacle detection distance threshold, the waypoint can be directly used as the target waypoint. I won't repeat them here.

According to an embodiment of the present disclosure, the perceived road boundary information is obtained from sensing data collected by the sensor, and the sensing data includes obstacle information at the edge of the road. For example: the perception data includes the position information of the fence or guardrail set on the edge of the road. It is specified that road boundary information is obtained from map data. For example, the specified road boundary information includes road boundary lines and lane lines determined based on traffic rules. The predetermined safety boundary information is determined based on the width of the self-driving vehicle itself. The planned path information can be used as the center line to form two boundary lines based on half the vehicle width or one vehicle width, and then the predetermined safety boundary information can be determined.

According to an embodiment of the present disclosure, the boundary information closest to the target waypoint is determined from the perceived road boundary information, the specified road boundary information, and the predetermined safety boundary information, and is determined as the target boundary information. The target area is divided based on target boundary information. Selecting the boundary information closest to the self-driving vehicle as the target boundary information makes the collision detection process of the self-driving vehicle more accurate and effectively improves the screening accuracy of obstacles.

Fig. 3 schematically shows a flow chart of determining a target area according to an embodiment of the present disclosure.

As shown in FIG. 3, the method includes operations S310-S350.

In operation S310, an interpolation process is performed on the current track point of the autonomous vehicle and the route points in the planned route information to obtain a target route point.

In operation S320, predetermined safety boundary information around each target waypoint is determined.

In operation S330, determine the perceived road boundary information of each target waypoint, determine the boundary information closest to the target waypoint from the predetermined safety boundary information and the perceived road boundary information, and save it.

In operation S340, map road boundary information of each target waypoint is determined, and boundary information closest to the target waypoint is determined from the map road boundary information and saved boundary information to obtain target boundary information.

In operation S350, a target area is determined based on the target boundary information.

According to an embodiment of the present disclosure, determining an initial obstacle whose existence is uncertain may include the following operations.

For example, based on the historical perception data sequence, the type of the obstacle to be identified is determined. The type is used to characterize whether the obstacle to be identified is an obstacle whose existence is uncertain. Based on the type, the initial obstacle is determined from the obstacles to be identified.

According to an embodiment of the present disclosure, the historical sensing data sequence includes data obtained by arranging sensing data at multiple historical moments in time sequence. Perception data at each historical moment may include data collected by multiple sensors. For example, perception data includes image data as well as point cloud data. If it is determined that the type of the obstacle to be identified is an obstacle with uncertain existence, the obstacle to be identified is determined as an initial obstacle, and when the type of the obstacle to be identified is determined to be an obstacle with a certain existence, the obstacle to be identified may be determined as a known obstacle. Send known obstacles to the path planning module for obstacle avoidance. According to an embodiment of the present disclosure, determining the type of the obstacle to be recognized based on the historical sensing data sequence may include the following operations.

For example, based on the obstacle category sequence, the category detection result is determined. Based on the sequence of locations, a location detection result is determined. Based on the speed sequence, a speed detection result is determined. Based on the category detection result, the position detection result and the speed detection result, the type of the obstacle to be recognized is determined.

According to an embodiment of the present disclosure, the historical perception data sequence may include an obstacle category sequence, a position sequence, and a speed sequence. But it is not limited to this. It may also include at least one or two of the following: obstacle category sequence, position sequence and speed sequence.

According to the embodiments of the present disclosure, for the sequence of obstacle categories, the image information of the obstacle to be identified can be collected by the monitoring device, input into the trained image recognition neural network, and the category of the obstacle to be identified can be determined. Save the recognition results corresponding to each historical moment into the obstacle category sequence. For the position sequence, the radar can be used to track the position of the obstacle to be recognized, so as to obtain the position information of the obstacle to be recognized at each historical moment, and save it in the position sequence. For the speed sequence, the speed sensor can be used to collect the movement speed corresponding to each historical moment of the obstacle to be identified, and save it in the speed sequence.

According to an embodiment of the present disclosure, the category detection result may be determined based on the obstacle category sequence. For example, based on the category corresponding to each historical moment in the obstacle category sequence, the category stability is determined, and the category stability is taken as the category detection result. Specifically, category stability is determined by determining the difference between obstacle categories corresponding to each historical moment. For example: at 10:00, the obstacle category is identified as a dog, and at 10:01, the obstacle category is identified as an umbrella, then the stability of the obstacle category is determined to be low.

According to an embodiment of the present disclosure, a position detection result may be determined based on a position sequence. For example, position stability is determined based on the position corresponding to each historical moment in the position sequence, and the position stability is taken as a position detection result. Specifically, position stability is determined by determining the fluctuation amplitude of the position difference at each historical moment. For example: the difference between 10:00 and 10:01 is 1 meter, and the difference between 10;01 and 10:02 is also 1 meter; the difference between 10:02 and 10:03 is still 1 meter, which means that the movement process of the obstacle is stable and the stability is high. You can set a position fluctuation threshold to measure the position stability. For example, if the fluctuation between the set difference reaches 5 meters, it is determined that the position stability of the obstacle is low.

According to an embodiment of the present disclosure, the speed detection result may be determined based on the speed sequence. For example, based on the speed corresponding to each historical moment in the speed sequence, the speed stability is determined, and the speed stability is taken as the speed detection result. Specifically, by determining the moving speed of the obstacle at each historical moment, for example: the moving speed of the obstacle at 10:00 is 48 km/h, the moving speed of the obstacle at 10:01 is also 48 km/h, and the moving speed of the obstacle at 10:03 is also 48 km/h, then it is determined that the moving stability of the obstacle is relatively high. The speed fluctuation threshold can be set to measure the position stability. For example, if the speed fluctuation value between two adjacent moments reaches the threshold 1km/h, it is determined that the position stability of the obstacle is low.

According to an embodiment of the present disclosure, determining the type of the obstacle to be identified based on the category detection result, the position detection result, and the speed detection result may include: determining the target stability based on the category stability, position stability, and speed stability and determining the type of the obstacle to be identified based on the target stability. For example, weighted summation is performed on class stability, position stability and velocity stability respectively to obtain target stability. For example, the higher stability is assigned a value of 1, and the lower stability is assigned a value of -1. The corresponding weight is set for the importance of each stability, and the weighted summation is performed to obtain the target stability. The higher the target stability is, the more stable the obstacle is. If it is greater than the preset threshold, it is determined that the obstacle to be recognized is an obstacle with a certain existence, and if it is less than the preset threshold, it is determined that the obstacle to be recognized is an obstacle with uncertain existence.

For example: set the weight of category stability to 6, the weight of position stability to 2, and the weight of speed stability to 2. Assuming that the preset threshold is 5, and based on the category stability, it is determined that the category stability is assigned a value of 1, the speed stability is assigned a value of -1, and the location stability is assigned a value of -1, then the assignment value corresponding to the target stability is determined to be 2, which is less than the preset threshold value, and the obstacle to be identified is determined to be an obstacle with uncertain existence.

According to the embodiments of the present disclosure, based on the combination of the obstacle category sequence, position sequence and speed sequence in the historical perception data sequence, the type of the obstacle to be identified can be determined, and the multi-sensory data can be fused to a certain extent to obtain multi-sensory fusion data. This improves the accuracy of determining whether the obstacle to be identified is an initial obstacle whose existence is uncertain.

According to the embodiments of the present disclosure, using the method for determining risky obstacles provided by the present disclosure improves the processing capability for obstacle collision risks, and especially effectively improves the safety of autonomous vehicles in the case of multi-sensor fusion.

According to an embodiment of the present disclosure, in a case where it is determined that the initial obstacle is an obstacle of uncertain existence within the target area, it may be preliminarily determined that the obstacle is an obstacle with unstable attributes. There is a certain false detection problem. A first-level collision detection can be performed on this type of obstacle to determine whether there is a risk of collision between the obstacle and the self-driving vehicle. In this way, while reducing the frequency of reporting to the cloud server, the safety of automatic driving is improved.

According to an embodiment of the present disclosure, for the operation S220 shown in FIG. 2 , performing a primary collision detection on an obstacle to obtain a primary collision detection result may include the following operations.

For example, determining the speed of movement of obstacles in the direction of the planned path. Based on the driving speed of the autonomous vehicle and the moving speed of the obstacle, the first-level collision detection result is obtained.

According to an embodiment of the present disclosure, when the driving speed of the self-driving vehicle is less than or equal to the moving speed of the obstacle, the obtained first-level collision detection result is used to indicate that there is no risk of collision between the self-driving vehicle and the obstacle, and the obstacle can be used as a known obstacle, and the obstacle information of the obstacle can be transmitted to the path planning module. When the driving speed of the self-driving vehicle is greater than the moving speed of the obstacle, the obtained first-level collision detection result is used to represent the risk of collision between the self-driving vehicle and the obstacle, and the second-level collision detection is performed on the obstacle.

According to an embodiment of the present disclosure, determining the moving speed of the obstacle in the direction of the planned path may include: decomposing the moving speed of the obstacle into moving speeds parallel to the planned path direction and moving speeds perpendicular to the planned path direction. In this way, the obstacle and the autonomous vehicle can be compared in the same direction of motion, so that the obtained first-level collision detection result is intuitive, simple and effective.

According to an embodiment of the present disclosure, determining the moving speed of the obstacle in the direction of the planned path may include the following operations: based on the current position of the obstacle, determining the second target road section from a plurality of target path points. The multiple target waypoints are determined based on the planned route information of the autonomous vehicle. Based on the second target road section and the moving speed of the obstacle in the world coordinate system, the moving speed of the obstacle in the direction of the planned path is determined.

Fig. 4 schematically shows a schematic diagram of a first-level collision detection according to an embodiment of the present disclosure.

As shown in Figure 4, the route AB is a planned route determined based on the planned route information of the autonomous vehicle. The two dotted lines on the left and right of the planned route are the target boundaries determined based on the target boundary information of the autonomous vehicle, and the area within the target boundary is the target area. Based on the historical perception data sequence, it is determined that the obstacle in the target area is an obstacle P with uncertain existence.

Based on the current position of the obstacle P, the second target road section can be determined from a plurality of target waypoints. As shown in FIG. 4 , multiple solid points on the route AB are multiple target waypoints. Based on the current position of the obstacle P, two target path points d ₁ and d ₂ closest to the obstacle P are determined, and the road section between d ₁ and d ₂ is taken as the second target road section. The driving direction between the second target road sections is the planned route direction.

As shown in FIG. 4 , the moving speed of the obstacle P is v ₁ , which is (v _x , v _y ) in the world coordinate system such as the X-axis and Y-axis coordinate system. V ₁ can be decomposed into velocity v _p in the direction parallel to the planned path and velocity v _m in the direction perpendicular to the planned path. From this, the moving velocity v _p of the obstacle P in the direction of the planned path is determined, and the conversion formula can be as follows:

According to an embodiment of the present disclosure, for the operation S230 shown in FIG. 2 , performing secondary collision detection on obstacles to obtain a secondary collision detection result may include the following operations.

For example, the longitudinal distance between the self-driving vehicle and the obstacle in the direction of the planned path after the predetermined reaction time is determined. Based on the longitudinal distance, the driving speed of the self-driving vehicle and the moving speed of the obstacle, determine the travel time in the direction of the planned path. Based on the driving time, the target position of the obstacle is determined. Based on the target position of the obstacle, a secondary collision detection result is obtained.

According to an embodiment of the present disclosure, the predetermined reaction time of the self-driving vehicle is determined according to the reaction sensitivity of the self-driving vehicle, and may be set to 1 s, for example. The self-driving vehicle still drives along the direction of the planned path and the planned driving speed within the predetermined reaction time, and determines the longitudinal distance between the self-driving vehicle and the obstacle in the direction of the path planning after the predetermined reaction time. Based on the longitudinal distance, travel speed and movement speed of obstacles, the driving time is determined. The travelable duration can be understood as the duration for the autonomous vehicle to be level with the obstacle in the direction of the planned path when the autonomous vehicle decelerates after a predetermined reaction duration. The self-driving vehicle travels at a decelerated speed within the driving time. During the driving process, the longitudinal distance between the self-driving vehicle and the obstacle in the direction of the planned path will gradually decrease until it is vertically equal. Therefore, the target position of the obstacle is determined based on the driving time, and further based on the target position, it is determined whether there is a risk of collision between the self-driving vehicle and the obstacle.

According to the embodiments of the present disclosure, in the process of secondary collision detection, the predetermined reaction time and the travelable time in the direction of the planned path are taken into consideration, and the processing accuracy can be improved while combining the actual real situation.

According to an embodiment of the present disclosure, determining the travelable duration in the direction of the planned path based on the longitudinal distance, the driving speed of the self-driving vehicle and the moving speed of the obstacle may include the following operations: determining the maximum acceleration of the self-driving vehicle for braking based on the longitudinal distance, the driving speed of the self-driving vehicle and the moving speed of the obstacle. If it is determined that the maximum acceleration is greater than the acceleration threshold, the exercisable duration is determined based on the acceleration threshold and the longitudinal distance.

According to an embodiment of the present disclosure, based on the longitudinal distance, the driving speed of the autonomous vehicle and the moving speed of the obstacle, the determined maximum acceleration is the acceleration required to stop without collision in the direction of the planned path. The acceleration threshold is the limit acceleration within the equipment performance range of the autonomous vehicle. In the case that the maximum acceleration is greater than the acceleration threshold, the travelable duration is determined based on the acceleration threshold and the longitudinal distance. In the case that the maximum acceleration is less than or equal to the acceleration threshold, the driving time can be determined based on the maximum acceleration and the longitudinal distance. The exercisable duration may include: the quotient of the longitudinal distance and the maximum acceleration or the acceleration threshold.

According to the embodiment of the present disclosure, the maximum acceleration is compared with the acceleration threshold, thereby determining the travelable duration based on the smaller value, and improving the detection accuracy of the secondary collision detection while combining reality.

According to an embodiment of the present disclosure, obtaining a secondary collision detection result based on the target position of the obstacle may include the following operations: determining respective target sub-positions of multiple edge points of the obstacle based on the target position of the obstacle. For each target sub-position, a first target road segment is determined from multiple target waypoints to obtain multiple first target road segments. The multiple target waypoints are determined based on the planned route information of the autonomous vehicle. Based on the plurality of first target road segments and the plurality of target sub-locations, a plurality of lateral distances are determined. The horizontal distance is the vertical distance between the target sub-position and the first target road segment matching the target sub-position. Based on a plurality of lateral spacings and lateral spacing thresholds, a secondary collision detection result is obtained.

According to the embodiments of the present disclosure, the moving speed of the obstacle can be decomposed into the direction of the planned path and the direction perpendicular to the planned path. When the obstacle and the self-driving vehicle are at the same level in the direction of the planned path, it is possible that the horizontal distance between the obstacle and the self-driving vehicle in the direction perpendicular to the planned path is already far apart, which is greater than the lateral distance threshold. In this case, there is no risk of collision between the obstacle and the self-driving vehicle. There is a risk of collision between the obstacle and the autonomous vehicle when the obstacle and the autonomous vehicle are at the same level in the direction of the planned path, and the lateral distance between the obstacle and the autonomous vehicle in the direction perpendicular to the planned path is less than or equal to the lateral distance threshold.

According to other embodiments of the present disclosure, based on the target position of the obstacle and the target position of the self-driving vehicle, the lateral distance between the two can be determined. But it is not limited to this. The center position of the obstacle may also be determined based on the target position of the obstacle, and the lateral distance may be determined based on the center position and the first target road section matching the center position. The first target road segment matching the central position is a road segment formed by two target route points closest to the central position among the plurality of target route points.

According to the embodiments of the present disclosure, multiple edge points of the obstacle are calculated, thereby determining multiple lateral distances, which can make the secondary collision detection result comprehensive and accurate. In addition, using multiple target waypoints as the estimated position of the self-driving vehicle can reduce the amount of data processing and improve processing efficiency while ensuring the accuracy of secondary collision detection.

Fig. 5 schematically shows a schematic diagram of secondary collision detection according to an embodiment of the present disclosure.

As shown in Figure 5, the route AB is a planned route determined based on the planned route information of the self-driving vehicle. The self-driving vehicle travels to the path point d0 and performs secondary collision detection on the obstacle P traveling to the position S. Determine the longitudinal distance L between the self-driving vehicle and the obstacle P in the direction of the planned path after the predetermined reaction time is determined. Based on the longitudinal distance, the driving speed of the self-driving vehicle and the moving speed of the obstacle P, determine the travel time in the direction of the planned path. Based on the available travel time, the target position E of the obstacle P is determined.

As shown in FIG. 5 , based on the target position E where the obstacle P is located, the respective target sub-positions of the four edge points of the obstacle P are determined. And determine the first target road segment matching each of the four edge points. For example, the first target road segment formed by the target path points d3 and d4 matches the edge point P1. Based on the target sub-position and the first target road section matching the target sub-position, determine a lateral distance, for example, a lateral distance H between the edge point P1 and the first target road section. Get multiple horizontal spacing. A minimum lateral spacing is determined from a plurality of lateral spacings. The minimum lateral spacing is the lateral spacing H. Comparing the lateral distance H with the lateral distance threshold, if it is determined that the lateral distance H is greater than the lateral distance threshold, it is determined that there is no risk of collision between the obstacle and the autonomous vehicle. When it is determined that the lateral distance H is less than or equal to the lateral distance threshold, it is determined that there is a risk of collision between the obstacle and the autonomous vehicle.

According to an embodiment of the present disclosure, assuming that the longitudinal distance is L, the maximum acceleration A of the automatic driving vehicle traveling at the current speed is determined based on the driving speed v _c of the automatic driving vehicle and the moving speed v _p of the obstacle, and the calculation formula is as follows:

According to an embodiment of the present disclosure, it is determined whether the maximum acceleration A is greater than an acceleration threshold Ta, and the setting of the acceleration threshold is related to the braking performance of the automatic driving vehicle. If it is greater than, it is determined that the maximum acceleration for braking of the self-driving vehicle is the acceleration threshold Ta. If not greater than, the maximum acceleration for braking of the self-driving vehicle is A.

According to an embodiment of the present disclosure, when it is determined that the maximum acceleration is greater than the acceleration threshold, the exercisable duration is determined based on the acceleration threshold and the longitudinal distance. Calculated as follows:

Among them, Ta represents the acceleration threshold, T _r represents the predetermined reaction time of the self-driving vehicle, and T _c represents the driving time.

According to an embodiment of the present disclosure, based on the drivable time and the initial position (x ₀ , y ₀ ) of the obstacle, the target position (x _n , y _n ) of the obstacle after the drivable time can be determined, and the calculation formula is as follows:

Fig. 6 schematically shows a flow chart of a risky obstacle determination method according to another embodiment of the present disclosure.

As shown in FIG. 6, the method includes the following operations S601-S610.

In operation S601, an obstacle whose existence is uncertain is determined.

In operation S602, it is determined whether the moving speed of the obstacle in the direction of the planned path is less than the driving speed of the self-driving vehicle. In the case that the moving speed is lower than the driving speed, operation S603 is performed, otherwise, operation S610 is performed.

In operation S603, the longitudinal distance between the autonomous driving vehicle and the obstacle in the direction of the planned path after the predetermined reaction time is determined.

In operation S604, a maximum acceleration of the autonomous vehicle for braking is determined based on the longitudinal distance, the driving speed of the autonomous vehicle, and the moving speed of the obstacle.

In operation S605, it is determined whether the maximum acceleration is greater than an acceleration threshold. If the maximum acceleration is greater than the acceleration threshold, perform operation S6061. In the case that the maximum acceleration is less than or equal to the acceleration threshold, operation S6062 is performed.

In operation S6061, based on the acceleration threshold and the longitudinal distance, the exercisable duration is determined.

In operation S6062, a travelable time is determined based on the maximum acceleration and the longitudinal distance.

In operation S607, the target position of the obstacle is determined based on the travelable time.

In operation S608, it is determined whether the lateral distance between the obstacle and the second target road section is greater than a lateral distance threshold. If it is determined that the lateral distance is less than or equal to the lateral distance threshold, operation S609 is performed. Otherwise, operation S610 is performed.

In operation S609, it is determined that there is a collision risk between the obstacle and the self-driving vehicle.

In operation S610, it is determined that there is no collision risk between the obstacle and the self-driving vehicle.

Fig. 7 schematically shows a block diagram of an apparatus for determining a risky obstacle according to an embodiment of the present disclosure.

As shown in FIG. 7 , a risky obstacle determination device 700 of this embodiment includes an obstacle determination module 710 , a primary detection module 720 , a secondary detection module 730 and a risk determination module 740 .

The obstacle determination module 710 is configured to determine an obstacle whose existence is uncertain. In an embodiment, the obstacle determining module 710 may be configured to perform the operation S210 described above, which will not be repeated here.

The first-level detection module 720 is configured to perform a first-level collision detection on obstacles to obtain a first-level collision detection result. In an embodiment, the first-level detection module 720 may be configured to perform operation S220 described above, which will not be repeated here.

The secondary detection module 730 is configured to perform a secondary collision detection on the obstacle to obtain a secondary collision detection result when it is determined that the primary collision detection result is used to represent the risk of collision between the self-driving vehicle and the obstacle. In an embodiment, the secondary detection module 730 may be configured to perform operation S230 described above, which will not be repeated here.

The risk determination module 740 is configured to determine that the obstacle is a target risk obstacle when it is determined that the secondary collision detection result is used to indicate that there is a collision risk between the autonomous vehicle and the obstacle. In an embodiment, the risk determination module 740 may be configured to perform the operation S240 described above, which will not be repeated here.

According to the embodiments of the present disclosure, two-stage collision risk detection is performed for obstacles with uncertain existence, and reporting is performed when the obstacle is determined to be a target risk obstacle, thereby reducing the frequency of reporting, reducing the data processing capacity of the cloud server, and further improving the remote processing capability for obstacles, effectively improving the safety of autonomous vehicles.

According to an embodiment of the present disclosure, the obstacle determination module 710 includes an area determination submodule, an area determination submodule, and an obstacle determination submodule.

The area determination sub-module is used to determine the target area based on the planning path information.

The initial determination sub-module is used to determine the initial obstacle whose existence is uncertain.

The obstacle determination submodule is used to determine the obstacle based on the initial obstacle when it is determined that the initial obstacle is located in the target area.

According to an embodiment of the present disclosure, the area determination submodule includes a waypoint determination unit, an information determination unit, a boundary determination unit, and an area determination unit.

The waypoint determination unit is configured to determine the target waypoint based on the planned route information.

The information determination unit is configured to determine perceived road boundary information, specified road boundary information, and predetermined safety boundary information based on the target waypoint, wherein the perceived road boundary information is obtained from sensing data, and the specified road boundary information is obtained from map data.

A boundary determining unit, configured to determine target boundary information from perceived road boundary information, specified road boundary information, and predetermined safety boundary information.

The area determination unit is configured to determine the target area based on the target boundary information.

According to an embodiment of the present disclosure, the initial determination submodule includes a type determination unit and an initial determination unit.

The type determination unit is configured to determine the type of the obstacle to be recognized based on the historical perception data sequence, wherein the type is used to represent whether the obstacle to be recognized is an obstacle whose existence is uncertain.

The initial determination unit is configured to determine initial obstacles from obstacles to be identified based on types.

According to an embodiment of the present disclosure, the type determination unit includes a category determination subunit, a position determination subunit, a speed determination subunit, and a type determination subunit.

The category determination subunit is used to determine the category detection result based on the obstacle category sequence.

The position determination subunit is used to determine the position detection result based on the position sequence.

The speed determination subunit is used to determine the speed detection result based on the speed sequence.

The type determination subunit is used to determine the type of the obstacle to be identified based on the category detection result, the position detection result and the speed detection result.

According to an embodiment of the present disclosure, the primary detection module 720 includes a motion determination submodule and a first result determination submodule.

The motion determining sub-module is used to determine the moving speed of the obstacle in the direction of the planned path.

The first result determination sub-module is used to obtain a first-level collision detection result based on the driving speed of the self-driving vehicle and the moving speed of the obstacle.

According to an embodiment of the present disclosure, the secondary detection module 730 includes a vertical determination sub-module, a duration determination sub-module, a position determination sub-module and a second result determination sub-module.

The longitudinal determination submodule is used to determine the longitudinal distance between the self-driving vehicle and the obstacle in the direction of the planned path after the predetermined reaction time.

The duration determination sub-module is used to determine the exercisable duration in the direction of the planned path based on the longitudinal distance, the driving speed of the self-driving vehicle and the moving speed of the obstacle.

The position determining submodule is used to determine the target position of the obstacle based on the driving time.

The second result determination submodule is used to obtain a secondary collision detection result based on the target position of the obstacle.

According to an embodiment of the present disclosure, the duration determination submodule includes a braking determination unit and a duration determination unit.

The braking determination unit is configured to determine the maximum acceleration of the automatic driving vehicle for braking based on the longitudinal distance, the driving speed of the automatic driving vehicle and the moving speed of the obstacle.

The duration determining unit is configured to determine the exercisable duration based on the acceleration threshold and the longitudinal distance when the maximum acceleration is determined to be greater than the acceleration threshold.

According to an embodiment of the present disclosure, the second result determination submodule includes a sub-position determination unit, a first road section determination unit, a distance determination unit and a second result determination unit.

The sub-position determining unit is configured to determine respective target sub-positions of multiple edge points of the obstacle based on the target position of the obstacle.

The first road section determination unit is configured to determine a first target road section from a plurality of target waypoints for each target sub-position to obtain a plurality of first target road sections, wherein the plurality of target waypoints are determined based on planning path information of the autonomous vehicle.

The distance determining unit is configured to determine multiple lateral distances based on the multiple first target road sections and the multiple target sub-positions, wherein the horizontal distance is the vertical distance between the target sub-position and the first target road section matching the target sub-position.

The second result determining unit is configured to obtain a secondary collision detection result based on a plurality of lateral distances and a lateral distance threshold.

According to an embodiment of the present disclosure, the motion determining submodule includes a second road segment determining unit and a motion determining unit.

The second road segment determination unit is configured to determine the second target road segment from a plurality of target way points based on the current position of the obstacle, wherein the plurality of target way points are determined based on the planning route information of the self-driving vehicle.

The motion determining unit is configured to determine the moving speed of the obstacle in the direction of the planned path based on the second target road section and the moving speed of the obstacle in the world coordinate system.

According to an embodiment of the present disclosure, the waypoint determination unit includes a waypoint determination subunit and a target determination subunit.

The waypoint determining subunit is used to determine the waypoint from the planning path information, wherein the path distance between the waypoint and the current track point of the automatic driving vehicle is greater than the obstacle detection distance threshold.

The target determining subunit is used to determine the target way point based on the current track point and way point of the automatic driving vehicle.

Modules, sub-modules, units, any multiple of sub-units according to the embodiments of the present disclosure, or at least part of the functions of any multiple of them may be implemented in one module. Any one or more of modules, submodules, units, and subunits according to the embodiments of the present disclosure may be implemented by being divided into multiple modules. Any one or more of modules, submodules, units, and subunits according to embodiments of the present disclosure may be at least partially implemented as hardware circuits, such as Field Programmable Gate Arrays (Field Programmable Gate Arrays, FPGAs), Programmable Logic Arrays (Programmable Logic Arrays, PLAs), system-on-chips, systems-on-substrates, systems-on-packages, application-specific integrated circuits (Application Specific Integrated Circuits, ASICs), or any other circuit that can be integrated or packaged. Realized by hardware or firmware in a reasonable manner, or by any one of the three implementations of software, hardware and firmware, or by an appropriate combination of any of them. Alternatively, one or more of the modules, submodules, units, and subunits according to the embodiments of the present disclosure may be at least partially implemented as computer program modules, and when the computer program modules are executed, corresponding functions may be performed.

According to an embodiment of the present disclosure, any number of the obstacle determination module 710, the primary detection module 720, the secondary detection module 730, and the risk determination module 740 may be implemented in one module/unit/subunit, or any one of the modules/units/subunits may be split into multiple modules/units/subunits. Alternatively, at least part of the functions of one or more modules/units/subunits of these modules/units/subunits can be combined with at least part of the functions of other modules/units/subunits and realized in one module/unit/subunit. According to an embodiment of the present disclosure, at least one of the obstacle determination module 710, the first-level detection module 720, the second-level detection module 730 and the risk determination module 740 may be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or any other reasonable way to integrate or package the circuit. It can be realized by any suitable combination of any of them. Alternatively, at least one of the obstacle determination module 710, the primary detection module 720, the secondary detection module 730 and the risk determination module 740 may be at least partially implemented as a computer program module, and when the computer program module is executed, corresponding functions may be performed.

It should be noted that the part of the device for determining risky obstacles in the embodiments of the present disclosure corresponds to the part of the method for determining risky obstacles in the embodiments of the present disclosure, and the description of the part of the device for determining risky obstacles refers to the part of the method for determining risky obstacles, which will not be repeated here.

According to the embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

According to an embodiment of the present disclosure, an electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the method according to the embodiment of the present disclosure.

According to an embodiment of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the method according to the embodiments of the present disclosure.

According to an embodiment of the present disclosure, a computer program product includes a computer program, and the computer program implements the method according to the embodiment of the present disclosure when executed by a processor.

According to an embodiment of the present disclosure, a self-driving vehicle is configured with the above-mentioned electronic device, and the configured electronic device can implement the method for determining risky obstacles described in the above-mentioned embodiment when its processor is executed.

Fig. 8 schematically shows a block diagram of an electronic device suitable for implementing a risky obstacle determination method according to an embodiment of the present disclosure. Electronic device is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. Electronic devices may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are by way of example only, and are not intended to limit implementations of the disclosure described and/or claimed herein.

As shown in FIG. 8, the device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to computer programs stored in a read-only memory (ROM) 802 or loaded from a storage unit 808 into a random access memory (RAM) 803. In the RAM 803, various programs and data necessary for the operation of the device 800 can also be stored. The computing unit 801 , ROM 802 , and RAM 803 are connected to each other through a bus 804 . An input/output (I/O) interface 805 is also connected to the bus 804 .

Multiple components in the device 800 are connected to an input/output (I/O) interface 805, including: an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a magnetic disk, an optical disk, etc.; and a communication unit 809, such as a network card, modem, wireless communication transceiver, etc. The communication unit 809 allows the device 800 to exchange information/data with other devices over a computer network such as the Internet and/or various telecommunication networks.

The computing unit 801 may be various general-purpose and/or special-purpose processing components having processing and computing capabilities. Some examples of computing units 801 include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, digital signal processors (DSPs), and any suitable processors, controllers, microcontrollers, etc. The computing unit 801 executes various methods and processes described above, such as a risky obstacle determination method. For example, in some embodiments, the risky obstacle determination method may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as storage unit 808 . In some embodiments, part or all of the computer program may be loaded and/or installed onto the device 800 via the ROM 802 and/or the communication unit 809 . When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the risky obstacle determination method described above can be performed. Alternatively, in other embodiments, the computing unit 801 may be configured in any other appropriate way (for example, by means of firmware) to execute the method for determining risky obstacles.

Various implementations of the systems and techniques described above herein may be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include being implemented in one or more computer programs executable and/or interpreted on a programmable system comprising at least one programmable processor, which may be a special purpose or general purpose programmable processor, capable of receiving data and instructions from and transmitting data and instructions to a storage system, at least one input device, and at least one output device.

Program codes for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to processors or controllers of general-purpose computers, special purpose computers, or other programmable data processing devices, so that the program codes cause the functions/operations specified in the flowcharts and/or block diagrams to be implemented when executed by the processors or controllers. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include one or more wire-based electrical connections, a portable computer disk, a hard disk, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described herein can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, voice input, or tactile input.

The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication, eg, a communication network. Examples of communication networks include: Local Area Network (LAN), Wide Area Network (WAN) and the Internet.

A computer system may include clients and servers. Clients and servers are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

It should be understood that steps may be reordered, added or deleted using the various forms of flow shown above. For example, each step described in the present disclosure may be executed in parallel, sequentially, or in a different order, as long as the desired result of the technical solution disclosed in the present disclosure can be achieved, no limitation is imposed herein.

The specific implementation manners described above do not limit the protection scope of the present disclosure. It should be apparent to those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made depending on design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present disclosure shall be included within the protection scope of the present disclosure.

Claims (26)

1. A risk obstacle determination method, comprising:

determining an obstacle with uncertainty of existence;

performing primary collision detection on the obstacle to obtain a primary collision detection result;

under the condition that the primary collision detection result is used for representing collision risk between the automatic driving vehicle and the obstacle, carrying out secondary collision detection on the obstacle to obtain a secondary collision detection result; and

And determining the obstacle as a target risk obstacle under the condition that the secondary collision detection result is determined to be used for representing the collision risk between the automatic driving vehicle and the obstacle.

2. The method of claim 1, wherein the determining the presence of the uncertain obstacle comprises:

determining a target area based on the planned path information;

determining an initial obstacle with uncertainty in presence; and

in the case where it is determined that the initial obstacle is located within the target area, the obstacle is determined based on the initial obstacle.

3. The method of claim 2, wherein the determining a target area based on planned path information comprises:

determining a target path point based on the planned path information;

determining perceived road boundary information, prescribed road boundary information, and predetermined safety boundary information based on the target route points, wherein the perceived road boundary information is acquired from perceived data, and the prescribed road boundary information is acquired from map data;

determining target boundary information from the perceived road boundary information, the prescribed road boundary information, and the predetermined safety boundary information; and

And determining the target area based on the target boundary information.

4. The method of claim 2, wherein the determining an initial obstacle of uncertainty of presence comprises:

determining a type of an obstacle to be identified based on a historical perception data sequence, wherein the type is used for representing whether the obstacle to be identified is an obstacle with uncertain existence; and

based on the type, the initial obstacle is determined from the obstacles to be identified.

5. The method of claim 4, wherein the historical sensory data sequence comprises an obstacle category sequence, a location sequence, a speed sequence;

the determining the type of the obstacle to be identified based on the historical perception data sequence comprises the following steps:

determining a category detection result based on the obstacle category sequence;

determining a position detection result based on the position sequence;

determining a speed detection result based on the speed sequence; and

and determining the type of the obstacle to be identified based on the category detection result, the position detection result and the speed detection result.

6. The method according to any one of claims 1 to 5, wherein the performing the first-stage collision detection on the obstacle to obtain a first-stage collision detection result includes:

Determining a movement speed of the obstacle in the direction of the planned path; and

and obtaining the primary collision detection result based on the running speed of the automatic driving vehicle and the movement speed of the obstacle.

7. The method according to any one of claims 1 to 6, wherein performing secondary collision detection on the obstacle to obtain a secondary collision detection result, comprises:

determining a longitudinal distance between the autonomous vehicle and the obstacle in a planned path direction after a predetermined reaction time period;

determining a length of exercisable time in the direction of the planned path based on the longitudinal distance, a travel speed of the autonomous vehicle, and a movement speed of the obstacle;

determining a target position of the obstacle based on the drivable time period; and

and obtaining the secondary collision detection result based on the target position of the obstacle.

8. The method of claim 7, wherein the determining a length of exercisable in the planned path direction based on the longitudinal spacing, a travel speed of the autonomous vehicle, and a speed of movement of the obstacle comprises:

determining a maximum acceleration of the autonomous vehicle for braking based on the longitudinal distance, a travel speed of the autonomous vehicle, and a movement speed of the obstacle; and

In the case where it is determined that the maximum acceleration is greater than an acceleration threshold, the exercisable duration is determined based on the acceleration threshold and the longitudinal spacing.

9. The method according to claim 7 or 8, wherein the obtaining the secondary collision detection result based on the target position of the obstacle includes:

determining a target sub-position of each of a plurality of edge points of the obstacle based on the target position of the obstacle;

determining a first target road section from a plurality of target path points for each target sub-position to obtain a plurality of first target road sections, wherein the plurality of target path points are determined based on planning path information of the automatic driving vehicle;

determining a plurality of lateral distances based on the plurality of first target road segments and the plurality of target sub-positions, wherein the lateral distances are vertical distances between the target sub-positions and the first target road segments matched with the target sub-positions; and

and obtaining the secondary collision detection result based on the transverse distances and the transverse distance threshold values.

10. The method of claim 6, wherein the determining a speed of movement of the obstacle in the planned path direction comprises:

Determining a second target road segment from a plurality of target path points based on the current position of the obstacle, wherein the plurality of target path points are determined based on planned path information of the autonomous vehicle; and

a speed of movement of the obstacle in the direction of the planned path is determined based on the second target segment and the speed of movement of the obstacle in the world coordinate system.

11. A method according to claim 3, wherein said determining a target path point based on said planned path information comprises:

determining a path point from the planned path information, wherein a path distance between the path point and a current track point of the automatic driving vehicle is larger than an obstacle detection distance threshold; and

and determining the target path point based on the current track point of the automatic driving vehicle and the path point.

12. A risk obstacle determination device, comprising:

an obstacle determination module for determining an obstacle whose presence is uncertain;

the first-level detection module is used for carrying out first-level collision detection on the obstacle to obtain a first-level collision detection result;

the secondary collision detection module is used for carrying out secondary collision detection on the obstacle under the condition that the primary collision detection result is used for representing collision risk between the automatic driving vehicle and the obstacle, so as to obtain a secondary collision detection result; and

And the risk determination module is used for determining the obstacle as a target risk obstacle under the condition that the secondary collision detection result is used for representing the collision risk between the automatic driving vehicle and the obstacle.

13. The apparatus of claim 12, wherein the obstacle determination module comprises:

the area determination submodule is used for determining a target area based on the planned path information;

an initial determination submodule for determining an initial obstacle whose existence is uncertain; and

and the obstacle determination submodule is used for determining the obstacle based on the initial obstacle when the initial obstacle is determined to be positioned in the target area.

14. The apparatus of claim 13, wherein the region determination submodule comprises:

a path point determining unit configured to determine a target path point based on the planned path information;

an information determination unit configured to determine, based on the target path point, perceived road boundary information, prescribed road boundary information, and predetermined safety boundary information, wherein the perceived road boundary information is acquired from perceived data, and the prescribed road boundary information is acquired from map data;

A boundary determining unit configured to determine target boundary information from the perceived road boundary information, the prescribed road boundary information, and the predetermined safety boundary information; and

and the area determining unit is used for determining the target area based on the target boundary information.

15. The apparatus of claim 13, wherein the initial determination submodule comprises:

a type determining unit, configured to determine a type of an obstacle to be identified based on a historical perception data sequence, where the type is used to characterize whether the obstacle to be identified is an obstacle whose existence is uncertain; and

and the initial determining unit is used for determining the initial obstacle from the obstacles to be identified based on the type.

16. The apparatus of claim 15, wherein the type determining unit comprises:

a category determination subunit, configured to determine a category detection result based on the obstacle category sequence;

a position determining subunit configured to determine a position detection result based on the position sequence;

a speed determination subunit, configured to determine a speed detection result based on the speed sequence; and

and the type determining subunit is used for determining the type of the obstacle to be identified based on the category detection result, the position detection result and the speed detection result.

17. The apparatus of any of claims 12 to 16, the primary detection module comprising:

a movement determination sub-module for determining a movement speed of the obstacle in the planned path direction; and

and the first result determining submodule is used for obtaining the primary collision detection result based on the running speed of the automatic driving vehicle and the movement speed of the obstacle.

18. The apparatus of any of claims 12 to 17, the secondary detection module comprising:

a longitudinal determination submodule for determining a longitudinal distance between the autonomous vehicle and the obstacle in a planned path direction after a predetermined reaction period;

a duration determination submodule for determining a exercisable duration in the planned path direction based on the longitudinal distance, the running speed of the autonomous vehicle, and the movement speed of the obstacle;

a position determination sub-module for determining a target position of the obstacle based on the drivable time period; and

and the second result determining submodule is used for obtaining the secondary collision detection result based on the target position of the obstacle.

19. The apparatus of claim 18, wherein the duration determination submodule comprises:

A brake determining unit for determining a maximum acceleration of the autonomous vehicle for braking based on the longitudinal distance, a running speed of the autonomous vehicle, and a moving speed of the obstacle; and

and the duration determining unit is used for determining the exercisable duration based on the acceleration threshold value and the longitudinal distance under the condition that the maximum acceleration is determined to be larger than the acceleration threshold value.

20. The apparatus of claim 18 or 19, wherein the second result determination submodule comprises:

a sub-position determining unit configured to determine a target sub-position of each of a plurality of edge points of the obstacle based on a target position of the obstacle;

a first route segment determining unit, configured to determine, for each of the target sub-positions, a first target route segment from a plurality of target route points, to obtain a plurality of first target route segments, where the plurality of target route points are determined based on planned route information of the autonomous vehicle;

a distance determining unit, configured to determine a plurality of lateral distances based on the plurality of first target road segments and the plurality of target sub-positions, where the lateral distances are vertical distances between the target sub-positions and first target road segments matched with the target sub-positions; and

And the second result determining unit is used for obtaining the secondary collision detection result based on the transverse intervals and the transverse interval threshold values.

21. The apparatus of claim 17, wherein the motion determination submodule comprises:

a second link determination unit configured to determine a second target link from a plurality of target route points based on a current position of the obstacle, wherein the plurality of target route points are determined based on planned route information of the autonomous vehicle; and

and the motion determining unit is used for determining the motion speed of the obstacle in the planned path direction based on the second target road section and the motion speed of the obstacle in the world coordinate system.

22. The apparatus of claim 14, wherein the waypoint determination unit comprises:

a path point determining subunit, configured to determine a path point from the planned path information, where a path distance between the path point and a current track point of the autonomous vehicle is greater than an obstacle detection distance threshold; and

and the target determining subunit is used for determining the target path point based on the current track point of the automatic driving vehicle and the path point.

23. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 11.

24. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1 to 11.

25. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 11.

26. An autonomous vehicle comprising: the electronic device of claim 23.