Mobile object control method, mobile object control device, and storage medium

1. A mobile object control method comprising:

recognizing physical objects near a mobile object and a route shape;

generating a target trajectory along which the mobile object is to travel autonomously based on a result of the recognition;

causing the mobile object to travel autonomously along the target trajectory; and

determining that an abnormality has occurred in a control system for causing the mobile object to travel autonomously by performing the recognition when a degree of deviation between a reference target trajectory and the target trajectory is greater than or equal to a first reference degree and output a determination result,

wherein the generation of the target trajectory includes:

setting the reference target trajectory based on a center of a lane in which the mobile object exists;

setting a first risk on coordinates in an assumed plane ahead of the mobile object based on existence of the physical objects, the first risk is set larger as the coordinates are closer to the physical objects and the first risk is set with respect to a current time point and future time points, the first risk with respect to future time points being predicted based on a change in the position of the physical objects;

setting a second risk on the coordinates in the assumed plane that is lowest at a position of the reference target trajectory, and increases as a distance between the coordinates and the reference target trajectory increases; and

generating the target trajectory based on the first risk and the second risk,

wherein, the determining that an abnormality has occurred in a control system is stopped when the aggregated values of first risks at trajectory points of the target trajectory with respect to a reference section is greater than or equal to a threshold value.

2. The mobile object control method according to claim 1 , further comprising:

calculating deviations between coordinate points that are on the reference target trajectory and the target trajectory, and correspond to the same point in a traveling direction of the mobile object; and

calculating a weighted sum of the deviations as the degree of deviation, wherein the weight for one deviation is set larger when the deviation is larger than adjacent deviations.

3. The mobile object control method according to claim 1 ,

wherein the generating includes setting a risk that is an index value representing a degree at which the mobile object should not approach based on the presence of the recognized physical object in an assumed plane represented in a two-dimensional plane when a space near the mobile object is viewed from above and generates the target trajectory so that the mobile object passes through a point where the risk is low, and

wherein the method further comprises stopping the determining when a degree of risk based on a risk value due to the presence of the physical object at each point of the target trajectory is higher than or equal to a second reference degree.

4. The mobile object control method according to claim 1 , further comprising:

acquiring surrounding environment information of the mobile object; and

making a probability of the determining lower when the environment information satisfies a predetermined condition.

5. The mobile object control method according to claim 1 , further comprising:

acquiring a speed of the mobile object; and

making a probability of the determining lower when the speed is higher than a reference speed.

6. The mobile object control method according to claim 1 , further comprising:

collecting data for each speed range of the mobile object; and

wherein the determining is performed for each speed range of the mobile object.

7. The mobile object control method according to claim 1 , further comprising:

collecting data for each speed range of the mobile object,

wherein

the determining of the abnormality is performed for each speed range of the mobile object, and

the determining of the abnormality includes determining that the abnormality has occurred in the control system when an abnormality has occurred in relation to two or more speed ranges.

8. A mobile object control device comprising:

a storage device storing a program; and

a hardware processor,

wherein the hardware processor executes the program stored in the storage device to

recognize physical objects near a mobile object and a route shape;

generate a target trajectory along which the mobile object is to travel autonomously based on a result of the recognition;

cause the mobile object to travel autonomously along the target trajectory; and

determine that an abnormality has occurred in a control system for causing the mobile object to travel autonomously by performing the recognition when a degree of deviation between a reference target trajectory and the target trajectory is greater than or equal to a first reference degree and output a determination result, wherein the generation of the target trajectory includes:

setting the reference target trajectory based on a center of a lane in which the mobile object exists;

setting a first risk on coordinates in an assumed plane ahead of the mobile object based on existence of the physical objects, wherein the first risk is set to be larger as the coordinates are closer to the physical objects and the first risk is set with respect to a current time point and future time points, the first risk with respect to future time points being predicted based on a change in the position of the physical objects;

setting a second risk on the coordinates in the assumed plane that is lowest at a position of the reference target trajectory, and increases as a distance between the coordinates and the reference target trajectory increases; and

generating the target trajectory based on the first risk and the second risk,

wherein, the determining that an abnormality has occurred in a control system is stopped when the aggregated values of first risks at trajectory points of the target trajectory with respect to a reference section is greater than or equal to a threshold value.

9. A computer-readable non-transitory storage medium storing a program to be executed by a computer to:

recognize physical objects near a mobile object and a route shape;

generate a target trajectory along which the mobile object is to travel autonomously based on a result of the recognition;

cause the mobile object to travel autonomously along the target trajectory; and

determine that an abnormality has occurred in a control system for causing the mobile object to travel autonomously by performing the recognition when a degree of deviation between a reference target trajectory and the target trajectory is greater than or equal to a first reference degree and output a determination result,

wherein the generation of the target trajectory includes:

setting the reference target trajectory based on a center of a lane in which the mobile object exists;

setting a first risk on coordinates in an assumed plane ahead of the mobile object based on existence of the physical objects, wherein the first risk is set to be larger as the coordinates are closer to the physical objects and the first risk is set with respect to a current time point and future time points, the first risk with respect to future time points being predicted based on a change in the position of the physical objects;

setting a second risk on the coordinates in the assumed plane that is lowest at a position of the reference target trajectory, and increases as a distance between the coordinates and the reference target trajectory increases; and

generating the target trajectory based on the first risk and the second risk,

wherein, the determining that an abnormality has occurred in a control system is stopped when the aggregated values of first risks at trajectory points of the target trajectory with respect to a reference section is greater than or equal to a threshold value.