WO2022130718A1

WO2022130718A1 - Information generation device

Info

Publication number: WO2022130718A1
Application number: PCT/JP2021/034453
Authority: WO
Inventors: 誠一佐藤
Original assignee: 日立Astemo株式会社
Priority date: 2020-12-18
Filing date: 2021-09-21
Publication date: 2022-06-23
Also published as: JP7470213B2; DE112021005100T5; JPWO2022130718A1

Abstract

Provided is an information generation device that is able to generate test scenarios of various scenes, without being limited to incident scenes, by performing an analysis using a formation format in which the positional relationship of a peripheral object is abstracted. This information generation device is provided with: a formation building unit 9 that sets a formation of a host vehicle and a peripheral object in accordance with a formation format in which, on the basis of analysis data for the relative positions and existence lanes of the peripheral object and the host vehicle, the peripheral object is assigned to each arbitrarily divided region in the periphery of the host vehicle; a formation transition determination unit 10 that determines whether an event has occurred in which the formation of the peripheral object has changed; and a data extraction unit 11 that, when the event has occurred, determines, on the basis of the timing at which the event occurred, an extraction range for an actual vehicle travel log used to generate a scenario corresponding to the event, and extracts data in said extraction range.

Description

Information generator

The present invention refers to a scenario generation method capable of generating a test scenario for each event by analyzing the formation of the own vehicle and surrounding objects, or a scenario generated by the scenario generation method. The present invention relates to an information generator suitable for a vehicle control method capable of performing vehicle travel control in real time.

In the automobile industry in recent years, the development of ADAS (advanced driver assistance system) and automated driving related technology is rapidly progressing, and adaptive cruise control, lane keep assist system, and emergency are available as functions to automate a part of driving operation. Automatic braking and the like have already been put into practical use.

In order to ensure the reliability of ADAS and autonomous driving systems, it is said that hundreds of millions of kilometers of driving tests or tens of thousands of scenario tests are required. There are practical limits to what can be done. Against this background, in recent years, performance tests of ADAS and automated driving systems that utilize simulation have come to be conducted. However, even when performing performance tests by simulation, it is necessary to prepare a huge number of test scenarios. The test scenario should include everything from everyday safety scenes to diverse and complex danger scenes. It is not realistic to manually create these test scenarios one by one, so a mechanism for automatic generation is required.

As a technique related to scenario generation, an information generation device that generates a test case of an incident scene described in Patent Document 1 is known. For example, in paragraph 0030 of Patent Document 1, there is a description that "the presence or absence of abnormal approach is determined by comparing the distance between moving objects and the threshold value (distance threshold value) of the inter-vehicle distance." An information generator that generates a test case using distance information with an object is disclosed.

Japanese Unexamined Patent Publication No. 2019-79296

However, since the information generator described in Patent Document 1 extracts an incident scene based on TTC (Time To Collision), it is not determined as an incident scene depending on the setting of the TTC threshold value, and a test case is generated. Leakage may occur.

Further, the information generator described in Patent Document 1 has a problem that it is not possible to generate a test scenario for a scene that is not actually an incident due to the characteristics of generating a test scenario of an incident scene using TTC as described above. There is also.

Therefore, in the present invention, an information generation device capable of generating test scenarios of various scenes without being limited to incident scenes by performing analysis using a formation format that abstracts the positional relationship of surrounding objects is provided. The purpose is to provide.

In order to solve the above problems, the information generation device of the present invention is an information generation device that generates a scenario of a scene in which the own vehicle travels based on the travel information of the own vehicle, based on the travel information of the own vehicle. A formation forming unit that forms a formation that abstracts the layout of the own vehicle and the surrounding object according to a formation format in which the surrounding object is assigned to each previously arbitrarily divided area around the own vehicle, and the surrounding object. In order to generate a scenario corresponding to an event in which the formation of the surrounding object changes, a formation transition determination unit for determining whether or not an event in which the formation of the surrounding object changes has occurred in the traveling information of the own vehicle is provided. It is characterized by that.

According to the information generator of the present invention, by analyzing the formation based on the existence lane and the positional relationship between the own vehicle and surrounding objects, a test scenario is generated including not only the scene in the dangerous area but also the scene in the normal area. can do. In addition, most of the performance evaluation (80%) of the automatic driving system has a large number of test scenarios that are automatically generated by the information generator of the present invention, which is the process that was mainly performed in the performance evaluation test and development using the actual vehicle in the past. Since the above) can be performed efficiently by simulation, the development speed of the automated driving system can be increased, and the development cost can be expected to be significantly reduced.

In addition, by collating the scene of the own vehicle running with the generated test scenario (event scenario for each event), it is appropriate for various scenes including not only dangerous scenes but also regular scenes. Moreover, it is possible to control the running of the own vehicle in real time.

Issues, configurations and effects other than those described above will be clarified by the explanation of the following embodiments.

The block diagram which shows the whole composition example of the scenario generation apparatus which becomes Example 1 of the information generation apparatus to which the scenario generation method based on the formation analysis which concerns on this invention is applied. The figure which shows the embodiment of the formation format. The figure which shows the processing flow in the formation analysis part. The figure which shows the example of the formation analysis. The figure which shows an example of the state transition in the formation transition judgment when there is one or no surrounding object. The figure explaining the condition in the state transition of FIG. The figure which shows an example of the state transition in the formation transition judgment in the surrounding object existing in front of the own vehicle in two objects. The figure explaining the condition in the state transition of FIG. The figure which shows the Example about the event information addition part. FIG. 3 is a block diagram showing an overall configuration example of a travel control device according to a second embodiment of an information generation device to which the formation analysis according to the present invention is applied.

The present invention generates a test scenario (event scenario) for each event by analyzing the formation of the own vehicle and surrounding objects. The definition of the formation in the present invention is not simply determined by the relative distance or the relative position of the own vehicle and the surrounding vehicle, but is an abstraction of the layout of the surrounding vehicle with respect to the own vehicle according to the formation format specified by the user. Point to.

Hereinafter, examples of the scenario generation method based on the formation analysis according to the present invention will be described with reference to the drawings.

[Example 1 (scenario generator)]
<Overall configuration>
FIG. 1 is a block diagram showing an overall configuration example of a scenario generation device to which a scenario generation method based on the formation analysis of the present invention is applied. The scenario generation device 3 exemplified here is a scenario generation device that outputs an event scenario 19 by inputting a vehicle (hereinafter, may be referred to as a own vehicle or a own vehicle) 1 or a driving log 2 acquired by a driving simulator 1A. be.

The scenario generation device 3 is configured as a computer including a processor such as a CPU (Central Processing Unit), a memory such as a ROM (Read Only Memory), a RAM (Random Access Memory), and an HDD (Hard Disk Drive). Each function of the scenario generator 3 is realized by the processor executing the program stored in the ROM. The RAM stores data including intermediate data of operations performed by a program executed by the processor.

The scenario generator 3 analyzes the formation of the own vehicle and surrounding objects based on the log analysis unit 4 for analyzing the information required for the formation analysis from the travel log 2 and the information analyzed by the log analysis unit 4. Then, the formation analysis unit 8 that outputs the extracted data 12 that is extracted by dividing the travel log 2 based on the analysis result, and the scenario event (cut-in, cut-out, rapid addition) to the extracted data 12. It is provided with an event information addition unit 13 for adding information (deceleration, wobbling, etc.) and a format conversion unit 18 for outputting an event scenario 19 in which the extraction data 12 to which event information is added is converted into a format suitable for the simulator environment. There is.

The travel log 2 records travel information that is data related to the travel scene of the vehicle 1, and is vehicle information (vehicle speed) that can be acquired by the vehicle-mounted network of the vehicle 1 (CAN (Control Area Network), Ethernet (registered trademark)). , Position, etc.) and log data that records detection information of surrounding objects that can be acquired by external recognition sensors (camera, radar, lidar, etc.), and simulation results obtained by driving in a virtual space like the driving simulator 1A. May be used as. In this embodiment, in addition to the vehicle 1 and the driving simulator 1A, the travel log 2 is applied with the data obtained by acquiring the information of the surrounding objects by the outside world recognition sensor installed in the infrastructure such as the security camera and the N system. Is also good.

<Log analysis unit>
The log analysis unit 4 analyzes the traveling lane of the own vehicle, the traveling lane of the surrounding object, and the relative position between the own vehicle and the surrounding object, which are mainly required for the formation analysis, from the traveling log 2. The log analysis unit 4 is composed of a vehicle lane analysis unit 5, a surrounding object lane analysis unit 6, and a relative position analysis unit 7.

(Vehicle lane analysis department)
The own vehicle lane analysis unit 5 determines the travel lane of the own vehicle by analyzing the data (GNSS and map data) related to the travel position of the own vehicle recorded in the travel log 2.

(Around object lane analysis unit)
The surrounding object lane analysis unit 6 has data on the traveling position of the surrounding object recorded in the traveling log 2 (the detection position of the surrounding object detected by the outside world recognition sensor, the relative position with the own vehicle, and the pair obtained by the vehicle-to-vehicle communication. By analyzing the position information of the vehicle, etc.), the traveling lane of the surrounding object is determined.

(Relative position analysis unit)
The relative position analysis unit 7 obtains relative position information by analyzing data related to the relative positions of the own vehicle and surrounding objects recorded in the travel log 2 (detection positions of surrounding objects detected by the outside world recognition sensor, etc.). Relative position information is required to determine the front-back relationship between the vehicle and surrounding objects, which cannot be determined from the travel lane information alone in the formation analysis.

<Formation analysis department>
The formation analysis unit 8 is composed of a formation formation unit 9, a formation transition determination unit 10, and a data extraction unit 11.

(Formation forming part)
The formation forming unit 9 assigns which area around the own vehicle the surrounding object belongs to based on the information analyzed by the log analysis unit 4 according to the formation format specified by the user.

Here, the formation format will be described with reference to FIG.

The formation format can specify in advance which format (for example, designations such as those in FIGS. 2A and 2B) the user uses to analyze the formation of the surrounding object.

The formation format is centered on the vehicle presence lane (“C”), the right area is “R”, the left area is “L”, the road shoulder area is “S”, the front area is “F”, and the front area “F”. The area around the vehicle is arbitrarily divided in advance, such as "FF" for the front area and "R" for the rear area.

In FIG. 2A, two regions of “F” and “FF” are set in the front region of the vehicle 20, “R” is set in the rear region, and “R”, “L” and “S” are set as the left and right regions. This is a setting example. When the formation format shown in FIG. 2A is used, the front area is divided into two, so that not only the scenario related to the preceding vehicle close to the own vehicle but also the behavior of the own vehicle can be indirectly affected by the preceding vehicle. It is possible to generate a scenario in which there is a possibility (for example, when the preceding vehicle interrupts in front of the preceding vehicle and the preceding vehicle suddenly brakes, the own vehicle also needs to suddenly decelerate).

FIG. 2B is a setting example in which the front region of the vehicle 20 is "F", the rear region is "R", and the left and right regions are "R", "L", and "S". In the formation format of FIG. 2B, a scenario can be generated in consideration of two objects of the road shoulder "S" in addition to the six objects of front, rear, left and right directly facing the own vehicle. If it is not necessary to consider a scene that indirectly affects the behavior of the own vehicle, such as the preceding vehicle, use the formation format shown in Fig. 2 (b) to narrow down the generation to the minimum necessary scenarios. Can be done.

Examples of the formation format have been described with reference to FIGS. 2 (a) and 2 (b), but the formation format can be freely (arbitrarily) set according to the scenario that the user wants to generate, and FIG. The setting is not limited to (a) and FIG. 2 (b).

(Formation transition judgment unit)
The formation transition determination unit 10 determines whether or not the formation (abstract layout of the surrounding object) of the surrounding object assigned to the area has changed according to the formation format specified by the formation forming unit 9. For example, if the preceding vehicle in the adjacent lane changes lanes to the driving lane of the own vehicle, it is determined that the formation has changed.

(Data extraction unit)
The data extraction unit 11 generates the extraction data 12 by dividing the travel log 2 before and after the formation is switched each time the formation transition determination unit 10 determines that the formation (abstracted layout) has changed. do.

<Event information granting department>
The event information addition unit 13 is composed of a lane change extraction unit 14, an acceleration extraction unit 15, and an event determination unit 16.

(Lane change extraction section)
The lane change extraction unit 14 analyzes whether or not a lane change has occurred due to a surrounding object based on the data (lane recognition information, lateral position, lateral speed, etc.) recorded in the extracted data 12 regarding the surrounding object. do. Depending on the specified formation format, it may be determined that the assigned area of the surrounding object has changed, assuming that the lane has changed. Taking the formation format shown in FIG. 2A as an example, one example is a case where a surrounding object belonging to “FR” is changed to the assignment to “FC”.

(Acceleration extraction unit)
The acceleration extraction unit 15 determines the magnitude / change of the acceleration of the surrounding object based on the data (velocity, acceleration, etc.) recorded in the extraction data 12 regarding the surrounding object.

(Event Judgment Department)
The event determination unit 16 determines whether or not the behavior information of the surrounding objects analyzed by the lane change extraction unit 14 and the acceleration extraction unit 15 corresponds to the event conditions predefined in the scenario definition file 17. Then, event information (cut-in, cut-out, sudden deceleration, sudden acceleration, etc.) is added to the extracted data 12.

<Format conversion unit>
The format conversion unit 18 generates an event scenario 19 by converting the extraction data 12 to which the event information is added by the event information addition unit 13 into a format that can be used in various simulation environments.

<Formation analysis by formation analysis department>
Here, the formation analysis, which is a feature of this embodiment, will be described in detail with reference to FIGS. 3 and 4. FIG. 3 shows a processing flow of the formation analysis unit 8, and FIG. 4 shows an example in which formation analysis is performed using a certain travel log 2.

The formation analysis unit 8 first reads the log analysis data (information analyzed by the log analysis unit 4) (S300), and then allocates surrounding objects to the area defined in the formation format based on the read log analysis data. Then, a formation is formed (S301).

Since the running log is read out step by step to form a formation, there is no formation registration history in the first step, so it is necessary to register the initial formation. Therefore, in step S302, it is determined whether or not there is a formation registration history.

If there is no formation registration history in step S302, the formation formed in that step is registered as the initial formation (S303).

Next, the analysis start point (t1) for starting the analysis of the formation for dividing the running log 2 for each formation is set (S304). This analysis start point (t1) can be arbitrarily set by the user, but basically it may be the same as the first step of reading the travel log.

In the traveling log 2 shown in FIG. 4, as the initial formation, the vehicle 41 is assigned to the adjacent right lane front “FR” and the vehicle 42 is assigned to the adjacent left lane front “FL” (definition area) around the own vehicle 40. The formation 401 is registered and set as the analysis start point (t1).

After the analysis start point (t1), the mode of the scene recorded in the travel log 2 will change with the passage of time. Therefore, the formation of surrounding objects formed based on the analysis data of the travel log 2 also changes with the passage of time. Therefore, in step S305, whether or not the formation of the surrounding object formed in each step has changed from the last registered formation, that is, whether or not an event has occurred in the traveling log 2 in which the formation of the surrounding object changes. Judge.

If it is determined in step S305 that there is no change in the formation, the process returns to the log analysis data reading process (S300) in the next step.

The case where there is a change in the formation of surrounding objects in step S305 will be described below.

First, the state transition of the formation when there is one surrounding object will be described with reference to FIGS. 5 and 6.

FIG. 5 is a diagram summarizing the state transitions of the formation when there is one or no surrounding object, and FIG. 6 is a table summarizing the conditions when the states in FIG. 5 are switched.

For example, if there is a surrounding object in front of the vehicle driving lane, it means that there is an object in front of the vehicle lane, and then the surrounding object existing in front of the vehicle traveling lane moves to the right lane. Transitions to the state of "there is an object in front of the right". At this time, the transition from the state of "with an object in front of the own lane" to the state of "with an object in front of the right" in FIG. 5 becomes "13", and the state transition occurs when the condition of No. 13 in FIG. 6 is satisfied.

Next, the state transition of the formation when there are a plurality of surrounding objects will be described with reference to FIGS. 7 and 8.

FIG. 7 shows a case where two surrounding objects exist in the front direction of the own vehicle, and FIG. 8 is a table summarizing the state transition conditions in FIG. 7.

For example, when it is determined that the transition to the formation (second formation) 402 is performed after the initial formation 401 shown in FIG. 4, the initial formation 401 is the “object 1 in front of the right lane and the object 2 in front of the left lane” in FIG. It is a state, and the next formation 402 transitions to the state of “the object 1 in front of the own lane and the object 2 in front of the left lane” in FIG. 7. At this time, the state transition is performed based on the condition of "1" in FIG. 7, and it is determined that the condition of No. 1 in FIG. 8 is satisfied, and it is determined that the formation has been switched.

Further, when transitioning from the formation (second formation) 402 to the formation (third formation) 403 in FIG. 4, the condition “13” is changed from the state of “the object 1 in front of the own lane and the object 2 in front of the left lane” in FIG. By satisfying ", there is an object 1 in front of the own lane and an object 2 in front of the object 1".

Here, the definitions of the front object 1 and the front object 2 shown in FIGS. 7 and 8 are merely examples, and the indexes such as the object 1 and the object 2 should be referred to as independent definitions between the states. For example, in a situation where an object exists in the areas "FL" and "FC", when the object in "FL" changes lanes to the own vehicle driving lane "C", put it in front of the object already in the "FC" area. For example, the object existing in "FL" will be assigned to the "FFC" area. Conversely, if you enter behind an object that is already in the "FC" area, the object that was in the "FL" will be assigned to the "FC" area, and the object that was already in the "FC" area will be "FFC". "It will be assigned to an object in the area.

Furthermore, the state transitions in FIGS. 7 and 8 are part of all formation transition judgment rule databases, and combinations can be flexibly defined by the front-back and left-right relationships between not only two objects but also a larger number of surrounding objects.

The processing after it is determined in step S305 of FIG. 3 that there is a formation change will be described.

If it is determined in step S305 of FIG. 3 that there is a formation change (that is, an event that changes the formation of a surrounding object has occurred), the formation at the time of determination is registered (S306).

Next, the timing at which the formation is switched (that is, the timing at which the formation of the surrounding object changes occurs) is set as the formation change point (t1_fc, t2_fc, ...) (S307).

Next, using the formation change point set in step S307 as a base point, a log data division point (t2, t3) is set after an arbitrary time from the formation change point (S308). The user can arbitrarily decide how many seconds before and after the formation change point is set as the log data division point. Basically, as shown in FIG. 4, the analysis start point t1 is set as the first log data division point, and each time the formation is changed, t2 (= t1_fc + α [s]) and t3 (= t2_fc + β [s]]. ) ... should be set automatically. However, if the same formation state continues for a long time after setting the log data division point and the log data during that period is unnecessary, it is possible to set the log data division point separately. As a result, the extraction range of the travel log 2 used to generate the scenario corresponding to the event in which the formation of the surrounding object changes is determined with the formation change point as the base point.

Here, depending on the purpose of use of the user of this embodiment, it may be possible to handle a series of flows of a plurality of events related to formation changes such as cut-in to cut-out in combination as one data. Therefore, the log data division point is arbitrarily set with the formation change point as the base point, but the log data division point is not determined only by the first formation change point, but the subsequent second formation change point. It is also possible to set in a form that includes. Specifically, the start point of the log data division is set based on the first formation change point, and the end point of the log data division is set based on the second formation change point. That is, the extraction range of the travel log 2 includes at least one of the log data division points (corresponding to the occurrence timing of the event in which the formation of the surrounding object changes).

In step S309, by dividing and extracting the travel log data based on the log data division point set in step S308, the extraction data 12 divided for each formation change (within the extraction range described above) is generated. do.

Note that S300 and S301 in FIG. 3 are executed by the formation forming unit 9, S302 to S305 are executed by the formation transition determination unit 10, and S306 and thereafter are executed by the data extraction unit 11.

<Event information is given by the event information giving department>
The event information giving unit 13 of FIG. 1 gives event information such as cut-in, cut-out, overtaking, sudden deceleration, etc. to each extracted data 12 generated by the formation analysis unit 8. It is necessary to analyze the behavior of surrounding objects in order to determine the event to be given.

The lane change extraction unit 14 extracts whether or not the surrounding object has changed lanes based on the lateral position and speed of the surrounding object, or information indicating whether or not the lane has been changed directly.

The acceleration extraction unit 15 extracts information such as sudden deceleration or rapid acceleration (that is, change in speed or acceleration) of the surrounding object based on the moving speed or acceleration information of the surrounding object. The definitions of sudden deceleration and sudden acceleration are not uniquely determined, but can be arbitrarily determined by the user.

In the event determination unit 16, the scenario definition includes the own vehicle lane, the surrounding object lane, and the relative position information analyzed by the log analysis unit 4, and the behavior information of the surrounding object obtained by the lane change extraction unit 14 and the acceleration extraction unit 15. It is determined whether or not the event condition defined in advance in the file 17 is satisfied, and the determined event information is added to the extracted data 12.

An embodiment of the event information giving unit 13 will be described with reference to FIG.

FIG. 9 is described according to the formations (401, 402, 403) formed based on the travel log shown in FIG. After the extraction data 90 of the front and rear sections t1 to t2 of the formation 402 and the extraction data 91 of the front and rear sections t2 to t3 of the formation 403 are generated (by the formation analysis unit 8), the event information is given to each extraction data. Is shown.

Further, FIG. 9 describes the details when event information is given to the extracted data 90. In order to add event information to the extracted data 90, the scenario definition file 92 in which the definition of the event (cut-in, cutout, follow-up, deceleration, sudden deceleration, wobbling, etc.) is described is read.

Here, although not described in FIG. 4 above, in the traveling log 2 of FIG. 4, the surrounding vehicle 41A decelerates after changing lanes from the right lane of the own vehicle 40A to the own vehicle traveling lane (at this time, the surrounding vehicle). 42A is traveling in the left lane), and then the surrounding vehicle 42B traveling in the left lane interrupts in front of the own vehicle 40B and the surrounding vehicle 41B, so that the surrounding vehicle (preceding vehicle) 41B is traveling in a staggered manner.

Therefore, in the extracted data 90 generated based on the formation 402, it is determined that a cut-in and sudden deceleration event has occurred as an event determination, and tag information of cut-in and sudden deceleration is added. Further, in the extracted data 91 generated based on the formation 403, it is determined that a cut-in and wobble event has occurred as an event judgment, and tag information of cut-in and wobble is added.

The scenario definition file 92 describes the definitions of cut-in, sudden deceleration, and wobbling.

The definition of cut-in satisfies both the condition that the surrounding object has transitioned from the adjacent lanes "R" and "L" to the own vehicle traveling lane "C" and that the relative distance to the surrounding object is ahead. It is supposed to be. The definition of the cut-in is only an example defined in this embodiment, and it is also possible to define the cut-in according to the lateral position and lateral speed of the surrounding object.

The definition of sudden deceleration is when the acceleration of the surrounding object is less than or equal to the preset threshold value (for example, -0.4 [G]) in the scenario definition file.

The definition of wobbling may be carried out by setting a certain threshold value from information such as the steering angle, lateral speed, lateral acceleration, or lane departure warning of surrounding objects, and determining how many times the threshold value is exceeded. ..

The vehicle is generated by generating the event scenario 19 by converting the extracted data 12 (90, 91) to which the event information is added as described above into a format that can be used in various simulation environments by the format conversion unit 18. In the scene in which 1 travels, the event scenario 19 can be generated including not only the scene in the dangerous area (incident scene) but also the scene in the regular area.

<Operation and effect of Example 1 (scenario generator)>
As described above, the scenario generator (information generator) 3 of the first embodiment is an information generator (in other words, an information generator) that generates a scenario of the scene in which the own vehicle travels based on the travel information of the own vehicle. , A scenario generator that generates a scenario used for simulation based on the actual vehicle driving and the driving log acquired by the driving simulator) 3. From the driving information (driving log) of the own vehicle, it is arbitrary in advance around the own vehicle. According to the formation format in which the surrounding objects are assigned to each of the divided areas, the formation forming unit 9 that forms a formation that abstracts the layout of the own vehicle and the surrounding objects, and the event that the formation of the surrounding objects changes. In order to generate the corresponding scenario, the formation transition determination unit 10 for determining whether or not an event in which the formation of the surrounding object changes has occurred in the travel information (travel log) of the own vehicle is provided.

Further, when the event occurs, the extraction range of the running information (running log) of the own vehicle used to generate the scenario corresponding to the event is extracted based on the occurrence timing (formation change point) of the event. A data extraction unit 11 for determining and extracting data in the extraction range is provided.

Further, when the event occurs, the data extraction unit 11 divides the travel information (travel log) of the own vehicle before and after the event occurrence timing (log data division point) and obtains data (of the event). (Data divided for each occurrence) is extracted.

Further, the extraction range of the travel information (travel log) of the own vehicle includes one or more of the occurrence timings of the event.

Whether or not the behavior information (at least one of lane change, speed change, or acceleration change) of the surrounding object in the data extracted by the data extraction unit 11 corresponds to a predefined event condition. The event information giving unit 13 that gives the event information of the scenario to the extracted data is provided.

In other words, the scenario generation device (information generation device) 3 of the first embodiment includes a log analysis unit 4 that analyzes the relative positions of the own vehicle and surrounding objects and the relative positions of the own vehicle and surrounding objects from the travel log data. The own vehicle and the surrounding object are assigned according to a formation format in which the surrounding object is assigned to each previously arbitrarily divided area around the own vehicle based on the analysis data of the existing lanes and the relative positions of the own vehicle and the surrounding object. The formation forming unit 9 that sets the formation of the above, the formation transition determination unit 10 that determines whether or not an event that changes the formation of the surrounding object has occurred, and the formation timing of the event when the event occurs. Based on this, the data extraction unit 11 that determines the extraction range of the actual vehicle travel log used to generate the scenario corresponding to the event and extracts the data of the extraction range, and the lane change by the surrounding object with respect to the extracted data. It also includes an event information giving unit 13 that analyzes and gives scenario information (cut-in, cut-out, sudden deceleration, sudden acceleration, etc.) based on acceleration / deceleration.

According to the first embodiment, by analyzing the formation based on the existence lane and the positional relationship between the own vehicle and the surrounding object, a test scenario is generated including not only the scene in the dangerous area but also the scene in the normal area. Can be done. In addition, most of the performance evaluation of the automatic driving system is based on the large number of test scenarios that are automatically generated by the scenario generator 3 of this embodiment, which is the process that was mainly performed in the performance evaluation test and development using the actual vehicle in the past. Since 80% or more) can be efficiently performed by simulation, the development speed of the automatic driving system can be increased, and the development cost can be expected to be significantly reduced.

[Example 2 (travel control device)]
<Overall configuration>
In the first embodiment described above, the scenario generation device 3 including the formation analysis unit 8 has been described, but the formation analysis unit 8 obtains information on surrounding objects obtained by the external world recognition sensor mounted on the vehicle during actual vehicle driving. Based on this, the formation analysis can be applied in real time, and it may be provided in the travel control device of the vehicle.

FIG. 10 is a block diagram showing an overall configuration example of a vehicle travel control device to which the formation analysis according to the present invention is applied.

The travel control device 101 exemplified here is an event scenario 119 stored based on formation analysis and actual travel by inputting vehicle information, external world information, infrastructure information, etc. acquired by vehicle 1B during actual vehicle travel. It is a driving control device that predicts an event by collating the scenes and provides driving control suitable for an actual driving scene.

The travel control device 101 is analyzed by the information receiving unit 102 that receives information necessary for formation analysis and travel control from the vehicle 1B, the information analysis unit 104 that analyzes the information used in the formation analysis, and the information analysis unit 104. A formation that analyzes the transition of the formation of the own vehicle and surrounding objects based on the information obtained, and outputs the extracted data 112 including at least one formation change point (occurrence timing of an event that changes the formation) based on the analysis result. The analysis unit 108 determines a scenario event (cut-in, cutout, rapid acceleration / deceleration, wobble, etc.) with respect to the extracted data 112, outputs the event determination result, and outputs the event determination result information. The event information giving unit 113 given to the extracted data 112, the storage unit 120 to generate the event scenario 119 by storing and accumulating the extracted data 112 to which the event information is given, the event scenario 119 and the actual running. Based on the event prediction unit 121 that predicts the event that the vehicle 1B will encounter in the near future by collating the scenes, the scene predicted by the event prediction unit 121, and the vehicle information and surrounding information obtained by the information receiving unit 102. The vehicle 1B is provided with a travel control unit 122 that instructs the vehicle 1B to perform appropriate travel control.

<Information receiver>
The information receiving unit 102 receives from the vehicle 1B the traveling information related to the traveling scene of the vehicle 1B including the own vehicle information, the outside world recognition information, the GNSS, the map information, the infrastructure information, and the like.

<Information analysis department>
The information analysis unit 104 analyzes the traveling lane of the own vehicle, the traveling lane of the surrounding object, and the relative position between the own vehicle and the surrounding object, which are mainly required for the formation analysis, from the traveling information received by the information receiving unit 102. The information analysis unit 104 has a vehicle lane analysis unit 105 that analyzes the travel lane of the vehicle, a peripheral object lane analysis unit 106 that analyzes the moving lane (travel lane) of the surrounding object, and a relative relationship between the vehicle and the peripheral object. It has a relative position analysis unit 107 that analyzes the position.

<Formation analysis department>
The formation analysis unit 108 has a formation forming unit 109 that forms a formation to which the surrounding object belongs to which area around the own vehicle according to the formation format specified by the user, and whether or not the formation has been switched (that is, the surrounding object). Includes at least one formation transition determination unit 110 that determines (whether or not an event that changes the formation of the formation has occurred) and a formation change point (that is, the occurrence timing of the event that changes the formation) that indicates the timing at which the formation is switched. It has a data extraction unit 111 that accumulates data in an extraction range for an arbitrary past X [s] in the form and extracts the extraction data 112. X can be arbitrarily set in consideration of the design intention.

<Event information granting department>
The event information addition unit 113 extracts whether or not the lane of the own vehicle or the surrounding vehicle has been changed based on the data (lane recognition information, lateral position, lateral speed, etc.) recorded in the extraction data 112 regarding the surrounding objects. The acceleration extraction unit 115 that extracts the degree of acceleration / deceleration (change) of the surrounding object based on the extraction unit 114 and the data (velocity, acceleration, etc.) about the surrounding object recorded in the extraction data 112, and the scenario definition file 117. The event (cut-in, cut-out, sudden deceleration, sudden acceleration, etc.) of the extracted data 112 is determined according to (event conditions defined in advance in), and what kind of event the own vehicle 1B is under at that time is determined. It has an event determination unit 116 that outputs as a signal.

<Memory>
Further, the travel control device 101 accumulates the event scenario 119 to which the event information is added by the storage unit 120 based on the event signal output from the event determination unit 116 of the event information addition unit 113.

<Event prediction department>
The event prediction unit 121 holds the event signal (corresponding to the actual driving scene in which the current own vehicle 1B is placed) output from the event determination unit 116 of the event information addition unit 113 and the data of the past similar event. By collating the event scenario 119, it is predicted what kind of event is likely to occur in the near future regarding the event occurring in the running own vehicle 1B. Further, the predicted event and its occurrence probability are input to the traveling control unit 122.

<Driving control unit>
In the travel control unit 122, the own vehicle 1B is traveling on a deceleration command related to travel control, a target travel speed command, a lateral movement amount command, etc., based on the predicted event and the occurrence probability input from the event prediction unit 121. It can be set appropriately according to the scene.

<Event prediction by the event prediction unit and command setting by the driving control unit>
For example, assuming the situation shown in FIG. 4, the vehicle 41A cuts in in front of the overtaking

vehicles

41A and 41B that have been cut in in front of the

own vehicles

40A and 40B (own vehicle traveling lane). It is assumed that an event scenario in which 41B suddenly brakes is already stored by the storage unit 120. When the vehicle is driving in the second lane of a three-lane road while the actual vehicle is running, the overtaking vehicle in the third lane is approaching, and the vehicle is also driving in the first lane. When the scene in which the current own vehicle is placed is collated with the past event scenario, it can be determined that there is a high possibility that an event in which the preceding vehicle cuts in and suddenly decelerates occurs (event prediction unit 121). It becomes possible to adjust the traveling control in real time, such as accelerating the deceleration timing of the ACC (Adaptive Cruise Control) (travel control unit 122).

<Operational effect of Example 2 (travel control device)>
As described above, the travel control device (information generation device) 101 of the second embodiment is based on the storage unit 120 that stores the event scenario to which the event information is assigned and the current travel information of the own vehicle. , The event prediction unit 121 that predicts an event that will occur in the current own vehicle in the future (high probability) by collating the scene in which the current own vehicle is placed with a past event scenario (storage unit 120). And the travel control unit 122 that controls the travel of the current own vehicle based on at least one of the event predicted by the event prediction unit 121 (prediction event) or the probability of occurrence of the predicted event.

According to the second embodiment, in addition to the operation and effect of the first embodiment described above, the scene of the own vehicle in motion is collated with the generated test scenario (event scenario for each event) to create a dangerous scene. Not only that, it is possible to appropriately and in real time control the running of the own vehicle in various scenes including the scene in the regular area.

Although the embodiments of the present invention have been described above with reference to the drawings, the specific configuration is not limited to the above-described embodiments, and even if there are design changes and the like within a range that does not deviate from the gist of the present invention. , They are included in the present invention.

For example, the above-mentioned examples have been described in detail in order to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the described configurations. Further, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Further, it is possible to add / delete / replace a part of the configuration of each embodiment with another configuration.

Further, each of the above configurations, functions, processing units, processing means, etc. may be realized by hardware by designing a part or all of them by, for example, an integrated circuit. Further, each of the above configurations, functions, and the like may be realized by software by the processor interpreting and executing a program that realizes each function. Information such as programs, tables, and files that realize each function can be stored in a memory, a hard disk, a storage device such as an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.

In addition, the control lines and information lines indicate those that are considered necessary for explanation, and do not necessarily indicate all control lines and information lines in the product. In practice, it can be considered that almost all configurations are interconnected.

1, 1B: Vehicle (own vehicle), 1A: Driving simulator 2: Driving log 3: Scenario generation device (information generation device) 4: Log analysis unit 5, 105: Own vehicle lane analysis unit, 6, 106 : Surrounding object lane analysis unit, 7, 107: Relative position analysis unit, 8, 108: Formation analysis unit, 9, 109: Formation formation unit, 10, 110: Formation transition determination unit, 11, 111: Data extraction unit, 12 , 112: Extracted data, 13, 113: Event information addition unit, 14, 114: Lane change extraction unit, 15, 115: Acceleration extraction unit, 16, 116: Event judgment unit, 17, 117: Scenario definition file, 18: Format converter, 19, 119: Event scenario, 40, 40A, 40B: Own vehicle, 41, 41A, 41B, 42, 42A, 42B: Surrounding object, 401: Initial formation, 402: Second formation, 403: Third Formation, 90, 91: Extracted data, 92: Scenario definition file, 101: Travel control device (information generation device), 102: Information receiving unit, 104: Information analysis unit, 120: Storage unit, 121: Event prediction unit, 122 : Driving control unit

Claims

In the information generation device that generates a scenario of the scene in which the own vehicle travels based on the travel information of the own vehicle.
From the traveling information of the own vehicle, a formation that abstracts the layout of the own vehicle and the surrounding object is formed according to a formation format in which a peripheral object is assigned to each previously arbitrarily divided area around the own vehicle. Formation formation part and
In order to generate a scenario corresponding to an event in which the formation of the surrounding object changes, a formation transition determination unit for determining whether or not an event in which the formation of the surrounding object changes has occurred in the traveling information of the own vehicle. An information generator, characterized in that it comprises.
In the information generator according to claim 1,
When the event occurs, data for extracting the data in the extraction range by determining the extraction range of the traveling information of the own vehicle used to generate the scenario corresponding to the event based on the occurrence timing of the event. An information generation device characterized by having an extraction unit.
In the information generator according to claim 2,
The data extraction unit is an information generation device, characterized in that, when the event occurs, the travel information of the own vehicle is divided and data is extracted before and after the event occurrence timing.
In the information generator according to claim 2,
An information generation device characterized in that the extraction range of travel information of the own vehicle includes one or a plurality of occurrence timings of the event.
In the information generator according to claim 2,
By determining whether or not the behavior information of the surrounding object in the data extracted by the data extraction unit corresponds to the predefined event conditions, the event information of the scenario is given to the extracted data. An information generation device including an event information addition unit.
In the information generator according to claim 5,
The information generation device, characterized in that the behavior information of the surrounding object includes at least one of the presence / absence of a lane change, a change in speed, or a change in acceleration of the surrounding object.
In the information generator according to claim 5,
A storage unit that stores the event scenario to which the event information is assigned, and a storage unit.
An event prediction unit that predicts future events in the current own vehicle by collating the scene in which the current own vehicle is placed with the past event scenario based on the current running information of the own vehicle. ,
An information generation device including a travel control unit that controls the travel of the current own vehicle based on at least one of the event predicted by the event prediction unit and the probability of occurrence of the predicted event.
In the information generator according to claim 1,
The travel information of the own vehicle includes a travel log acquired by the actual vehicle travel of the own vehicle or a driving simulator, or a travel information acquired by the own vehicle during the actual vehicle travel. ..