JP7501379B2 - Autonomous Vehicle - Google Patents

Description

The present invention relates to an autonomous vehicle.

In the autonomous vehicle disclosed in Patent Document 1, if there are avoidable areas on either side of the vehicle's direction of travel within a search area extending in the direction of the vehicle's travel path based on the detection results of a sensor mounted on the vehicle, the vehicle will take evasive action to head toward the closest avoidable area.

JP 2020-42469 A

However, simply taking evasive action to head toward the closest avoidable area from the vehicle may result in a longer arrival time at the destination.

The autonomous vehicle for solving the above problem comprises a vehicle, a sensor mounted on the vehicle, a search unit that searches for whether or not there are multiple avoidable areas in which the vehicle can avoid the obstacle in a horizontal direction intersecting the traveling direction from the obstacle when the sensor detects that an obstacle is present in a search area extending in the direction of the path along which the vehicle travels, and an evaluation unit that evaluates each avoidable area based on an allowable speed for returning to the original traveling path through the avoidable area when the search unit determines that the multiple avoidable areas exist, and an avoidance traveling execution unit that selects an avoidable area highly evaluated by the evaluation unit from among the multiple avoidable areas and causes the vehicle to perform an avoidance action by traveling through the avoidable area.

According to this, each avoidable area is evaluated based on the permissible speed for returning to the original driving route through the avoidable area, and the vehicle takes evasive action by selecting the avoidable area with the highest evaluation from among multiple avoidable areas and driving through that avoidable area.

Therefore, when the vehicle takes evasive action by simply heading toward the closest avoidable area, for example, even if the width of the avoidable area is narrow or the speed is limited in the avoidable area, the area closest to the vehicle will be selected, which may result in a longer arrival time at the destination, whereas with this method, the vehicle can take an efficient avoidance action.

In addition, in an autonomous vehicle, the evaluation unit may determine the permissible speed based on the width of the avoidable area in a direction intersecting the traveling direction and the distance from the vehicle to the avoidable area.

In addition, in an autonomous vehicle, the evaluation unit may determine the permissible speed based on the width of the avoidable area in a direction intersecting the traveling direction, the distance from the vehicle to the avoidable area, and the speed limit in a speed limit area on the path to return to the traveling path.

In addition, in an autonomous vehicle, the evaluation unit may determine the permissible speed based on the width of the avoidable area in a direction intersecting the traveling direction, the distance from the vehicle to the avoidable area, and whether the route for returning to the traveling route includes a section in which at least one of a temporary stop and slowing down is performed.

The present invention allows for efficient avoidance actions.

FIG. Block diagram of an autonomous vehicle. 4 is a flowchart showing an obstacle avoidance process. FIG. 4 is a schematic diagram showing a process for searching an avoidable area. FIG. 11 is an explanatory diagram showing evaluation of a condition. FIG. 4 is a schematic diagram showing a process for searching an avoidable area. FIG. 11 is an explanatory diagram showing evaluation of a condition. FIG. 4 is a schematic diagram showing a process for searching an avoidable area. FIG. 11 is an explanatory diagram showing evaluation of a condition. 4 is a time chart for explaining intersection control. 4 is a time chart for explaining intersection control. 4 is a time chart for explaining intersection control. FIG. 4 is a schematic diagram showing a process for searching an avoidable area.

An embodiment of the autonomous vehicle will be described below.
1, the autonomous vehicle 10 includes a vehicle 20, a sensor 31 mounted on the vehicle 20, and a control device 32 mounted on the vehicle 20. The vehicle 20 includes a vehicle body 21 and a plurality of wheels 22. The autonomous vehicle 10 is a transport vehicle that transports loads.

The wheels 22 in this embodiment are omnidirectional wheels. An omnidirectional wheel is a wheel that rotates together with the axle and also allows movement in the axial direction of the axle. Four wheels 22 are provided. By controlling the rotation speed and direction of the wheels 22, the vehicle 20 can move in all directions while maintaining the orientation of the body 21, move while changing the orientation of the body 21, and change the orientation of the body 21 without moving. Note that the above-mentioned "omnidirectional" refers to the direction of movement on the road surface or floor surface on which the vehicle 20 is traveling.

As shown in FIG. 2, the autonomous vehicle 10 includes a drive mechanism 41 that drives the wheels 22. The drive mechanism 41 includes a motor 42 for rotating the wheels 22, and a motor driver 43 that drives the motor 42. Although not shown in the figure, the number of motors 42 and motor drivers 43 provided is the same as the number of wheels 22. The motor driver 43 controls the rotation speed of the motor 42 in response to a command from the control device 32. The control device 32 can control the traveling direction of the vehicle 20 by controlling the rotation speed of the motor 42 via the motor driver 43.

Next, the sensor 31 and the control device 32 will be described in detail.
The sensor 31 used is one that can cause the control device 32 to detect an obstacle. The sensor 31 in this embodiment is a laser range finder. A laser range finder is a distance meter that measures distance by irradiating a laser to the surroundings and receiving light reflected from the area where the laser hits. In this embodiment, a two-dimensional laser range finder that irradiates a laser while changing the irradiation angle in the horizontal direction is used.

If the part where the laser hits is considered to be the irradiation point, the sensor 31 measures the distance to the irradiation point in correspondence with the irradiation angle. In other words, the sensor 31 can measure relative coordinates that indicate the distance between the vehicle 20 and the irradiation point. In this embodiment, the horizontal laser irradiation angle of the sensor 31 is 270°.

As shown in FIG. 1, if the center of the irradiation angle is taken as reference axis B, the irradiation angle is in the range of reference axis B ±135 degrees. The relative coordinates measured by sensor 31 are coordinates in a Cartesian coordinate system in which the direction in which reference axis B extends is the Y axis and the direction perpendicular to the Y axis is the X axis.

As shown in FIG. 2, the control device 32 includes a CPU 33 and a storage unit 34 including RAM, ROM, and the like. Various programs for controlling the vehicle 20 are stored in the storage unit 34. The control device 32 may include dedicated hardware for performing at least some of the various processes, such as an application specific integrated circuit (ASIC). The control device 32 may be configured as a circuit including one or more processors that operate according to a computer program, one or more dedicated hardware circuits such as ASICs, or a combination thereof. The processor includes a CPU and memory such as RAM and ROM. The memory stores program code or instructions configured to cause the CPU to execute processes. The memory, i.e., computer readable medium, includes anything that can be accessed by a general-purpose or dedicated computer.

As shown in FIG. 1, when the vehicle 20 is driven, the control device 32 generates a driving route R. The driving route R is a route from the current position, which is the starting point of the vehicle 20, to the final destination. When the vehicle 20 is moving straight, the driving route R and the reference axis B face in the same direction. In FIG. 1, the driving route R and the reference axis B overlap. Methods for generating the driving route R include a method of generating a route by converting a drivable route into a grid map, a potential method using a potential field, and the like.

The control device 32 sets a position on the travel route R a predetermined distance away from the vehicle 20 as a target point Pt, and controls the drive mechanism 41 to travel toward the target point Pt. The control device 32 drives the vehicle 20 so that the reference axis B of the sensor 31 faces the traveling direction of the vehicle 20. The control device 32 also performs obstacle avoidance processing to keep the distance between the vehicle 20 and the obstacle at or above a predetermined value so that the vehicle 20 does not come into contact with the obstacle.

The control device 32 detects obstacles that exist in the traveling direction of the vehicle 20. The detection of obstacles is performed from the detection results of the sensor 31. If the X and Y coordinates of the illuminated point of the sensor 31 exist within the search area As, the control device 32 identifies the illuminated point as an obstacle and determines that an obstacle exists.

The search area As is a rectangular area, one of the two long sides LS1, LS2 extending in the X-axis direction passes through a predetermined position Pc of the vehicle 20, and the other is located in the traveling direction of the vehicle 20. The search area As is an area extending from the vehicle 20 in the direction of the route along which the vehicle 20 travels. Here, the direction of the route along which the vehicle 20 travels refers to the traveling direction of the vehicle 20 on the set travel route R. The predetermined position Pc is an arbitrary position of the vehicle 20, and is a fixed position determined in advance. In this embodiment, the predetermined position Pc is the center of the vehicle body 21 in the Y-axis direction and the center of the vehicle body in the X-axis direction. The search area As is an area that is linearly symmetrical with respect to the reference axis B.

The dimension of the search area As in the Y-axis direction is determined based on how far the obstacle search should be performed along the path along which the vehicle 20 is traveling. In this embodiment, the distance in the Y-axis direction from the predetermined position Pc to the target point Pt is the dimension of the search area As in the Y-axis direction, that is, the dimension of the short sides SS1, SS2 of the search area As. The dimension of the search area As in the Y-axis direction is set according to, for example, the time required to move the vehicle 20 from the detection of an obstacle to a position where the obstacle can be avoided. Note that the target point Pt does not necessarily need to be located on the long side LS2 in order to avoid obstacles, and may be set at a position that is appropriately shifted in consideration of the agility and target tracking ability of the vehicle 20 during avoidance.

The dimension of the search area As in the X-axis direction is determined based on how far the search for obstacles should be performed in a direction intersecting the direction of the path along which the vehicle 20 is traveling. The dimension of the search area As in the X-axis direction varies depending on the location where the vehicle 20 is expected to travel, and varies depending on the layout of the room, the width of the road, etc. If the location where the autonomous vehicle 10 will travel is determined in advance, the dimension of the search area As in the X-axis direction may be a fixed value that is preset depending on the location. If the location where the autonomous vehicle 10 will travel is not determined, the control device 32 may change the dimension of the search area As in the X-axis direction depending on the location where the autonomous vehicle 10 will travel.

Next, the operation will be described.
The control device 32 executes an obstacle avoidance process shown in Fig. 3. This obstacle avoidance process is repeatedly performed while the vehicle 20 is traveling.

In step S10, the control device 32 determines whether or not an obstacle is present within the search area As. If no obstacle is present within the search area As as shown in FIG. 1, the control device 32 ends the process.

3. If, in step S10 of FIG. 3, obstacles O1, O2, and O3 are present within the search area As as shown in FIG. 4, the control device 32 proceeds to step S11 of FIG.
In step S11, the control device 32 searches for whether or not there is an avoidable area in which the vehicle 20 can avoid the obstacle in the X direction, which is a horizontal direction intersecting the traveling direction. The avoidable area is a space in which the vehicle 20 can proceed in the Y direction while keeping the distance from the obstacle at a predetermined value or more when the vehicle 20 is lined up with the obstacle in the X direction. In other words, it is a space in which the vehicle 20 can pass beside the obstacle while keeping the distance from the obstacle at a predetermined value or more. The predetermined value is set to a value that can prevent contact between the obstacle and the vehicle 20, taking into consideration the detection accuracy of the sensor 31 and the possibility that the obstacle is a moving body.

The control device 32 determines that a space within the search area As whose dimension in the X-axis direction from the obstacle is equal to or greater than a threshold is an avoidable area. The threshold is set based on the dimension within which the vehicle 20 can travel while maintaining a distance between the vehicle 20 and the obstacle equal to or greater than a predetermined value.

In the example shown in FIG. 4, three obstacles O1, O2, and O3 are lined up in the X-axis direction. The central obstacle (hereinafter referred to as the central obstacle) O1 is located on the travel route R (on the reference axis B), an obstacle (hereinafter referred to as the right obstacle) O2 is located to the right of the central obstacle O1, and an obstacle (hereinafter referred to as the left obstacle) O3 is located to the left of the central obstacle O1. The width (distance) W1 in the X-axis direction between the central obstacle O1 and the right obstacle O2 is equal to or greater than a threshold value, and the space between the central obstacle O1 and the right obstacle O2 is the right avoidable area AA1. The width (distance) W2 between the central obstacle O1 and the left obstacle O3 is equal to or greater than a threshold value, and the space between the central obstacle O1 and the left obstacle O3 is the left avoidable area AA2.

Therefore, in the example shown in FIG. 4, the search area As has a right-side avoidable area AA1 formed between the central obstacle O1 and the right-side obstacle O2, and a left-side avoidable area AA2 formed between the central obstacle O1 and the left-side obstacle O3.

In the example shown in FIG. 4, the distance in the X direction from the vehicle to the center of the right-side avoidable area AA1 is L1. The distance in the X direction from the vehicle to the center of the left-side avoidable area AA2 is L2.

If there is no avoidable area in the search area As in step S11 of FIG. 3, the control device 32 ends the process. Furthermore, in step S12, the control device 32 determines whether or not there are multiple avoidable areas in the search area As. If there is only one avoidable area in the search area As, the control device 32 proceeds to step S15 and performs an avoidance action by changing the vehicle to travel through the avoidable area.

When a plurality of avoidable regions exist within the search region As in step S12 (see FIGS. 4, 6 and 8), the control device 32 performs the process of step S13.
In step S13, the control device 32 calculates the average travel speed for evaluating each avoidable area based on the speed permitted for returning to the original travel route through the avoidable area. In other words, when it is determined that there are multiple avoidable areas, each avoidable area is evaluated using the average travel speed (the speed permitted for returning to the original travel route through the avoidable area) as the evaluation value.

FIG. 5 is a table of conditions and evaluations used in the example shown in FIG.
In Fig. 5, the conditions are the width W of the avoidable area and the distance L in the X direction from the vehicle to the avoidable area. The average running speed is specified as an evaluation for each condition (W value, L value). For example, when W = 0.8 m, L = 0.1 m, the average running speed is 3 km/h. When W = 0.8 m, L = 0.2 m, the average running speed is 2 km/h. When W = 0.8 m, L = 0.3 m, the average running speed is 1.5 km/h.

When W = 1.2 m and L = 0.1 m, the average running speed is 7 km/h. When W = 1.2 m and L = 0.2 m, the average running speed is 6 km/h. When W = 1.2 m and L = 0.3 m, the average running speed is 5 km/h.

In this way, the control device 32 determines the average driving speed (the speed permitted for passing through the avoidable area and returning to the original driving route) based on the width W of the avoidable area in the direction intersecting the traveling direction, and the distance L from the vehicle to the avoidable area. The higher the average driving speed, the higher the evaluation.

In the example of Figure 4, the right-side avoidable area AA1 has W1 = 0.8 m and L1 = 0.1 m, and as shown in Figure 5, the average driving speed in the right-side avoidable area AA1 is 3 km/h. Also, in the example of Figure 4, the left-side avoidable area AA2 has W2 = 1.2 m and L2 = 0.3 m, and as shown in Figure 5, the average driving speed in the left-side avoidable area AA2 is 5 km/h. Therefore, the average driving speed in the left-side avoidable area AA2 is higher than that in the right-side avoidable area AA1, and is therefore rated higher.

The control device 32 selects an avoidable area with a high evaluation in step S14 in FIG. 3. For example, in the example in FIG. 4, the left-side avoidable area AA2, which has a high average driving speed, is selected as the avoidable area with a high evaluation. In other words, the right-side avoidable area AA1 is close to the vehicle, but has a narrow width W, so the average driving speed is slow. Therefore, the left-side avoidable area AA2 is selected.

In the case of Fig. 6 instead of Fig. 4, the following is done: In the case of Fig. 6, the table of conditions and evaluations shown in Fig. 7 is used instead of Fig. 5.
In the case of Fig. 6, the left avoidable area AA2 between the central obstacle O1 and the left obstacle O3 is a speed limit area of 1 km/h. Therefore, when W = 1.2 m and L = 0.3 m in Fig. 7, the average traveling speed is not 5 km/h, but is ultimately set to 1 km/h due to the speed limit area.

In this way, the control device 32 determines the average driving speed (the speed permitted for passing through the avoidable area to return to the original driving route) based on the width W of the avoidable area in the direction intersecting the direction of travel, the distance L from the vehicle to the avoidable area, and the speed limit in the speed limit area on the route to return to the driving route.

In the example of Figure 6, the right-side avoidable area AA1 has W1 = 0.8 m and L1 = 0.1 m, and as shown in Figure 7, the average driving speed in the right-side avoidable area AA1 is 3 km/h. Also, in the example of Figure 6, the left-side avoidable area AA2 has W2 = 1.2 m and L2 = 0.3 m, and as shown in Figure 7, the average driving speed in the left-side avoidable area AA2 is 1 km/h. Therefore, the average driving speed in the right-side avoidable area AA1 is higher than that in the left-side avoidable area AA2, and is rated higher.

Therefore, when the control device 32 selects an avoidable area with a high evaluation in step S14 of FIG. 3, in the example of FIG. 6, it selects the right-side avoidable area AA1, which has a high average driving speed, as the avoidable area with a high evaluation.

In the case of Fig. 8 instead of Fig. 4, the following is done: In the case of Fig. 8, the table of conditions and evaluations shown in Fig. 9 is used instead of Fig. 5.
In the case of Fig. 8, a stop sign Mst is installed in the left avoidable area AA2 between the central obstacle O1 and the left obstacle O3, and a stop line Lst is drawn on the road surface. As shown in Fig. 10, deceleration starts at timing t1, the vehicle stops at timing t2, and the period from t2 to t3 is the temporary stop period. Acceleration starts at timing t3, and the period from t4 to t5 is the period during which the vehicle travels at a low speed. Acceleration starts at t5, and the vehicle travels at a normal speed from t6 onwards.

As a result, when W=1.2 m and L=0.3 m in FIG. 9, the average traveling speed is not 5 km/h, but is reduced to 2 km/h by the operation of the intersection control.
In this way, when the control device 32 includes an intersection control section with a stop sign Mst, it determines the average driving speed (the speed permissible for passing through the avoidable area to return to the original driving route) based on the width W of the avoidable area in the direction intersecting the direction of travel, the distance L from the vehicle to the avoidable area, and whether the route to return to the driving route includes a section where the vehicle must stop.

In the example of Figure 8, the right-side avoidable area AA1 has W1 = 0.8 m and L1 = 0.1 m, and as shown in Figure 9, the average driving speed in the right-side avoidable area AA1 is 3 km/h. Also, in the example of Figure 8, the left-side avoidable area AA2 has W2 = 1.2 m and L2 = 0.3 m, and as shown in Figure 9, the average driving speed in the left-side avoidable area AA2 is 2 km/h. Therefore, the average driving speed in the right-side avoidable area AA1 is higher than that in the left-side avoidable area AA2, and is therefore rated higher.

Therefore, when the control device 32 selects an avoidable area with a high evaluation in step S14 of FIG. 3, in the example of FIG. 8, it selects the right-side avoidable area AA1, which has a high average driving speed, as the avoidable area with a high evaluation.

The tables shown in Figs. 5, 7, and 9 are stored in the memory unit 34, and the average traveling speed is determined from at least the width W of the avoidable area and the distance L in the X direction from the vehicle to the avoidable area. The search area As is an area with a limited width (size), and if the obstacles O1, O2, and O3 extend in the depth direction (forward in the direction of travel), it is not possible to determine how long the vehicle will continue to avoid the obstacles. In other words, it is not known how long the vehicle will need to continue to avoid the obstacles. If the obstacles O1, O2, and O3 extend far in the depth direction (forward in the direction of travel), the vehicle will need to continue to avoid the obstacles for a longer distance and for a longer period of time. Therefore, the average traveling speed is used for evaluation under the idea that if the traveling speed is high, there is a possibility that the obstacles can be reached in an earlier time. In other words, if the average traveling speed is high, there is a possibility that the time available for passing through will be shorter.

The map stores routes that can be traveled when there are no obstacles (stationary objects, moving objects such as people, etc.). Information such as whether the area is a speed limit area or whether an intersection is approaching is stored as map information. There are no obstacles on the map, and obstacles are detected by sensor 31. The direction of vehicle 20 travelling at the average travel speed is indicated by the outlined arrow in Figs. 4, 6 and 8. The average travel speed is used because, for example, in Fig. 4, straight travel approaching the centre of left avoidable area AA2, straight travel passing between obstacles O1 and O3, and straight travel returning to the original travel route R are relatively fast travellable, while turning travel between these straight travellables is slow travell. The average speed during this series of travels is the average travel speed.

In FIG. 10, the intersection control involves slowing down after a temporary stop, but as shown in FIG. 11, in the case where only a temporary stop is performed, that is, deceleration begins at timing t10 and the vehicle stops at timing t11, with t11 to t12 being the temporary stop period. Acceleration begins at timing t12, and the vehicle travels at the normal driving speed from t13 onwards. In addition, in the case where only slowing down is performed as shown in FIG. 12, that is, deceleration begins at timing t20, and the vehicle travels at a slow speed from t21 to t22 being the slow driving period. Acceleration begins at timing t22, and the vehicle travels at the normal driving speed from t23 onwards. In essence, the permissible speed is determined based on whether the route to return to the travel route includes a section where at least one of a temporary stop and slowing down is performed.

After selecting the area with the high evaluation in step S14 in FIG. 3, in step S15, the control device 32 causes the vehicle 20 to perform an avoidance action by traveling through the avoidable area with the high evaluation among the multiple avoidable areas. In other words, the control device 32 controls the vehicle 20 to move toward the avoidable area.

When performing evasive maneuvers, the traveling direction of the vehicle 20 and the direction of the travel route R no longer coincide, and the traveling direction of the vehicle 20 becomes inclined relative to the direction of the travel route R. The control device 32 changes the orientation of the search area As to match the orientation of the travel route R.

As shown in FIG. 3, when the control device 32 finishes the process of step S15, it returns the vehicle 20 to the driving route R in step S16 and ends the process. In detail, if an evasive action is being taken in step S15, the control device 32 controls the drive mechanism 41 so that the vehicle 20 returns to the driving route R. As a result, the X coordinate of the vehicle 20 approaches the X coordinate of the driving route R, and the vehicle 20 returns to the driving route R. Also, if an evasive action is not being taken, the vehicle 20 maintains a state of traveling on the driving route R.

As described above with reference to Figures 4 to 12, when selecting an avoidable area, the width W of the avoidable area, the distance L from the vehicle to the avoidable area, areas where speed changes (speed limit areas, intersections where the vehicle must stop), etc. are taken into consideration. Then, the average driving speed required to pass through each avoidable area is calculated as an evaluation value, and an avoidable area with the highest driving speed during avoidance is selected, allowing for efficient avoidance operations.

For example, as explained using Figures 4 and 5, when selecting an avoidable area based on the size (width W) of the avoidable area, the narrow avoidable area AA1 on the right side of Figure 4 is close to the vehicle, but because the avoidable area AA1 is narrow, the passing speed is slow.

Here, typical driving control involves decelerating if there is an obstacle ahead in the direction of travel, and stopping if there is an obstacle immediately in front, but with this type of control, the speed tends to decrease when passing through a narrow space. Also, while the table in Figure 5 is calculated in advance, if the relationship for finding the average driving speed under the conditions of the W and L values can be expressed by a formula, it may be calculated in real time. In this case, while the average driving speed in Figure 5 is a discontinuous value, it is possible to obtain the average driving speed as a continuous value by interpolation.

As explained using Figures 6 and 7, when selecting an avoidable area based on the speed limit area, the avoidable area AA2 on the left side of Figure 4 was evaluated the highest in Figure 5, but if it is in a speed limit area as in Figure 6, the average driving speed will decrease. Therefore, since the driving speed will not increase even if it is wide, it is more time efficient to pass through the narrow avoidable area AA1 on the right side of Figure 6.

As explained using Figures 8 and 9, when intersection control is required, the vehicle stops at the location, slows down for a certain period of time, and then resumes driving. Since the vehicle stops temporarily, if obstacles O1 and O3 are placed on both sides making it difficult to see to the left and right, i.e., if obstacles O1 and O3 are placed on the stop line Lst drawn on the map, the side of the stop line Lst in front of it is a sidewalk, and slowing down is required to prevent the vehicle from suddenly running out into the road. The left avoidable area AA2 in Figure 4 was the most highly rated in Figure 5, but if a temporary stop is required, the average driving speed drops. Therefore, since the speed does not increase even if the area is wider, it is more time efficient to pass through the narrow avoidable area AA1 on the right.

According to the above embodiment, the following effects can be obtained.
(1) The autonomous vehicle 10 includes a vehicle 20 as a vehicle, a sensor 31 mounted on the vehicle 20, and a control device 32. When the detection result of the sensor 31 indicates that obstacles O1, O2, and O3 exist within a search area As extending in the direction of the path along which the vehicle 20 travels, the control device 32 as a search unit performs the following procedure. The control device 32 searches for whether or not there are multiple avoidable areas AA1 and AA2 that can be avoided by the vehicle 20 in the X direction (in a broader sense, a direction intersecting the direction of travel in the horizontal direction) from the obstacles O1, O2, and O3. When it is determined that multiple avoidable areas AA1 and AA2 exist, the control device 32 as an evaluation unit evaluates each of the avoidable areas AA1 and AA2 based on an average travel speed as a speed permitted for returning to the original travel path through the avoidable areas AA1 and AA2. The control device 32, which serves as an avoidance driving execution unit, selects an avoidance possible area that is highly evaluated from among a plurality of avoidance possible areas, and causes the vehicle 20 to take evasive action by driving through the avoidance possible area.

According to this, the avoidable areas AA1 and AA2 are evaluated based on the average driving speed as the speed permitted for returning to the original driving route through the avoidable areas AA1 and AA2, and the vehicle 20 takes evasive action by selecting the avoidable area with the highest evaluation from among the multiple avoidable areas and driving through that avoidable area.

Therefore, when evasive action is taken to simply head toward the closest avoidable area from the vehicle, even if the width of the avoidable area is narrow or there is a speed limit in the avoidable area, the area closest to the vehicle will be selected, which may result in a longer arrival time at the destination, whereas with this method, efficient avoidance action can be taken.

(2) Specifically, the control device 32 as an evaluation unit determines the average driving speed as an allowable speed based on the width W of the avoidable area in a direction intersecting the traveling direction and the distance L from the vehicle to the avoidable area. Alternatively, the control device 32 as an evaluation unit determines the average driving speed as an allowable speed based on the width W of the avoidable area in a direction intersecting the traveling direction, the distance L from the vehicle to the avoidable area, and the speed limit of the speed limit area on the route to return to the traveling route. Alternatively, the control device 32 as an evaluation unit determines the average driving speed as an allowable speed based on the width W of the avoidable area in a direction intersecting the traveling direction, the distance L from the vehicle to the avoidable area, and the fact that the route to return to the traveling route includes a section where at least one of a temporary stop and slowing down is performed. This makes it possible to handle various cases.

The embodiment is not limited to the above, and may be embodied as follows, for example.
If the road is gravel, the tires will dig into the gravel and it will be difficult to gain speed. Taking this into consideration, the average driving speed can be lowered by increasing the number of judgment conditions according to the road conditions (gravel road, etc.). Gravel roads can also be detected as map information using a dedicated sensor. This allows the vehicle to avoid gravel roads and improve the situation where the vehicle slows down.

The avoidable area may be selected by determining whether the object is stationary or moving (such as a human) and taking into consideration the type of obstacle on either side of the avoidable area. For example, in a situation where a person is walking in a way that narrows the avoidable area, the avoidable area in the direction the person is facing will become narrower in the future and the average running speed will become slower, so this can be improved by taking this into consideration and lowering the average running speed. The type of obstacle can be detected using sensor 31, which is a laser sensor, and the direction of the person's feet can also be detected by sensor 31.

○ The obstacle avoidance process shown in FIG. 3 is repeated while the vehicle 20 is traveling, but the results of the previous process may be reflected in the current process, taking into account factors such as the depth of the obstacle in the previous process. In other words, the previous passing conditions may also be reflected. This can improve errors in the first evaluation (calculation). If the driving conditions in the previous run are known (for example, if the passing time was long), and it is known that the current run, which is the second run, will be slower than the average driving speed in the table, the previous average driving speed is reflected, and if another route is faster, the other route is selected for the current run. For example, this is useful when the obstacle is long in the forward direction (depth direction).

In addition, for evaluation tables such as Figure 5, the theoretically calculated table can be adjusted to suit the actual environment by updating the table values using deep learning or the like using data from actual driving in the past. This makes it possible to reflect the effects of actual disturbances such as those in a factory (e.g., driving on a gravel road).

The distance from the vehicle to the center of the avoidable area has been shown as an example, but it is not limited to the center and may be a location a specified distance away from the closer obstacle. For example, in FIG. 13, which replaces FIG. 4, the distances α and β to locations P10 and P11, which are 1 m away from obstacle O1 for avoidable areas AA1 and AA2, may be set as the distances to the avoidable areas.

○ A stereo camera may be used as the sensor 31. The stereo camera allows the control device 32 to recognize the surrounding environment from a disparity image obtained by capturing images of the surrounding environment using multiple cameras. A disparity image shows pixel differences that occur between cameras when multiple cameras capture images of the same feature point. A feature point is a part where parallax is obtained, such as the edge of an object. The control device 32 can measure the distance to the feature point from the disparity image. The control device 32 can detect an obstacle from a point cloud, which is a collection of feature points.

The sensor 31 may be an ultrasonic sensor that measures distance by emitting ultrasonic waves.
The wheels 22 may be wheels other than omnidirectional wheels, i.e., wheels that do not allow movement in the direction of the rotation axis of the wheels 22. In this case, the traveling direction of the vehicle 20 may be changed by using two-wheel speed difference control in which the rotation speeds of the two wheels are made different to perform steering. Alternatively, an individual steering mechanism may be provided for each wheel, and the traveling direction may be changed by individually steering each wheel.

The autonomous vehicle 10 is not limited to a transport vehicle that transports loads, and may be an autonomous vacuum cleaner or the like.
The running object is not limited to a vehicle that runs on wheels and may be, for example, a multi-legged running object.

10...autonomous vehicle, 20...vehicle, 31...sensor, 32...controller, As...search area, AA1, AA2...avoidable area, L...distance, O1, O2, O3...obstacle, R...travel path, W...width.

Claims (3)

A running body,
A sensor mounted on the traveling body;
a search unit that, when an obstacle is present within a search area extending in a direction of a path along which the moving object travels based on a detection result of the sensor, searches for whether or not there are a plurality of avoidable areas in a horizontal direction intersecting the traveling direction from the obstacle, in which the moving object can be avoided;
an evaluation unit that, when the search unit determines that a plurality of avoidable areas exist, evaluates each avoidable area based on an allowable speed for a series of travels that pass through the avoidable areas and return to the original travel route;
an avoidance travel execution unit that selects an avoidance possible area that is highly evaluated by the evaluation unit from among the plurality of avoidance possible areas, and causes the traveling object to take an avoidance action by traveling through the avoidance possible area ,
The evaluation unit determines, for each of the multiple avoidable areas, an average driving speed that is an average of the speeds allowable for the series of drives based on the width in a direction intersecting the direction of travel and the distance from the vehicle to the avoidable area, and evaluates each avoidable area so that the higher the average driving speed, the higher the evaluation . The autonomous vehicle according to claim 1, characterized in that the evaluation unit determines the average driving speed based on a width of the avoidable area in a direction intersecting the traveling direction, a distance from the vehicle to the avoidable area, and a speed limit in a speed limited area on a path to return to the driving path. The autonomous vehicle according to claim 1, characterized in that the evaluation unit determines the average traveling speed based on a width of the avoidable area in a direction intersecting the traveling direction, a distance from the vehicle to the avoidable area, and a route for returning to the traveling route including a section in which at least one of a temporary stop and slowing down is performed.