Vehicle autonomous escape method, system, device and vehicle

Abstract

The application relates to a vehicle autonomous escape method, system, device and vehicle. The method comprises the following steps: acquiring a state parameter of the vehicle; determining whether the vehicle is in a trapped state according to the state parameter of the vehicle, and determining a trapped type of the vehicle in the case that the vehicle is in the trapped state; if the vehicle is in the trapped state, adjusting a rotation angle of a wheel or a rotation speed of the wheel according to an escape strategy corresponding to the trapped type until the state parameter of the vehicle meets a target escape state. The method can significantly improve the escape ability of the vehicle.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle control, in particular to a vehicle autonomous escape method, system, device and vehicle. BACKGROUND

[0002] The application range of current autonomous vehicles has been extended from structured urban environments to more unstructured road surfaces. Unstructured road surfaces such as sand, mud, ice and snow are prone to vehicle entrapment due to their rugged and variable terrain conditions and low friction coefficient characteristics, making it difficult for vehicles to travel normally, posing challenges to vehicle safety and efficient operation.

[0003] The prior art often uses fixed escape modes, such as adjusting the stuck wheel load or redistributing the wheel torque. This single control method cannot adapt to diverse stuck scenarios and road conditions, resulting in low vehicle escape capability. SUMMARY

[0004] Therefore, it is necessary to provide a vehicle autonomous escape method, system, device and vehicle capable of improving the vehicle escape capability to solve the above technical problems.

[0005] In a first aspect, the present application provides a vehicle autonomous escape method, comprising:

[0006] obtaining a state parameter of the vehicle;

[0007] determining whether the vehicle is in a stuck state according to the state parameter of the vehicle, and a stuck type of the vehicle in the case that the vehicle is in the stuck state;

[0008] if the vehicle is in the stuck state, adjusting the steering angle or the rotational speed of the wheel according to the stuck type corresponding escape strategy until the state parameter of the vehicle meets a target escape state.

[0009] In one embodiment, the stuck type includes single wheel stuck or coaxial wheel stuck, and adjusting the steering angle or the rotational speed of the wheel according to the stuck type corresponding escape strategy until the state parameter of the vehicle meets the target escape state, comprising:

[0010] obtaining a target escape angle, the target escape angle being a wheel steering angle at which the contact area and friction of the stuck wheel and the ground are maximum;

[0011] controlling the vehicle to move forward or backward based on the target escape angle until the state parameter of the vehicle meets the target escape state.

[0012] In one embodiment, obtaining the target escape angle comprises:

[0013] determining the target escape angle based on a reinforcement learning algorithm, the reinforcement learning algorithm comprising any one of: a deep Q-network; a deep deterministic policy gradient; a proximal policy optimization.

[0014] In one of the embodiments, the stuck type further comprises a same-side wheel stuck, and the adjusting of the steering angle or the rotation speed of the wheel is performed according to the escape strategy corresponding to the stuck type until the state parameter of the vehicle meets the target escape state, comprising:

[0015] adjusting the height of the stuck wheel to the lowest position and increasing the rotation speed of the stuck wheel;

[0016] if the stuck wheel cannot be escaped by increasing the rotation speed of the stuck wheel, then locking the stuck wheel, adjusting the rotation speed of the wheel on the other side of the vehicle until the state parameter of the vehicle meets the target escape state.

[0017] In one of the embodiments, the stuck type further comprises a same-side wheel stuck or a same-side and same-axle wheel stuck, and the adjusting of the steering angle or the rotation speed of the wheel is performed according to the escape strategy corresponding to the stuck type until the state parameter of the vehicle meets the target escape state, comprising:

[0018] determining a first stuck wheel according to the stuck degree of each stuck wheel, the stuck degree of the first stuck wheel being the lightest;

[0019] adjusting the steering angle or the rotation speed of the wheel based on the escape strategy corresponding to the single wheel stuck until the first stuck wheel is escaped;

[0020] for the remaining stuck wheels, adjusting the steering angle or the rotation speed of the wheel according to the escape strategy corresponding to the latest stuck type of the vehicle until the state parameter of the vehicle meets the target escape state.

[0021] In one of the embodiments, the state parameter of the vehicle comprises the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle and the acceleration of each wheel of the vehicle; and determining whether the vehicle is in the stuck state according to the state parameter of the vehicle, comprising:

[0022] determining the slip ratio corresponding to each wheel according to the speed of each wheel and the speed of the vehicle, respectively;

[0023] determining the acceleration difference corresponding to each wheel according to the acceleration of each wheel and the acceleration of the vehicle, respectively;

[0024] determining whether the vehicle is in the stuck state according to the slip ratio and the acceleration difference.

[0025] In one of the embodiments, determining whether the vehicle is in the stuck state according to the slip ratio and the acceleration difference, comprising:

[0026] If the slip ratio of at least one wheel is greater than a first preset threshold and the difference in acceleration between the wheels is greater than a second preset threshold, then the vehicle is determined to be in a trapped state.

[0027] Secondly, this application also provides a vehicle autonomous extrication system, which includes a data acquisition device and a controller;

[0028] The data acquisition device is used to collect the vehicle's status parameters and send them to the controller.

[0029] The controller is used to acquire the vehicle's status parameters; determine whether the vehicle is in a tangled state based on the vehicle's status parameters, and if the vehicle is in a tangled state, what type of tangled state it is; if the vehicle is in a tangled state, adjust the wheel angle or wheel speed according to the tangled state type and the corresponding escaping strategy until the vehicle's status parameters meet the target escaping state.

[0030] Thirdly, this application also provides a vehicle autonomous extrication device, comprising:

[0031] The acquisition module is used to acquire the vehicle's status parameters;

[0032] The determination module is used to determine whether the vehicle is in a trapped state based on the vehicle's status parameters, and if the vehicle is in a trapped state, the type of trapped state.

[0033] The vehicle traction module is used to adjust the wheel angle or wheel speed according to the traction strategy corresponding to the traction type when the vehicle is in a traction state, until the vehicle's state parameters meet the target traction state.

[0034] In one embodiment, the type of entrapment includes entrapment of a single wheel or entrapment of coaxial wheels. The entrapment module is specifically used to obtain the target entrapment angle, which is the wheel angle when the contact area and friction between the entrapped wheel and the ground are at their maximum. Based on the target entrapment angle, the vehicle is controlled to move forward or backward until the vehicle's state parameters meet the target entrapment state.

[0035] In one embodiment, the escape module is specifically used to determine the target escape angle based on a reinforcement learning algorithm, which includes any of the following: deep Q-network; deep deterministic policy gradient; proximal policy optimization.

[0036] In one embodiment, the entrapment type also includes entrapment of the same-side wheel. The entrapment module is specifically used to adjust the height of the entrapped wheel to the lowest position and increase the rotation speed of the entrapped wheel. If increasing the rotation speed of the entrapped wheel cannot entrap the vehicle, the entrapped wheel is locked and the rotation speed of the other wheel of the vehicle is adjusted until the vehicle's state parameters meet the target entrapment state.

[0037] In one of the embodiments, the trapped type further includes both of the wheels trapped or both of the wheels trapped on the same side and the same axis, the trapped module is specifically configured to determine a first trapped wheel according to the trapped degree of each trapped wheel, the first trapped wheel has the lightest trapped degree; adjust the rotation angle of the wheel or the rotation speed of the wheel based on the trapped strategy corresponding to the single wheel trapped, until the first trapped wheel is trapped; for the remaining trapped wheels, adjust the rotation angle of the wheel or the rotation speed of the wheel according to the trapped strategy corresponding to the latest trapped type of the vehicle, until the state parameter of the vehicle meets the target trapped state.

[0038] In one of the embodiments, the state parameter of the vehicle includes the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle and the acceleration of each wheel; the determination module is specifically configured to determine the slip ratio corresponding to each wheel according to the speed of each wheel and the speed of the vehicle respectively; determine the acceleration difference corresponding to each wheel according to the acceleration of each wheel and the acceleration of the vehicle respectively; determine whether the vehicle is in the trapped state according to the slip ratio and the acceleration difference.

[0039] In one of the embodiments, the determination module is specifically configured to determine that the vehicle is in the trapped state if the slip ratio of at least one wheel is greater than a first preset threshold value and the acceleration difference of the wheel is greater than a second preset threshold value.

[0040] In a fourth aspect, the present application also provides a vehicle, comprising a memory and a processor, the memory stores a computer program, and the processor implements the method of any one of the first aspect when executing the computer program.

[0041] In a fifth aspect, the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method of any one of the first aspect.

[0042] In a sixth aspect, the present application also provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the method of any one of the first aspect.

[0043] The vehicle autonomous trapped method, system, device and vehicle described above acquire the state parameter of the vehicle; determine whether the vehicle is in the trapped state according to the state parameter of the vehicle, and the trapped type of the vehicle in the case that the vehicle is in the trapped state; if the vehicle is in the trapped state, adjust the rotation angle of the wheel or the rotation speed of the wheel according to the trapped strategy corresponding to the trapped type, until the state parameter of the vehicle meets the target trapped state. In the above method, the rotation angle or the rotation speed of the wheel can be adjusted according to the trapped strategy corresponding to different trapped types, so as to make the vehicle trapped, and thus the trapped strategies corresponding to different trapped types are different, the trapped can be performed pertinently, the vehicle can be helped to trapped more accurately, and the trapped ability of the vehicle is improved. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the following will briefly introduce the drawings needed to be used in the description of the embodiments of the present application or the related art. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other related drawings can also be obtained based on these drawings without creative labor.

[0045] Figure 1 An application environment diagram of a vehicle autonomous escape method in an embodiment;

[0046] Figure 2 A flowchart of a vehicle autonomous escape method in an embodiment;

[0047] Figure 3 A flowchart of a step of determining whether a vehicle is in a trapped state according to a state parameter of the vehicle in an embodiment;

[0048] Figure 4 A flowchart of a step of adjusting an angle of rotation of a wheel or a rotation speed of the wheel in an embodiment;

[0049] Figure 5 A schematic diagram of a trapped type being single wheel trapped in an embodiment;

[0050] Figure 6 A schematic diagram of a trapped type being coaxial wheel trapped in an embodiment;

[0051] Figure 7 A flowchart of a step of adjusting an angle of rotation of a wheel or a rotation speed of the wheel in another embodiment;

[0052] Figure 8 A schematic diagram of a trapped type being same side wheel trapped in an embodiment;

[0053] Figure 9 A flowchart of a step of adjusting an angle of rotation of a wheel or a rotation speed of the wheel in another embodiment;

[0054] Figure 10 A flowchart of a vehicle autonomous escape method in another embodiment;

[0055] Figure 11 A structural block diagram of a vehicle autonomous escape device in an embodiment;

[0056] Figure 12 An internal structure diagram of a vehicle in an embodiment. DETAILED DESCRIPTION

[0057] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not intended to limit the present application.

[0058] It should be noted that the terms "first", "second", etc. used in the present application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "include" and "have" and any variations thereof used in the present application are intended to cover non-exclusive inclusion. The term "a plurality of" used in the present application means two or more. The term "and / or" used in the present application means one of the options or any combination of multiple options.

[0059] The vehicle autonomous escape method provided by the embodiments of the present application can be applied to the application environment as shown in Figure 1 The vehicle autonomous escape system 10 includes a collection device 100 and a controller 200. The collection device 100 is used to collect the state parameters of the vehicle and send the state parameters to the controller. The controller 200 is used to obtain the state parameters of the vehicle, determine whether the vehicle is in a trapped state according to the state parameters of the vehicle, and determine the trapped type of the vehicle in the case that the vehicle is in the trapped state. If the vehicle is in the trapped state, the rotation angle or the rotation speed of the wheel is adjusted according to the trapped type corresponding to the escape strategy, until the state parameters of the vehicle meet the target escape state.

[0060] For example, the vehicle autonomous escape system can be provided in any vehicle to help the vehicle escape when the vehicle is in a trapped state.

[0061] In an exemplary embodiment, as shown in Figure 2 A vehicle autonomous escape method is provided. The controller 200 in the Figure 1 application environment is taken as an example to illustrate the method, which includes the following steps 201 to 203. Wherein:

[0062] Step 201, obtaining the state parameters of the vehicle.

[0063] The state parameters of the vehicle can be various data and indicators used to describe the running state and performance of the vehicle.

[0064] Optionally, the state parameters of the vehicle can include the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle, and the acceleration of each wheel of the vehicle. It can also include the roll angle of the vehicle, the pitch angle of the vehicle, the roll angle acceleration of the vehicle, and the pitch angle acceleration of the vehicle. The embodiments of the present application are not limited in this regard.

[0065] Optionally, the state parameter of the vehicle can be collected by a collection device arranged on the vehicle and sent to the controller in real time.

[0066] In step 202, it is determined whether the vehicle is in a trapped state according to the state parameter of the vehicle, and the trapped type of the vehicle in the case that the vehicle is in the trapped state.

[0067] Optionally, the vehicle can not be able to normally travel or move when the vehicle is in the trapped state, and in this case, the state parameter of the vehicle will change abnormally, so whether the vehicle is in the trapped state can be determined based on the state parameter of the vehicle.

[0068] For example, the vehicle in the trapped state can be due to the wheels of the vehicle being trapped in a pit.

[0069] Optionally, the trapped type of the vehicle can include single wheel trapped, coaxial wheel trapped, same side wheel trapped, different side wheel trapped, or same side and coaxial wheel trapped.

[0070] In one possible implementation, whether the wheels are trapped can be determined according to the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle, and the acceleration of each wheel, so as to determine whether the vehicle is in the trapped state according to whether the wheels are trapped.

[0071] In another possible implementation, the state parameter of the vehicle can be input into a pre-trained neural network model, and whether the vehicle is in the trapped state can be determined according to the output of the neural network model.

[0072] Optionally, the neural network model can be obtained by training an initial neural network model with historical state parameters of various vehicles and corresponding historical trapped states as training samples.

[0073] For example, the neural network model can be a classification model, and the historical trapped state can include: not trapped, single wheel trapped, coaxial wheel trapped, same side wheel trapped, different side wheel trapped, or same side and coaxial wheel trapped, so the neural network model can determine whether the vehicle is in the trapped state and the specific trapped type according to the state parameter of the vehicle.

[0074] In step 203, if the vehicle is in the trapped state, the angle of rotation or the speed of rotation of the wheel is adjusted according to the trapped type corresponding to the escape strategy until the state parameter of the vehicle meets the target escape state.

[0075] Optionally, the escape strategies corresponding to different trapped types are different, so after the trapped type is determined, the corresponding escape strategy can be determined according to the trapped type, and the escape strategy is the adjustment mode of the angle of rotation or the speed of rotation of the wheel.

[0076] For example, the escape strategy can be adjusting the rotation angle of each wheel of the vehicle to a target rotation angle, or adjusting the rotation speed of each wheel, or can include intermittently increasing the throttle in a pulse pushing manner, controlling the vehicle to repeatedly move forward and backward, and using the potential energy accumulated by the backward movement to obtain greater forward driving power.

[0077] Optionally, the target escape state can be a state parameter of the vehicle when the vehicle can move normally.

[0078] In one possible implementation, the target escape state can include a target coordinate position, a target roll angle, a target pitch angle, a target roll angle acceleration, and a target pitch angle acceleration of the vehicle. The stuck type, the roll angle, the pitch angle, the roll angle acceleration, and the pitch angle acceleration of the vehicle can be input into a preset escape model, and the target escape state can be generated according to the escape model.

[0079] Optionally, when the rotation angle of the wheel or the rotation speed of the wheel is adjusted once, the adjusted state parameter of the vehicle can be obtained, and the adjusted state parameter is compared with the target escape state. If the adjusted state parameter is the same as or has a small difference from the target escape state, it is determined that the vehicle is escaped, and the rotation angle of the wheel or the rotation speed of the wheel does not need to be adjusted again. Otherwise, the rotation angle of the wheel or the rotation speed of the wheel is adjusted according to the escape strategy.

[0080] In another possible implementation, the target escape state can be determined according to the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle, and the acceleration of each wheel of the vehicle, so that it is determined that the wheels are not stuck. It can be understood that when the rotation angle of the wheel or the rotation speed of the wheel is adjusted once, the adjusted state parameter of the vehicle can be obtained, and then it is determined whether the wheels are stuck according to the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle, and the acceleration of each wheel of the vehicle, until it is determined that the wheels are no longer stuck.

[0081] According to the state parameter of the vehicle, it is determined whether the vehicle is in a stuck state, and the stuck type of the vehicle when the vehicle is in the stuck state. If the vehicle is in the stuck state, the rotation angle of the wheel or the rotation speed of the wheel is adjusted according to the escape strategy corresponding to the stuck type, until the state parameter of the vehicle meets the target escape state. In the above method, the rotation angle or the rotation speed of the wheel can be adjusted according to the escape strategy corresponding to different stuck types, so that the vehicle is escaped. Therefore, the escape strategies corresponding to different stuck types are different, the stuck vehicle can be escaped in a targeted manner, the vehicle can be escaped more accurately, and the escape ability of the vehicle is improved.

[0082] In one example embodiment, optionally, as Figure 3As shown, the state parameters of the vehicle include the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle, and the acceleration of each wheel; determining whether the vehicle is in a trapped state according to the state parameters of the vehicle includes the following steps 301 to 303. Wherein:

[0083] Step 301, respectively according to the speed of each wheel and the speed of the vehicle, determine the slip ratio corresponding to each wheel.

[0084] Wherein, the slip ratio can be the difference between the actual speed and the theoretical speed of the wheel during driving, which can be used to measure the degree of wheel slip.

[0085] Optionally, in the embodiment of the application, the vehicle speed is the theoretical speed of the wheel, therefore, the slip ratio corresponding to a wheel can be determined by the following formula:

[0086]

[0087] Step 302, respectively according to the acceleration of each wheel and the acceleration of the vehicle, determine the acceleration difference value corresponding to each wheel.

[0088] Optionally, for a wheel, the acceleration difference value of the wheel can be determined by subtracting the acceleration of the vehicle from the acceleration of the wheel, it can be understood that in the embodiment of the application, when determining the acceleration difference value, the absolute value of the subtraction result can be taken.

[0089] Exemplarily, taking a vehicle including a left front wheel, a right front wheel, a left rear wheel and a right rear wheel as an example, the speed and acceleration values are collected by sensors arranged on each wheel and the vehicle: the vehicle speed , the acceleration of the vehicle , the left front wheel speed , the left front wheel rim acceleration , the right front wheel speed , the right front wheel rim acceleration , the left rear wheel speed , the left rear wheel rim acceleration , the right rear wheel speed , the right rear wheel rim acceleration , four-wheel slip ratios , , , , and the difference between the four-wheel rim acceleration and the vehicle acceleration , , , .

[0090] Step 303, determining whether the vehicle is in a trapped state according to each slip ratio and each acceleration difference value.

[0091] Optionally, after determining the slip ratio and the acceleration difference of each wheel, the slip ratio can be compared with a first preset threshold and / or the acceleration difference can be compared with a second preset threshold, so as to determine the trapped state of each wheel, and determine whether the vehicle is in the trapped state according to the trapped state of the wheel.

[0092] In one possible implementation, if the slip ratio of at least one wheel is greater than the first preset threshold, or the acceleration difference of at least one wheel is greater than the second preset threshold, it is determined that the vehicle is in the trapped state.

[0093] In one possible implementation, if the slip ratio of at least one wheel is greater than the first preset threshold, and the acceleration difference of the wheel is greater than the second preset threshold, it is determined that the vehicle is in the trapped state.

[0094] It can be understood that considering the slip ratio and the acceleration difference at the same time can effectively improve the accuracy of determining the trapped state of the wheel and avoid misjudgment.

[0095] For example, when the slip ratio of only one wheel is greater than the first preset threshold and the acceleration difference of the wheel is greater than the second preset threshold, the vehicle is in the trapped state, and the trapped type of the current vehicle is single-wheel trapped; when the slip ratio of multiple wheels is greater than the first preset threshold and the acceleration difference of each wheel is greater than the second preset threshold, and these wheels are coaxial at the same time, the trapped type of the current vehicle is coaxial-wheel trapped; when the slip ratio of multiple wheels is greater than the first preset threshold and the acceleration difference of each wheel is greater than the second preset threshold, and these wheels are on the same side at the same time, the trapped type of the current vehicle is same-side-wheel trapped; when the slip ratio of multiple wheels is greater than the first preset threshold and the acceleration difference of each wheel is greater than the second preset threshold, and these wheels are on different sides at the same time, the trapped type of the current vehicle is different-side-wheel trapped; when the slip ratio of multiple wheels is greater than the first preset threshold and the acceleration difference of each wheel is greater than the second preset threshold, and these wheels include both coaxial and same-side wheels at the same time, the trapped type of the current vehicle is both same-side-wheel trapped and coaxial-wheel trapped.

[0096] In the above method, the slip ratio corresponding to each wheel is determined according to the speed of each wheel and the speed of the vehicle, the acceleration difference corresponding to each wheel is determined according to the acceleration of each wheel and the acceleration of the vehicle, and whether the vehicle is in the trapped state is determined according to the slip ratio and the acceleration difference. The vehicle trapped state can be accurately determined by using multiple sensor data, and the misjudgment rate of the vehicle trapped state can be reduced.

[0097] In one example embodiment, optionally, as Figure 4As shown, the trapped types include single wheel trapped or coaxial wheel trapped, and the turning angle of the wheel or the rotation speed of the wheel is adjusted according to the corresponding escape strategy of the trapped type until the state parameter of the vehicle meets the target escape state, including steps 401 to 402. Wherein:

[0098] Step 401, obtaining a target escape angle.

[0099] Wherein, the target escape angle is the wheel turning angle when the trapped wheel has the maximum contact area and friction with the ground.

[0100] In one possible implementation, the target escape angle can be determined based on a reinforcement learning algorithm.

[0101] Wherein, the reinforcement learning algorithm includes any one of the following: deep Q network; deep deterministic policy gradient; proximal policy optimization.

[0102] Optionally, the reinforcement learning algorithm learns how to make optimal decisions by trial and error in an environment, where the environment can be the external world in which the vehicle is located, the feedback information can be provided to the vehicle, and the action can be the behavior that the vehicle can take under a certain state parameter, for example, the action can be the turning angle of the wheel.

[0103] Wherein, the Q network is used to evaluate the long-term value of taking a certain action in a given state, and by comparing the Q values of different actions, the vehicle can decide which action is optimal in the current state, for example, which action is optimal can be to choose the turning angle of the wheel that maximizes the contact area and friction of the trapped wheel with the ground.

[0104] Optionally, the deep Q network is an algorithm that combines deep learning and reinforcement learning, which uses a deep neural network to approximate the Q network, the input is the state, and the output is the Q value of each action, which can solve the storage and calculation problems of traditional Q learning in the face of large-scale state space.

[0105] Optionally, the deep deterministic policy gradient (DDPG) is a reinforcement learning algorithm for continuous action space. It combines the ideas of deep learning and deterministic policy gradient, using two deep neural networks, one is the Actor network, which outputs the action selected in the current state; the other is the Critic network, which evaluates the goodness of the action. This algorithm has the following characteristics: using deterministic policy, that is, outputting only one action at each state; using deep neural network for parameter learning, which can handle high-dimensional state space; using experience replay technology to improve sample utilization and learning efficiency. It can perform well in continuous control tasks and learn stable strategies.

[0106] Optionally, Proximal Policy Optimization (PPO) is a policy gradient-based deep reinforcement learning algorithm that learns the optimal policy by optimizing policy parameters. It uses the proximal policy optimization method to constrain the gradient in the policy update process, improving the stability and convergence speed of the algorithm. This algorithm has the following characteristics: using the proximal policy optimization method to improve the stability and convergence speed of the algorithm; supporting continuous action space, which can handle complex control tasks; using advantage function to better evaluate the goodness of actions. It has fast convergence speed and is not sensitive to parameter settings, and can learn a stable policy.

[0107] It can be understood that it can be any reinforcement learning algorithm or its improved algorithm, for example, value-based reinforcement learning algorithm (Q-learning, etc.) or policy-based algorithm (Actor-Critic, etc.), and the type of reinforcement learning algorithm is not limited in the embodiment of the application.

[0108] In another possible implementation manner, the state parameters of the vehicle and the trapped type can be input into a pre-trained neural network model, and the target escape angle is determined according to the output of the neural network model.

[0109] Step 402, based on the target escape angle, control the vehicle to move forward or backward until the state parameters of the vehicle meet the target escape state.

[0110] Optionally, after determining the target escape angle of the trapped wheel, the wheel rotation angle of other non-trapped wheels can be determined based on the reinforcement learning algorithm or the pre-trained neural network model, and then the vehicle is controlled to move forward or backward based on the target escape angle and the wheel rotation angle of the other wheels.

[0111] It can be understood that the trapped wheel can be the leading wheel, and the wheel rotation angle of the leading wheel is mainly adjusted. The wheel rotation angle of the auxiliary wheel is fine-tuned to cooperate with the adjustment of the wheel rotation angle of the leading wheel.

[0112] Optionally, when the trapped type is single wheel trapped, the leading wheel is the trapped wheel, for example, as shown in Figure 5 , which is a schematic diagram when the trapped type is single wheel trapped, and the arrow direction is the vehicle driving direction. Taking the left front wheel as an example, the rotation angle of the left front wheel is adjusted; when the trapped type is coaxial wheel trapped, the leading wheel is the two coaxial wheels, which need to be adjusted synchronously. For example, as shown in Figure 6As shown, the arrow direction is the vehicle driving direction, and the stuck wheel types are the left front wheel and the right front wheel, and the turning angles of the left front wheel and the right front wheel are adjusted synchronously.

[0113] Optionally, the process of controlling the vehicle to move forward or backward until the state parameters of the vehicle meet the target escape state in the embodiment of the application is similar to the above-mentioned step 203, and will not be described here.

[0114] In the above method, the target escape angle is obtained, and the vehicle is controlled to move forward or backward based on the target escape angle until the state parameters of the vehicle meet the target escape state. In this way, when the stuck wheel type is single wheel stuck or coaxial wheel stuck, the wheel turning angle can be continuously adjusted and the vehicle can be controlled to move forward or backward until the target escape angle is reached, so that the contact area and friction between the wheel and the ground can be maximized, and then whether the vehicle escapes or not can be determined according to the state parameters of the vehicle at this time, so as to fundamentally improve the escape friction, avoid blind increase of the throttle to cause tire idling, and at the same time, in the case of coaxial wheel stuck, the turning of the two wheels and the driving force are synchronously controlled to balance the load of the two wheels, ensure the effective transmission of the power, and prevent the continuous idling of the wheel on one side to waste energy. The ability of the vehicle to escape can be improved. At the same time, the vehicle is controlled in the way of pulse pushing, the throttle is intermittently increased, the vehicle is repeatedly controlled to move forward and backward, and the power is applied at the appropriate time to convert potential energy into kinetic energy, so as to help the vehicle to overcome the obstacle and realize escape. This energy management-based strategy can effectively improve the success rate of escape and reduce the damage to the power system of the vehicle.

[0115] In an exemplary embodiment, as shown in Figure 7 The stuck type also includes same-side wheel stuck, and the turning angle or the rotating speed of the wheel is adjusted according to the escape strategy corresponding to the stuck type until the state parameters of the vehicle meet the target escape state, including the following steps 701 to 702. Among them:

[0116] Step 701, adjust the height of the stuck wheel to the lowest position, and increase the rotating speed of the stuck wheel.

[0117] Optionally, when the stuck type is same-side wheel stuck, the allowable wheel turning angle range of the wheel is too small at this time, and it is not suitable to continue to use angle control to help the vehicle to escape, but to increase the wheel-ground adhesion or reduce the driving force, so that the driving force of the vehicle is no longer greater than the wheel-ground adhesion.

[0118] Optionally, the height of the stuck wheel can be adjusted to the lowest position by adjusting the height of the active suspension, the adhesion between the ground and the tire is increased, and the rotating speed of the stuck wheel is increased. At this time, the state parameters of the vehicle can be obtained, and whether the vehicle escapes or not can be determined according to the state parameters of the vehicle.

[0119] If the vehicle cannot be unstuck by increasing the rotational speed of the stuck wheel, the stuck wheel is locked, and the rotational speed of the wheel on the other side of the vehicle is adjusted until the state parameter of the vehicle meets the target state of the unstuck state.

[0120] If the vehicle cannot be unstuck by increasing the rotational speed of the stuck wheel, the stuck wheel on the same side can be locked, and the rotational speed of the wheel on the other side of the vehicle is adjusted to reduce the driving force of the stuck wheel, and the torque borne by the stuck wheel is distributed to the wheel on the other side, thereby helping the vehicle to complete the unstuck.

[0121] For example, as shown in FIG. 6, when the stuck wheel is the right front wheel and the right rear wheel, the vehicle can be helped to be unstuck by lowering the height of the active suspension and the height of the stuck wheel, or by reducing the torque distribution of the stuck wheel. Figure 8

[0122] In the above method, by adjusting the height of the stuck wheel to the lowest position and increasing the rotational speed of the stuck wheel, if the vehicle cannot be unstuck by increasing the rotational speed of the stuck wheel, the stuck wheel is locked, and the rotational speed of the wheel on the other side of the vehicle is adjusted until the state parameter of the vehicle meets the target state of the unstuck state. In this way, by adjusting the rotational speed of the wheel, the driving force of the vehicle is no longer greater than the wheel-ground adhesion, thereby improving the unstuck ability of the vehicle.

[0123] In an exemplary embodiment, as shown in FIG. 7, the stuck type also includes the stuck of wheels on different sides or the stuck of wheels on the same side and the same axis, and the rotational angle or the rotational speed of the wheel is adjusted according to the unstuck strategy corresponding to the stuck type until the state parameter of the vehicle meets the target state of the unstuck state, including steps 901 to 903. Among them: Figure 9

[0124] Step 901: Determine the first stuck wheel according to the stuck degree of each stuck wheel, and the stuck degree of the first stuck wheel is the lightest.

[0125] For example, as shown in FIG. 7, when the stuck wheel is the right front wheel and the right rear wheel, the vehicle can be helped to be unstuck by lowering the height of the active suspension and the height of the stuck wheel, or by reducing the torque distribution of the stuck wheel.

[0126] Optionally, a step-by-step unstuck strategy can be used to determine the first stuck wheel with the lightest stuck degree from the multiple stuck wheels, then help the first stuck wheel to be unstuck according to the unstuck strategy corresponding to the single wheel stuck, and then help the vehicle to be unstuck according to the unstuck strategy corresponding to the current stuck type of the wheel.

[0127] ​​In a possible implementation, if the stuck wheel is determined according to the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle, and the acceleration of each wheel, the wheel with the minimum difference between the slip ratio and the first preset threshold value among the stuck wheels can be determined as the first stuck wheel, or the wheel with the minimum difference between the acceleration difference and the second preset threshold value among the stuck wheels can be determined as the first stuck wheel, or the wheel with the minimum difference between the slip ratio and the first preset threshold value and the minimum difference between the acceleration difference and the second preset threshold value among the stuck wheels can be determined as the first stuck wheel.

[0128] In another possible implementation, if the stuck state of the vehicle is determined according to the neural network model, the output of the neural network model can further include the stuck degree, and therefore, the first stuck wheel can be determined according to the output of the neural network model.

[0129] Step 902: Adjust the rotation angle or rotation speed of the wheel based on the escape strategy corresponding to the single-wheel stuck, until the first stuck wheel is unstuck.

[0130] Optionally, the escape strategy when the single wheel is stuck can be used to obtain a target escape angle, and the vehicle is controlled to move forward or backward based on the target escape angle, until the first stuck wheel is unstuck.

[0131] For example, the first stuck wheel being unstuck can mean that the slip ratio of the first stuck wheel is less than or equal to the first preset threshold value and / or the acceleration difference of the first stuck wheel is less than or equal to the second preset threshold value.

[0132] Step 903: For the remaining stuck wheels, adjust the rotation angle or rotation speed of the wheel according to the escape strategy corresponding to the latest stuck type of the vehicle, until the state parameter of the vehicle meets the target escape state.

[0133] Optionally, after the first stuck wheel is unstuck, the latest state parameter of the vehicle can be obtained, and the latest stuck type of the vehicle can be determined according to the latest state parameter.

[0134] For example, when the latest stuck type is single wheel stuck, the pulse push method can be used to intermittently increase the throttle, control the vehicle to repeatedly move forward and backward, and obtain more forward power by means of the potential energy accumulated by backward movement. At the same time, the stuck wheel is set as the leading wheel, and the wheel angle is continuously adjusted. The remaining wheels are auxiliary wheels, and the wheel angle is adjusted in coordination with the stuck wheel. The target escape angle is determined by using a deep Q network, the wheel angle is maximized to maximize the contact area between the wheel and the ground, and the friction is maximized. Finally, the target escape angle is continuously controlled to repeatedly move forward and backward until the vehicle is escaped. When the latest stuck type is coaxial wheel stuck, the escape strategy is similar to that of single wheel stuck. The vehicle is controlled to repeatedly move forward and backward by intermittently increasing the throttle. The difference is that the leading wheel is changed to the two stuck wheels on the same axis, and the wheel angles of the two stuck wheels need to be adjusted synchronously. The target escape angle needs to make the contact area between the two wheels on the same axis and the ground and the friction be larger. When the latest stuck type is same side wheel stuck, the height of the active suspension can be adjusted to the lowest position to increase the adhesion between the ground and the tire and increase the speed of the stuck wheel. Whether the vehicle can escape is detected. If the vehicle cannot escape, the stuck wheel is locked, the driving force of the stuck wheel is reduced, the torque borne by the stuck wheel is distributed to the wheel on the other side, and the vehicle is helped to escape. The embodiments of the present application will not be described here.

[0135] In the above method, the first stuck wheel is determined according to the stuck degree of each stuck wheel, the stuck degree of the first stuck wheel is the lightest, the wheel angle or the wheel speed of the first stuck wheel is adjusted based on the escape strategy corresponding to single wheel stuck, until the first stuck wheel escapes, and for the remaining stuck wheels, the wheel angle or the wheel speed is adjusted according to the escape strategy corresponding to the latest stuck type of the vehicle, until the state parameter of the vehicle meets the target escape state. In this way, the vehicle can be escaped based on the strategy of hierarchical escape, the relatively slight stuck wheel can be solved first, the trap can be deepened, and the vehicle can be escaped gradually.

[0136] Optionally, when the vehicle cannot escape according to the above method, it can be detected whether there is a rescue vehicle around, and a stuck signal can be sent to the surrounding or a manual escape instruction can be sent to the user.

[0137] For example, a preset escape time can be obtained, and when the time for making the vehicle escape according to the escape strategy is greater than the preset escape time, it can be determined that the vehicle cannot escape.

[0138] As an optional implementation, as shown in Figure 10 The vehicle autonomous escape method provided in the embodiments of the present application can include the following specific steps:

[0139] Step 1001, obtaining the state parameters of the vehicle, the state parameters of the vehicle including the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle, and the acceleration of each wheel.

[0140] Step 1002, determining the slip ratio of each wheel according to the speed of each wheel and the speed of the vehicle, respectively.

[0141] Step 1003, determining the acceleration difference of each wheel according to the acceleration of each wheel and the acceleration of the vehicle, respectively.

[0142] Step 1004, if the slip ratio of at least one wheel is greater than a first preset threshold, and the acceleration difference of the wheel is greater than a second preset threshold, determining that the vehicle is in a trapped state, and determining the trapped type of the vehicle.

[0143] Step 1005, when the trapped type is single wheel trapped or coaxial wheel trapped, determining a target escape angle based on a reinforcement learning algorithm.

[0144] The reinforcement learning algorithm includes any one of the following: deep Q network; deep deterministic policy gradient; proximal policy optimization, and the target escape angle is a wheel angle at which the contact area and friction between the trapped wheel and the ground are maximum.

[0145] Step 1006, based on the target escape angle, controlling the vehicle to move forward or backward until the state parameters of the vehicle meet a target escape state.

[0146] Step 1007, when the trapped type is same-side wheel trapped, adjusting the height of the trapped wheel to the lowest position and increasing the rotational speed of the trapped wheel.

[0147] Step 1008, if increasing the rotational speed of the trapped wheel cannot make the vehicle escape, locking the trapped wheel, adjusting the rotational speed of the other wheel of the vehicle until the state parameters of the vehicle meet the target escape state.

[0148] Step 1009, when the trapped type is opposite-side wheel trapped or both same-side and coaxial wheel trapped, determining a first trapped wheel according to the trapped degree of each trapped wheel, the trapped degree of the first trapped wheel being the lightest.

[0149] Step 1010, adjusting the rotational angle or the rotational speed of the wheel based on the escape strategy corresponding to single wheel trapped until the first trapped wheel escapes.

[0150] Step 1011, for the remaining trapped wheels, adjusting the rotational angle or the rotational speed of the wheel according to the escape strategy corresponding to the latest trapped type of the vehicle until the state parameters of the vehicle meet the target escape state.

[0151] It should be understood that although each step in the flowchart involved in each embodiment as described above is shown in sequence according to the arrow, these steps are not necessarily executed in the order indicated by the arrow. Unless otherwise specified herein, there is no strict order limitation for the execution of these steps, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but can be alternately executed with at least part of other steps or steps or stages in other steps. It can be understood that the steps in different embodiments can be freely combined as needed, and various non-contradictory schemes formed by the combination all belong to the scope of protection of the present application.

[0152] Based on the same inventive concept, the embodiments of the present application also provide a vehicle autonomous escape system for implementing the vehicle autonomous escape method described above. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme described in the above method, and therefore the specific limitations in one or more vehicle autonomous escape system embodiments provided below can refer to the limitations of the vehicle autonomous escape method described above, which will not be described here again.

[0153] In one exemplary embodiment, as shown in Figure 1 A vehicle autonomous escape system includes a collection device 100 and a controller 200.

[0154] The collection device 100 is configured to collect state parameters of the vehicle and send the state parameters to the controller.

[0155] The controller 200 is configured to obtain the state parameters of the vehicle, determine whether the vehicle is in a trapped state according to the state parameters of the vehicle, and determine the trapped type of the vehicle in the case that the vehicle is in the trapped state; and adjust the rotation angle or rotation speed of the wheels according to the trapped type corresponding to the escape strategy when the vehicle is in the trapped state, until the state parameters of the vehicle meet the target escape state.

[0156] Optionally, the collection device 100 in the vehicle autonomous escape system can include multiple sensors for collecting the roll angle, pitch angle, roll angle acceleration, pitch angle acceleration, vehicle speed, vehicle acceleration, and speed and acceleration of each wheel of the vehicle.

[0157] Optionally, the collection device 100 in the vehicle autonomous escape system can further include a camera, which is a 360-degree panoramic camera, for monitoring obstacles around the vehicle body and the vehicle bottom in real time, assisting the decision system to select a safe escape path, and avoiding collision and re-trapping after escape.

[0158] Based on the same inventive concept, the embodiments of the present application also provide a vehicle autonomous escape device for implementing the vehicle autonomous escape method described above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more vehicle autonomous escape device embodiments provided below can refer to the limitations of the vehicle autonomous escape method described above, which will not be repeated here.

[0159] In one exemplary embodiment, as shown in Figure 11 An autonomous vehicle escape device 1100 is provided, comprising: an acquisition module 1101, a determination module 1102, and an escape module 1103, wherein:

[0160] The acquisition module 1101 is configured to acquire a state parameter of the vehicle.

[0161] The determination module 1102 is configured to determine whether the vehicle is in a trapped state according to the state parameter of the vehicle, and determine a trapped type of the vehicle if the vehicle is in the trapped state.

[0162] The escape module 1103 is configured to adjust a rotation angle or a rotation speed of a wheel according to an escape strategy corresponding to the trapped type if the vehicle is in the trapped state, until the state parameter of the vehicle meets a target escape state.

[0163] In one embodiment, the trapped type includes single wheel trapping or coaxial wheel trapping, and the escape module 1103 is specifically configured to acquire a target escape angle, the target escape angle being a wheel rotation angle at which a contact area and friction of the trapped wheel with the ground are maximum; control the vehicle to move forward or backward based on the target escape angle, until the state parameter of the vehicle meets the target escape state.

[0164] In one embodiment, the escape module 1103 is specifically configured to determine the target escape angle based on a reinforcement learning algorithm, the reinforcement learning algorithm including any one of the following: deep Q network; deep deterministic policy gradient; and proximal policy optimization.

[0165] In one embodiment, the trapped type further includes same-side wheel trapping, and the escape module 1103 is specifically configured to adjust a height of the trapped wheel to a lowest position and increase a rotation speed of the trapped wheel; if the trapped wheel cannot be escaped by increasing the rotation speed of the trapped wheel, lock the trapped wheel and adjust a rotation speed of another wheel of the vehicle, until the state parameter of the vehicle meets the target escape state.

[0166] In one of the embodiments, the trapped type further includes both of the wheels trapped or both of the wheels trapped on the same side and the same axis, the escape module 1103 is specifically configured to determine a first trapped wheel according to the trapped degree of each trapped wheel, the trapped degree of the first trapped wheel is the lightest; adjust the rotation angle or the rotation speed of the wheel based on the escape strategy corresponding to the single wheel trapped until the first trapped wheel escapes; for the remaining trapped wheels, adjust the rotation angle or the rotation speed of the wheel according to the escape strategy corresponding to the latest trapped type of the vehicle until the state parameter of the vehicle meets the target escape state.

[0167] In one of the embodiments, the state parameter of the vehicle includes the speed of the vehicle, the acceleration of the vehicle, the speed of each wheel of the vehicle and the acceleration of each wheel; the determination module 1102 is specifically configured to determine the slip ratio corresponding to each wheel according to the speed of each wheel and the speed of the vehicle respectively; determine the acceleration difference corresponding to each wheel according to the acceleration of each wheel and the acceleration of the vehicle respectively; determine whether the vehicle is in the trapped state according to the slip ratio and the acceleration difference.

[0168] In one of the embodiments, the determination module 1102 is specifically configured to determine that the vehicle is in the trapped state if the slip ratio of at least one wheel is greater than a first preset threshold value and the acceleration difference of the wheel is greater than a second preset threshold value.

[0169] The above-mentioned various modules in the vehicle autonomous escape device can be realized by software, hardware and their combinations in whole or in part. The above-mentioned various modules can be embedded in or independent of the processor in the vehicle in hardware form, or can be stored in the memory in the vehicle in software form, so as to be called and executed by the processor to perform the operations corresponding to the above-mentioned various modules.

[0170] In one of the exemplary embodiments, a vehicle is provided, which can be a controller, and the internal structure diagram thereof can be as shown in Figure 12 The vehicle includes a processor, a memory, an input / output interface (Input / Output, I / O for short) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the vehicle is used to provide computing and control capability. The memory of the vehicle includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the running of the operating system and the computer program in the non-volatile storage medium. The database of the vehicle is used to store data. The input / output interface of the vehicle is used to exchange information between the processor and external devices. The communication interface of the vehicle is used to communicate with the external terminal through network connection. The computer program is executed by the processor to realize a vehicle autonomous escape method.

[0171] Those skilled in the art can understand that Figure 12 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the vehicle to which the scheme of the present application is applied. A specific vehicle can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0172] In an exemplary embodiment, a vehicle is provided, comprising a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of any of the method embodiments described above.

[0173] In an embodiment, a computer readable storage medium is provided, having stored thereon a computer program, the computer program being executed by a processor to implement the steps of any of the method embodiments described above.

[0174] In an embodiment, a computer program product is provided, comprising a computer program, the computer program being executed by a processor to implement the steps of any of the method embodiments described above.

[0175] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiment methods. In the embodiments provided in the present application, any reference to memory, database or other medium can include at least one of non-volatile memory and volatile memory. The non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. The volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., without being limited thereto.

[0176] The technical features of the above embodiments can be combined arbitrarily. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, but as long as the combinations of the technical features do not exist, they should be considered as the scope of the present application.

[0177] The above-described embodiments are merely illustrative of several embodiments of the present application, and the description is relatively specific and detailed, but should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A method for autonomous vehicle extrication, characterized in that, The method includes: Obtain the vehicle's status parameters; Determine whether the vehicle is in a trapped state based on the vehicle's status parameters, and if the vehicle is in a trapped state, determine the type of trapped state. If the vehicle is in a trapped state, the wheel angle or wheel speed is adjusted according to the escaping strategy corresponding to the trapped type until the vehicle's state parameters meet the target escaping state; the target escaping state includes the vehicle's target coordinate position, target roll angle, target pitch angle, target roll acceleration, and target pitch acceleration; The types of entrapment include entrapment of wheels on opposite sides or entrapment of wheels on the same side and coaxially. The step of adjusting the wheel angle or wheel speed according to the entrapment type and the corresponding escape strategy until the vehicle's state parameters meet the target escape state includes: The first trapped wheel is determined based on the degree of entrapment of each trapped wheel, with the first trapped wheel being the least entrapped. Adjust the wheel's turning angle or rotation speed based on the traction strategy corresponding to a single stuck wheel until the first stuck wheel is freed; For the remaining stuck wheels, adjust the wheel angle or wheel speed according to the latest entrapment type of the vehicle to meet the target entrapment state.

2. The method according to claim 1, characterized in that, The latest entrapment type is either a single wheel stuck or multiple wheels stuck on the same axle. The wheel's turning angle or rotational speed is adjusted according to the corresponding entrapment strategy until the vehicle's state parameters meet the target entrapment state, including: Obtain the target escape angle, which is the wheel angle when the contact area and friction between the trapped wheel and the ground are at their maximum; Based on the target escape angle, control the vehicle to move forward or backward until the vehicle's state parameters meet the target escape state.

3. The method according to claim 2, characterized in that, The method of obtaining the target's escape angle includes: The escape angle of the target is determined based on a reinforcement learning algorithm, which includes any one of the following: Deep Q-network; Deep deterministic policy gradient; Near-end strategy optimization.

4. The method according to claim 1, characterized in that, The latest entrapment type is entrapment of the same-side wheels. Based on the corresponding entrapment strategy, the wheel angle or wheel speed is adjusted until the vehicle's state parameters meet the target entrapment state, including: Adjust the height of the stuck wheel to the lowest position and increase the rotation speed of the stuck wheel; If increasing the rotational speed of the stuck wheel fails to free the vehicle, then the stuck wheel is locked, and the rotational speed of the other wheel of the vehicle is adjusted until the vehicle's state parameters meet the target freeing state.

5. The method according to claim 1, characterized in that, The vehicle's state parameters include the vehicle's speed, the vehicle's acceleration, the speed of each wheel of the vehicle, and the acceleration of each wheel. Determining whether the vehicle is trapped based on its status parameters includes: The slip ratio of each wheel is determined based on the speed of each wheel and the speed of the vehicle. The acceleration difference between each wheel is determined based on the acceleration of each wheel and the acceleration of the vehicle. Whether the vehicle is trapped is determined based on the slip ratio and the acceleration difference.

6. The method according to claim 5, characterized in that, The step of determining whether the vehicle is trapped based on the slip ratios and acceleration differences includes: If the slip ratio of at least one wheel is greater than a first preset threshold, and the difference in acceleration of the wheels is greater than a second preset threshold, then the vehicle is determined to be in a trapped state.

7. A vehicle autonomous extrication system, characterized in that, The system includes a data acquisition device and a controller; The acquisition device is used to acquire the vehicle's status parameters and send the status parameters to the controller; The controller is used to acquire the vehicle's state parameters; determine whether the vehicle is in a tangled state based on the vehicle's state parameters, and, if the vehicle is in a tangled state, the type of tangled state; if the vehicle is in a tangled state, adjust the wheel angle or wheel speed according to the tangled state type and the corresponding escaping strategy until the vehicle's state parameters meet the target escaping state; the target escaping state includes the vehicle's target coordinate position, target roll angle, target pitch angle, target roll acceleration, and target pitch acceleration; the tangled state type includes tangled wheels on opposite sides or tangled wheels on the same side and coaxial. If all wheels are stuck, the step of adjusting the wheel angle or wheel speed according to the traction strategy corresponding to the traction type until the vehicle's state parameters meet the target traction state includes: determining the first stuck wheel based on the degree of traction of each stuck wheel, wherein the first stuck wheel has the least degree of traction; adjusting the wheel angle or wheel speed based on the traction strategy corresponding to a single stuck wheel until the first stuck wheel is freed; for the remaining stuck wheels, adjusting the wheel angle or wheel speed according to the traction strategy corresponding to the latest traction type of the vehicle until the vehicle's state parameters meet the target traction state.

8. A vehicle autonomous extrication device, characterized in that, The device includes: The acquisition module is used to acquire the vehicle's status parameters; The determination module is used to determine whether the vehicle is in a trapped state based on the vehicle's state parameters, and if the vehicle is in a trapped state, the type of trapped state of the vehicle. The vehicle traction module is used to adjust the wheel angle or wheel speed according to the traction strategy corresponding to the traction type when the vehicle is in a traction state, until the vehicle's state parameters meet the target traction state. The target traction state includes the vehicle's target coordinate position, target roll angle, target pitch angle, target roll acceleration, and target pitch acceleration. The types of entrapment include wheels stuck on opposite sides or wheels stuck on the same side and on the same axle. The escaping module is specifically used to determine the first stuck wheel based on the degree of entrapment of each stuck wheel, with the first stuck wheel being the least entrapped; adjust the wheel's turning angle or wheel speed based on the escaping strategy corresponding to a single stuck wheel until the first stuck wheel is freed; for the remaining stuck wheels, adjust the wheel's turning angle or wheel speed according to the escaping strategy corresponding to the latest entrapment type of the vehicle until the vehicle's state parameters meet the target escaping state.

9. A vehicle comprising a memory and a processor, said memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

Images