CN110895406B

CN110895406B - Method and device for testing unmanned equipment based on interferent track planning

Info

Publication number: CN110895406B
Application number: CN201911096978.1A
Authority: CN
Inventors: 白钰; 任冬淳; 许笑寒; 李潇
Original assignee: Beijing Sankuai Online Technology Co Ltd
Current assignee: Beijing Sankuai Online Technology Co Ltd
Priority date: 2019-11-11
Filing date: 2019-11-11
Publication date: 2021-02-02
Anticipated expiration: 2039-11-11
Also published as: CN110895406A

Abstract

The specification discloses a method and a device for testing unmanned equipment based on interferent track planning, wherein the acquired historical motion state of an interferent and the current motion state of the unmanned equipment are input into a planning model to obtain a planned track of the interferent, when the interferent runs along the planned track, the current motion state of the unmanned equipment is interfered, the problem that whether the interferent can interfere with the operation of the unmanned equipment or not is solved, the planned track is sent to the interferent, the interferent runs along the planned track, the behavior action planned by the unmanned equipment is determined, the unmanned equipment is tested according to the behavior action, and the testing efficiency of the unmanned equipment is greatly improved.

Description

Method and device for testing unmanned equipment based on interferent track planning

Technical Field

The application relates to the technical field of testing unmanned equipment, in particular to an unmanned equipment testing method and device based on interferent track planning.

Background

Before the unmanned equipment is formally put into use, the unmanned equipment needs to be tested, and whether the unmanned equipment can be put into use or not is judged according to a test result.

When testing the unmanned device, a common method is to set a plurality of interferents (generally including vehicles, pedestrians, and the like), and make the interferents operate according to a preset behavior rule, and when the interferents may interfere with the operation of the unmanned device, observe the operation condition of the unmanned device, so as to test the unmanned device, therefore, the operation of the interferents is very important for the test of the unmanned device.

The operation of setting the interfering object is generally performed by operating the interfering object at a preset speed according to a preset route, or by specifying a start point and an end point of the interfering object and randomly generating a running speed and a running route of the interfering object by a computer program. However, by setting the operation of the interfering object by the two methods, it cannot be determined whether the interfering object interferes with the operation of the unmanned aerial vehicle, so that the efficiency of testing the unmanned aerial vehicle is extremely low.

Disclosure of Invention

The embodiment of the specification provides a method and a device for testing unmanned equipment based on interferent track planning, so as to partially solve the problems in the prior art.

The embodiment of the specification adopts the following technical scheme:

the specification provides a method for testing unmanned equipment based on interferent track planning, which comprises the following steps:

acquiring a historical motion state of an interfering object and a current motion state of the unmanned equipment;

inputting the historical motion state of the interferent and the current motion state of the unmanned equipment into a pre-trained planning model to obtain a planning track output by the planning model, wherein the planning track is used for generating interference on the current motion state of the unmanned equipment when the interferent runs along the planning track;

sending the planned track to the interferent so that the interferent runs according to the planned track;

and determining the behavior action planned by the unmanned equipment under the condition that the interferent runs according to the planned track, and testing the unmanned equipment according to the behavior action.

Optionally, the pre-training of the planning model specifically includes:

acquiring a historical motion state of the interfering object and a historical motion state of the unmanned equipment as training samples;

inputting the obtained training samples into a planning model to be trained to obtain a predicted planning track output by the planning model to be trained;

determining the interference degree of the estimated planning track on the unmanned equipment;

and training the planning model to be trained by taking the maximum interference degree as a training target.

Optionally, determining the interference degree of the pre-estimated planned trajectory on the unmanned aerial vehicle specifically includes:

determining an estimated speed change value, an estimated acceleration change value and an estimated direction angle change value of the unmanned equipment when the interferent runs along the estimated planning track;

and determining the interference degree of the interferent on the historical motion state of the unmanned equipment according to at least one of the determined estimated speed change value, the estimated acceleration change value and the estimated direction angle change value.

Optionally, the planning model to be trained is a reinforcement learning model;

training the planning model to be trained by taking the maximum interference degree as a training target, specifically comprising:

determining the type of the estimated planning track according to the interference degree;

and training the planning model to be trained by taking the maximum reward as a training target according to the preset rewards corresponding to all types.

Optionally, the obtaining of the historical motion state of the interfering object specifically includes:

aiming at each type, selecting a preset number of estimated planning tracks from the estimated planning tracks of the type output by the planning model to be trained in the last training process;

and acquiring the motion state of the interference object running according to the selected estimated planning track, and taking the motion state as at least part of the historical motion state of the interference object acquired in the current training process.

Optionally, in the last training process, selecting a preset number of estimated planning trajectories from the estimated planning trajectories of the type output by the planning model to be trained, specifically includes:

and selecting a preset number of estimated planning tracks from the estimated planning tracks of the type output by the planning model to be trained in the last training process according to the total amount of the estimated planning tracks to be selected and the preset proportion for the type.

Optionally, determining a behavior action planned by the unmanned aerial vehicle when the interfering object travels according to the planned trajectory, and testing the unmanned aerial vehicle according to the behavior action specifically includes:

determining a behavior action planned by the unmanned equipment under the condition that at least one interfering object runs according to the corresponding planned track;

determining a motion state of the unmanned equipment according to the behavior action;

and testing the unmanned equipment according to the determined motion state of the unmanned equipment.

The present specification provides a testing apparatus for an unmanned aerial vehicle based on interferent trajectory planning, the apparatus comprising:

the acquisition module is used for acquiring the historical motion state of the interferent and the current motion state of the unmanned equipment;

the output module is used for inputting the historical motion state of the interferent and the current motion state of the unmanned equipment into a planning model trained in advance to obtain a planning track output by the planning model, and the planning track is used for generating interference on the current motion state of the unmanned equipment when the interferent runs along the planning track;

the sending module is used for sending the planned track to the interferent so that the interferent can run according to the planned track;

and the testing module is used for determining the behavior action planned by the unmanned equipment under the condition that the interferent runs according to the planned track, and testing the unmanned equipment according to the behavior action.

The storage medium is characterized by storing a computer program, and the computer program is executed by a processor to implement the above method for testing the unmanned aerial vehicle based on the interferent trajectory planning.

The electronic device provided by the specification comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, and is characterized in that the processor executes the program to realize the method for testing the unmanned device based on the interferent track planning.

The embodiment of the specification adopts at least one technical scheme which can achieve the following beneficial effects:

the method includes the steps that the acquired historical motion state of the interferent and the current motion state of the unmanned equipment are input into a planning model, a planning track of the interferent is obtained, when the interferent runs along the planning track, the interference is generated on the current motion state of the unmanned equipment, the problem that whether the interferent can generate interference on the operation of the unmanned equipment or not cannot be determined is solved, the planning track is sent to the interferent, the interferent runs along the planning track, the behavior action planned by the unmanned equipment is determined, the unmanned equipment is tested according to the behavior action, and the testing efficiency of the unmanned equipment is greatly improved.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:

fig. 1 is a flowchart of a method for testing an unmanned aerial vehicle based on an interferent trajectory planning, provided in an embodiment of the present disclosure;

fig. 2 is a schematic diagram of an interference object that interferes with the operation of the unmanned aerial vehicle when the interference object travels along a planned trajectory at a certain time according to an embodiment of the present disclosure;

fig. 3 is a schematic diagram of interference caused to the operation of the unmanned aerial vehicle when at least one interfering object travels according to a planned trajectory at a certain time according to an embodiment of the present disclosure;

FIG. 4 is a flowchart of a method for pre-training a planning model provided in an embodiment of the present disclosure;

FIG. 5 is a flow chart of another method for pre-training a planning model provided in an embodiment of the present disclosure;

fig. 6 is a schematic structural diagram of a testing apparatus of an unmanned aerial vehicle based on interferent trajectory planning according to an embodiment of the present disclosure;

fig. 7 is a schematic diagram of an electronic device corresponding to fig. 1 provided in an embodiment of the present disclosure.

Detailed Description

In order to make the objects, technical solutions and advantages of the present disclosure more apparent, the technical solutions of the present disclosure will be clearly and completely described below with reference to the specific embodiments of the present disclosure and the accompanying drawings. It should be apparent that the described embodiments are only some of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments in the present specification without any inventive step are within the scope of the present application.

The technical solutions provided by the embodiments of the present application are described in detail below with reference to the accompanying drawings.

Fig. 1 is a flowchart of a method for testing an unmanned aerial vehicle based on an interferent trajectory planning, which may specifically include the following steps:

s100: and acquiring the historical motion state of the interference object and the current motion state of the unmanned equipment.

The method for testing the unmanned aerial vehicle based on the interferent track planning, provided by the specification, can be applied to a real unmanned aerial vehicle test scene and a simulated unmanned aerial vehicle test scene, one or more interferents can be set in the test scene when the unmanned aerial vehicle is tested, the planned track is sent to the interferents, the interferents are made to run according to the planned track, so that the behavior action planned by the unmanned aerial vehicle under the condition that the interferents run according to the planned track is determined, and the unmanned aerial vehicle is tested according to the behavior action.

In a real unmanned device test scenario, the interferent may include a real vehicle (which may include conventional vehicles such as buses and trucks), a pedestrian, a movable obstacle with intelligent computation, and the like.

In the simulated test scene of the unmanned aerial vehicle, the interferent may include a simulated vehicle, a simulated pedestrian, and the like, and in order to better obtain the test effect of the unmanned aerial vehicle, generally, the setting of the interferent information is based on a real vehicle and a real pedestrian, for example, when the interferent is assumed to be a simulated pedestrian and the interferent information of the interferent is set based on the real pedestrian, the set movement speed (i.e., the interferent information) does not exceed the fastest movement speed of a human; for another example, if the interfering object is a simulated vehicle and the interfering object information of the interfering object is set with reference to a real vehicle, the set running speed (i.e., the interfering object information) does not exceed the fastest running speed of the vehicle.

For convenience of description, the following description will be made only by taking a simulated test scenario of the unmanned aerial vehicle as an example.

The method comprises the steps of obtaining historical motion states of the interference object, obtaining historical motion states of the interference object at the previous moment, and obtaining historical motion states of the interference object in a period of time before the current moment, wherein the motion states can comprise the running speed, the running acceleration, the running direction, the position information and the like of a vehicle. In the simulated test scene of the unmanned aerial vehicle, the historical motion state of the interfering object can be randomly generated historically, or can be the historical motion state when the interfering object runs according to a historical planning track.

In addition, when the historical motion state of the interfering object is obtained, the historical motion state of at least one interfering object may be obtained. For example, in practical application, vehicles are mainly used on an urban main road, so that when a scene for testing the unmanned equipment is the urban main road, simulated vehicles can be selected as interference objects as far as possible, and meanwhile, the acquired historical motion state of a real vehicle is used as a reference, and the interference object information of the interference objects is set; for another example, in practical application, pedestrians are mainly used near sidewalks, so that when the scene for testing the unmanned device is near the sidewalks, simulated pedestrians can be selected as much as possible as interferents, and meanwhile, the obtained historical motion state of the real pedestrians is used as a reference to set interferent information of the interferents. For other scenarios of testing the unmanned device, no further description is given here.

S102: inputting the historical motion state of the interferent and the current motion state of the unmanned equipment into a pre-trained planning model to obtain a planning track output by the planning model, wherein the planning track is used for generating interference on the current motion state of the unmanned equipment when the interferent runs along the planning track.

Through the step S100, the historical motion state of the interfering object and the current motion state of the unmanned aerial vehicle are obtained, the obtained information is input into the planning model, and the planning trajectory output by the planning model for the interfering object is obtained, and when the unmanned aerial vehicle is tested, the interfering object is caused to travel according to the planning trajectory to generate interference on the current motion state of the unmanned aerial vehicle, for example, the interfering object suddenly changes to a lane where the unmanned aerial vehicle is located, and for example, the interfering object in front of the unmanned aerial vehicle suddenly brakes, and the unmanned aerial vehicle needs to plan a trip as an action according to the motion state of the interfering object, so that the current motion state is correspondingly changed, and the test efficiency can be effectively improved.

In this specification, the planning trajectory output by the planning model may be used to generate interference on the current motion state of the unmanned aerial vehicle when the interfering object travels on the planning trajectory, because the planning model is trained in advance, and when the planning model is trained in advance, the planning model to be trained is trained by using the maximum interference generated on the operation of the unmanned aerial vehicle when the interfering object travels on the estimated planning trajectory output by the planning model to be trained, so that, when the trained planning model is applied to test the unmanned aerial vehicle, the historical motion state of the interfering object and the current motion state of the unmanned aerial vehicle are input into the trained planning model, and the trained planning model may output the planning trajectory in which the interfering object interferes with the current motion state of the unmanned aerial vehicle. The method for training the planning model in advance will be described in detail below.

As can be seen from the above, the planned trajectory is a trajectory that can interfere with the current motion state of the unmanned aerial vehicle when the interfering object travels along the planned trajectory, and among trajectories that can interfere with the current motion state of the unmanned aerial vehicle, there are a normal travel trajectory (for example, a trajectory that allows the vehicle to follow traffic regulations when traveling) and an abnormal travel trajectory (for example, a trajectory that allows the vehicle to turn around on the road) according to the traveling habits of human beings.

S104: and sending the planned track to the interferent so that the interferent runs according to the planned track.

The planned track of the interfering object is obtained through the step S102, the planned track is sent to the interfering object, when the interfering object travels according to the planned track, interference is generated on the operation of the unmanned device, for example, the interfering object travels on an adjacent lane of the unmanned device, when the interfering object travels according to the planned track, the interfering object will change lanes to the lane where the unmanned device is located, and the motion state of the interfering object and the motion state of the unmanned device at a certain time can be as shown in fig. 2.

Fig. 2 is a schematic diagram of interference generated by an interfering object when the interfering object travels according to a planned trajectory at a certain time according to an embodiment of the present specification, in fig. 2, a is the unmanned device, B is the interfering object, an arrow a indicates a motion state of the unmanned device a, where a direction of the arrow a indicates a traveling direction of the unmanned device, a length of the arrow a indicates a traveling speed of the unmanned device, an arrow B indicates a motion state of the interfering object B, and the direction and the length of the arrow B have the same function as those of the arrow a. As can be seen from fig. 2, when the interfering object B travels along the planned trajectory, it may interfere with the motion state of the unmanned aerial vehicle a.

S106: and determining the behavior action planned by the unmanned equipment under the condition that the interferent runs according to the planned track, and testing the unmanned equipment according to the behavior action.

When the interfering object is caused to travel along the planned track in step S104, the behavior action planned by the unmanned aerial vehicle may be determined, where the behavior action refers to actions such as acceleration, deceleration, lane change and the like performed by the unmanned aerial vehicle for ensuring normal travel and safe travel according to the travel condition of the interfering object, and corresponding actions are performed, for example, in fig. 2, since the interfering object B travels along the planned track to the lane where the unmanned aerial vehicle a is located, the unmanned aerial vehicle a may determine the behavior action of deceleration or acceleration according to the distance from the interfering object B, so as to avoid a traffic accident with the interfering object B.

Secondly, according to the behavior action, the motion state of the unmanned equipment can be determined, and according to the above example, when the distance between the unmanned equipment A and the interference object B is smaller than the preset distance threshold value, in order to avoid a traffic accident with the interference object B, the deceleration behavior action can be planned, so that the interference object B smoothly changes the road and the unmanned equipment A safely runs, and according to the deceleration behavior action planned by the unmanned equipment A, the motion state of the unmanned equipment A can be determined to be that the running direction is kept unchanged and the running speed is reduced. Of course, the robot a may plan an acceleration behavior or a lane change behavior, and determine the motion state in the behavior according to the specific behavior of the robot a.

And finally, testing the unmanned equipment according to the determined motion state of the unmanned equipment, and continuing to use the above example, when the unmanned equipment A plans a deceleration action, the determined motion state is that the driving direction is kept unchanged and the driving speed is reduced, and according to the determined motion state, when the interference object B changes the lane to the lane where the unmanned equipment A is located, observing whether the unmanned equipment A and the interference object B have traffic accidents or not, and testing the obstacle avoidance performance of the unmanned equipment A.

In this specification, when testing the unmanned aerial vehicle, a plurality of interferents may be set, and when at least one interferent is traveling according to a corresponding planned trajectory, a behavior action planned by the unmanned aerial vehicle is determined, a motion state of the unmanned aerial vehicle is determined according to the behavior action, and the unmanned aerial vehicle is tested according to the determined motion state of the unmanned aerial vehicle. Specifically, the historical motion states of the multiple interferents can be obtained, the planned trajectories of the multiple interferents can be obtained through the planning model, and when the multiple interferents travel according to the corresponding planned trajectories, interference is generated on the unmanned aerial vehicle, and the unmanned aerial vehicle is tested, as shown in fig. 3.

Fig. 3 is a schematic diagram of interference generated by at least one interfering object when the interfering object travels along a planned trajectory at a certain time according to an embodiment of the present disclosure. With respect to fig. 2, fig. 3 is added with an interfering object C, D, E, and arrows c, d, and e respectively show the motion state of the interfering object C, D, E at a certain time, and it can be seen from fig. 3 that when all the interfering objects B, C, D, E travel along the corresponding planned trajectory, interference occurs to the unmanned aerial vehicle a, and it can be determined that the behavior action planned by the unmanned aerial vehicle a is deceleration, and according to the planned behavior action, it can be determined that the motion state of the unmanned aerial vehicle a is that the traveling direction is kept unchanged and the traveling speed is reduced, so as to avoid occurrence of a traffic accident. In the scenes shown in fig. 2 and 3, the interferent runs according to the planned track, and effectively interferes with the operation of the unmanned equipment, so that the efficiency of testing the unmanned equipment is greatly improved.

As can be seen from the step S102, the planning model is pre-trained, so that the present specification provides a method for pre-training the planning model, as shown in fig. 4, fig. 4 is a flowchart of the method for pre-training the planning model provided in the embodiment of the present specification, and specifically includes the following steps:

s400: and acquiring the historical motion state of the interfering object and the historical motion state of the unmanned equipment as training samples.

Specifically, the related content in step S100 may be referred to, and the historical motion state of the interfering object and the historical motion state of the unmanned aerial vehicle may be obtained from a historical record of a real unmanned aerial vehicle test scenario or a simulated unmanned aerial vehicle test scenario.

S402: and inputting the obtained training samples into a planning model to be trained to obtain an estimated planning track output by the planning model to be trained.

S404: and determining the interference degree of the estimated planning track on the unmanned equipment.

The estimated planned trajectory of the interfering object is obtained through the step S402, and when the interfering object travels along the estimated planned trajectory, an estimated speed variation value, an estimated acceleration variation value, and an estimated direction angle variation value of the unmanned aerial vehicle may be determined. Specifically, since the information of the training sample is history information, it is assumed that the interfering object travels in a predicted planned track historically, and it is possible to determine a predicted behavior action planned by the unmanned device, determine a predicted motion state of the unmanned device traveling according to the predicted behavior action, compare the predicted motion state of the unmanned device traveling with a real historical motion state in a history record, and determine a predicted change degree of the unmanned device traveling, where the predicted change degree of the unmanned device may include information such as a predicted speed change value, a predicted acceleration change value, and a predicted direction angle change value.

And then, according to at least one of the estimated speed change value, the estimated acceleration change value and the estimated direction angle change value, determining the interference degree of the interference object on the historical motion state of the unmanned equipment. The interference degree may be numerical information or percentage information of the change of the historical motion state of the unmanned aerial vehicle, and specifically, the calculation method of the interference degree may be preset, and the numerical value or percentage of the interference degree of the estimated planned trajectory on the unmanned aerial vehicle may be determined according to at least one of the determined estimated speed change value, the estimated acceleration change value, and the estimated direction angle change value. When at least one of the estimated speed change value, the estimated acceleration change value and the estimated direction angle change value is larger, the interference degree of the interference object on the historical motion state of the unmanned equipment is larger under the condition that the interference object runs in the estimated planning track historically. Therefore, the estimated change degree of the unmanned aerial vehicle is positively correlated with the interference degree of the interfering object on the historical motion state of the unmanned aerial vehicle. That is, the larger the estimated degree of change of the unmanned aerial vehicle is, the larger the degree of interference of the interfering object with the historical motion state of the unmanned aerial vehicle is.

S406: and training the planning model to be trained by taking the maximum interference degree as a training target.

After the interference degree of the estimated planned trajectory on the unmanned aerial vehicle is determined through the step S404, the type of the estimated planned trajectory may be determined according to the interference degree. Specifically, the determined interference level may be classified, and it should be noted that the interference level classification types corresponding to different interferent types may not be consistent, but the classification methods are the same, for example, the interference degree can be divided into serious interference, slight interference, non-interference and other types, the interference degree value or percentage interval corresponding to each type can be preset, determining the type of the estimated planning track according to the value or percentage of the interference degree corresponding to the estimated planning track, inputting the interference degree corresponding to the estimated planning track into a classification model trained in advance to obtain the type of the interference degree output by the classification model, the classification model may be a random forest model, a Support Vector Machine (SVM), or the like in a Machine learning model, and details of the classification model are not repeated in this specification with respect to the pre-training.

In this specification, the to-be-trained planning model may be a Reinforcement Learning (RL) model, rewards corresponding to each type may be preset, and the to-be-trained RL model is trained by using a maximized reward as a training target according to the preset rewards corresponding to each type, that is, the to-be-trained RL model is trained by using a maximized interference degree as a training target according to the reward corresponding to the type to which the interference degree of the preset estimated planning trajectory belongs. For example, when the type of the interference degree is a severe interference type, the weight of the reward of the severe interference type is set to be the maximum, and since the greater the interference degree is, the higher the reward is, the training target is the maximum interference degree, that is, the training target is the maximum reward.

Of course, the planning model in this specification may also be other machine learning models, for example, a neural network model, when the planning model is the neural network model, losses corresponding to each type may be preset, and the neural network model to be trained is trained with the minimum loss as a training target according to the preset losses corresponding to each type, and details about a specific training method are not repeated in this specification.

The present specification further provides another method for pre-training a planning model, as shown in fig. 5, fig. 5 is a flowchart of another method for pre-training a planning model provided in an embodiment of the present specification, and specifically may include the following steps:

s500: and aiming at each type, selecting a preset number of estimated planning tracks from the estimated planning tracks of the type output by the planning model to be trained in the last training process.

S502: and acquiring the motion state of the interference object running according to the selected estimated planning track, and taking the motion state as at least part of the historical motion state of the interference object acquired in the current training process.

S504: and inputting the obtained training samples into a planning model to be trained to obtain an estimated planning track output by the planning model to be trained.

S506: and determining the interference degree of the estimated planning track on the unmanned equipment.

S508: and training the planning model to be trained by taking the maximum interference degree as a training target.

When training the pre-estimation model to be trained, the interference degree is maximized as a training target. The abnormal travel trajectory mentioned above (e.g., the trajectory of a vehicle turning its place on a road) may also cause interference with the operation of the unmanned equipment. And the meaning of the abnormal running track is not great when the unmanned equipment is tested. Therefore, when testing the unmanned equipment, the output planned track of the pre-estimated model after training is expected to interfere with the operation of the unmanned equipment, and the output planned track is avoided as much as possible to be an abnormal running track, so that the testing efficiency is improved, and meaningless tests are reduced. Therefore, in the current training process, according to the total amount of the estimated planning trajectory to be selected and the preset proportion for the type of the interference degree, a preset number of estimated planning trajectories are selected from the estimated planning trajectories of the type of the interference degree output by the planning model to be trained in the last training process, that is, according to the preset total amount of the estimated planning trajectory to be selected and the preset proportion of the type of each interference degree, the estimated planning trajectory is selected from the estimated planning trajectories of each type in the last training process, and the motion state corresponding to the selected estimated planning trajectory is used as at least part of the historical motion state in the current training process.

Specifically, when the planning model to be trained is trained, parameters of the planning model to be trained may be set. The parameters may include the number of training iterations, the number of training samples, the total amount of estimated planning trajectories required to be selected in the current training process, a learning rate, a forgetting rate, and the like. For example, the number of training iterations may be set to 5000, the number of training samples may be set to 200, the total number of estimated planning trajectories to be selected in the current training process may be set to 100, and for the complete process of 5000 iterative training of the estimated model to be trained, if the interference degree is divided into three types of severe interference, slight interference, and non-interference, since the interference degree is maximized as the training target, when the type of the interference degree is configured, the trajectory proportion of the severe interference type may be set to a larger proportion, and the trajectory proportion of the non-interference type may be set to a smaller proportion. For example, the ratio of the track occupation ratio of the severe interference type, the slight interference type and the non-interference type when the estimated planning track is selected is preset to be 7:2:1, and in the nth iterative training process (wherein N is less than or equal to 5000), 100 estimated planning tracks can be selected in proportion from the severe interference type, the slight interference type and the non-interference type in the estimated planning track obtained in the nth iterative training process (namely, 70 estimated planning tracks are selected from the severe interference type, 20 estimated planning tracks are selected from the slight interference type, and 10 estimated planning tracks are selected from the non-interference type). And aiming at the selected 100 predicted planned tracks, obtaining the motion state of the interferent running according to the predicted planned tracks, and taking the obtained 100 motion states as part of historical motion states in the Nth iterative training process. That is, in the nth iterative training process, 200 training samples include 100 training samples composed of motion states corresponding to the estimated planned trajectory obtained in the nth-1 st iterative training process.

After obtaining the historical motion state of the interfering object in the Nth iterative training process, performing iterative training on the estimation model to be trained according to the contents of the steps S504-S508. Since the contents of steps S504 to S508 correspond to the same contents of steps S402 to S406, the details are not repeated here.

The method for testing the unmanned aerial vehicle based on the interferent track planning, which is provided by the specification, is particularly applicable to the field of distribution by using the unmanned aerial vehicle, for example, a distribution scene of express delivery, takeout and the like by using the unmanned aerial vehicle.

Based on the method for testing the unmanned aerial vehicle based on the interferent trajectory planning shown in fig. 1, an embodiment of the present specification further provides a schematic structural diagram of a testing apparatus for the unmanned aerial vehicle based on the interferent trajectory planning, as shown in fig. 6.

Fig. 6 is a schematic structural diagram of a testing apparatus for an unmanned aerial vehicle based on interferent trajectory planning, provided in an embodiment of the present specification, where the apparatus includes:

an obtaining module 601, configured to obtain a historical motion state of an interfering object and a current motion state of the unmanned device;

an output module 602, configured to input a historical motion state of the interfering object and a current motion state of the unmanned aerial vehicle into a pre-trained planning model, so as to obtain a planning trajectory output by the planning model, where the planning trajectory is used to generate interference on the current motion state of the unmanned aerial vehicle when the interfering object travels along the planning trajectory;

a sending module 603, configured to send the planned trajectory to the interfering object, so that the interfering object travels according to the planned trajectory;

the testing module 604 is configured to determine a behavior action planned by the unmanned device when the interfering object travels according to the planned trajectory, and test the unmanned device according to the behavior action.

Optionally, the apparatus further comprises:

a training module 605, configured to obtain a historical motion state of the interfering object and a historical motion state of the unmanned aerial vehicle in advance, as training samples; inputting the obtained training samples into a planning model to be trained to obtain a predicted planning track output by the planning model to be trained; determining the interference degree of the estimated planning track on the unmanned equipment; and training the planning model to be trained by taking the maximum interference degree as a training target.

Optionally, the training module 605 is specifically configured to determine an estimated speed change value, an estimated acceleration change value, and an estimated direction angle change value of the unmanned aerial vehicle when the interfering object travels along the estimated planned trajectory; and determining the interference degree of the interferent on the historical motion state of the unmanned equipment according to at least one of the determined estimated speed change value, the estimated acceleration change value and the estimated direction angle change value.

Optionally, the planning model to be trained is a reinforcement learning model;

the training module 605 is specifically configured to determine a type to which the estimated planned trajectory belongs according to the interference degree; and training the planning model to be trained by taking the maximum reward as a training target according to the preset rewards corresponding to all types.

Optionally, the training module 605 is specifically configured to, for each type, select a preset number of estimated planning trajectories from the estimated planning trajectories of the type output by the planning model to be trained in the last training process; and acquiring the motion state of the interference object running according to the selected estimated planning track, and taking the motion state as at least part of the historical motion state of the interference object acquired in the current training process.

Optionally, the training module 605 is specifically configured to select a preset number of estimated planning trajectories from the estimated planning trajectories of the type output by the planning model to be trained in the last training process according to the total amount of the estimated planning trajectories to be selected and a preset ratio for the type.

Optionally, the testing module 604 is specifically configured to determine a behavior planned by the unmanned aerial vehicle when at least one interfering object travels according to a corresponding planned trajectory; determining a motion state of the unmanned equipment according to the behavior action; and testing the unmanned equipment according to the determined motion state of the unmanned equipment.

The present specification further provides a computer-readable storage medium, which stores a computer program, where the computer program is used to execute the method for testing the unmanned aerial vehicle based on the interferent trajectory planning provided in fig. 1.

Based on the method for testing the unmanned aerial vehicle based on the interferent trajectory planning shown in fig. 1, the embodiment of the present specification further provides a schematic structural diagram of the electronic device shown in fig. 7. As shown in fig. 7, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, but may also include hardware required for other services. The processor reads a corresponding computer program from the nonvolatile memory into the memory and then runs the computer program to implement the method for testing the unmanned equipment based on the interferent trajectory planning described in fig. 1.

Of course, besides the software implementation, the present specification does not exclude other implementations, such as logic devices or a combination of software and hardware, and the like, that is, the execution subject of the following processing flow is not limited to each logic unit, and may be hardware or logic devices.

In the 90 s of the 20 th century, improvements in a technology could clearly distinguish between improvements in hardware (e.g., improvements in circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements in process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually making an Integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as abel (advanced Boolean Expression Language), ahdl (alternate Hardware Description Language), traffic, pl (core universal Programming Language), HDCal (jhdware Description Language), lang, Lola, HDL, laspam, hardward Description Language (vhr Description Language), vhal (Hardware Description Language), and vhigh-Language, which are currently used in most common. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functions of the various elements may be implemented in the same one or more software and/or hardware implementations of the present description.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

As will be appreciated by one skilled in the art, embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

This description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.

The above description is only an example of the present specification, and is not intended to limit the present specification. Various modifications and alterations to this description will become apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present specification should be included in the scope of the claims of the present specification.

Claims

1. An interferent trajectory planning-based test method for unmanned equipment, the method comprising:

inputting the historical motion state of the interferent and the current motion state of the unmanned equipment into a pre-trained planning model to obtain a planning track output by the planning model, wherein the planning track is used for generating interference on the current motion state of the unmanned equipment when the interferent runs along the planning track, and the planning model to be trained is trained by taking the maximum interference generated on the running of the unmanned equipment as a training target when the interferent runs along the estimated planning track output by the planning model to be trained during pre-training of the planning model;

2. The method of claim 1, wherein pre-training the planning model specifically comprises:

3. The method of claim 2, wherein determining the degree of interference of the pre-estimated planned trajectory with the unmanned device comprises:

4. The method of claim 2, wherein the planning model to be trained is a reinforcement learning model;

5. The method of claim 4, wherein obtaining the historical movement state of the interfering object specifically comprises:

6. The method of claim 5, wherein selecting a predetermined number of estimated planning trajectories from the type of estimated planning trajectories output by the planning model to be trained in the last training process comprises:

7. The method of claim 1, wherein determining a planned behavior action of the unmanned aerial vehicle when the interfering object travels according to the planned trajectory, and testing the unmanned aerial vehicle based on the behavior action comprises:

8. An unmanned equipment testing device based on interferent track planning, the device comprising:

the output module is used for inputting the historical motion state of the interferent and the current motion state of the unmanned equipment into a planning model trained in advance to obtain a planning track output by the planning model, wherein the planning track is used for generating interference on the current motion state of the unmanned equipment when the interferent runs along the planning track, and the planning model to be trained is trained by taking the maximum interference generated on the operation of the unmanned equipment as a training target when the interferent runs along the estimated planning track output by the planning model to be trained during the pre-training of the planning model;

9. A computer-readable storage medium, characterized in that the storage medium stores a computer program which, when executed by a processor, implements the method of any of the preceding claims 1-7.

10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any of claims 1-7 when executing the program.