Automatic driving vehicle control parameter determination method and device, vehicle and storage medium

Abstract

The invention discloses an automatic driving vehicle control parameter determination method and device, a vehicle and a storage medium, and relates to the technical field of automatic driving control, and the method comprises the steps: obtaining the current state of an automatic driving vehicle; obtaining an initial control parameter based on a preset control parameter model and the current state; determining a current control quantity based on the initial control parameter; and updating the initial control parameter according to the current control quantity to obtain a target control parameter. The preset control parameter model is obtained through deep learning construction training, the control parameters can be directly output through the preset control parameter model, the initial control parameters are adjusted on line through the current control quantity, and the method adapts to different control scenes. Different control parameters can be adopted for different vehicle states, so that expression of the control parameters has interpretability, and more artificial constraints can be added to enable the control parameters to be easier to adjust. And compared with adjustment of the control parameters depending on artificial experience, the difficulty of adjustment of the control parameters is reduced.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving control technology, and in particular to a method, device, vehicle and storage medium for determining control parameters of an autonomous driving vehicle. Background Art

[0002] There are many parameters that need to be adjusted in the current autonomous driving control function. There are two main ways to express the current control parameters. The first method uses an end-to-end method to implicitly express the control parameters. The input is the current vehicle information, and the output is the desired state. The error between the desired state and the reference state is used to train the implicitly expressed parameters, which is not interpretable. The second method uses an explicitly calibrated control model to explicitly express the control parameters. The adjustment of the parameters depends on manual experience, and the adjustment of the parameters is difficult. Summary of the invention

[0003] The main purpose of this application is to provide a method, device, vehicle and storage medium for determining control parameters of an autonomous driving vehicle, aiming to solve the technical problem that the adjustment of control parameters depends on manual experience, the adjustment is difficult and lacks explainability.

[0004] To achieve the above objectives, the present application proposes a method for determining control parameters of an autonomous driving vehicle, the method comprising:

[0005] Get the current state of the autonomous vehicle;

[0006] Obtaining initial control parameters based on a preset control parameter model and the current state;

[0007] Determine a current control amount based on the initial control parameter;

[0008] The initial control parameter is updated according to the current control amount to obtain a target control parameter.

[0009] In one embodiment, the step of updating the initial control parameter according to the current control amount to obtain the target control parameter includes:

[0010] Controlling the autonomous driving vehicle according to the current control amount;

[0011] Obtaining an updated state of the autonomous driving vehicle under control of the current control amount;

[0012] The initial control parameter is updated through the update state to obtain the target control parameter.

[0013] In one embodiment, the step of updating the initial control parameter by using the update state to obtain the target control parameter includes:

[0014] Calculating cost information between the updated state and the preset state;

[0015] Adjusting the parameters of the preset control parameter model according to the cost information to obtain an updated preset control parameter model;

[0016] The target control parameter is obtained through the updated preset control parameter model and the updated state.

[0017] In one embodiment, the step of calculating the cost information between the updated state and the preset state includes:

[0018] Calculating error data between the updated state and the preset state, and vehicle ride comfort data;

[0019] Respectively obtaining weight coefficients affecting the error data and the vehicle ride comfort data;

[0020] Cost information between the updated state and a preset state is calculated based on the weight coefficient, the error data, and the vehicle ride comfort data.

[0021] In one embodiment, the step of determining the current control amount based on the initial control parameter comprises:

[0022] Obtaining control requirements, and determining a type to be controlled according to the control requirements;

[0023] The current control amount is determined according to the type to be controlled and the initial control parameter.

[0024] In one embodiment, the step of determining the current control amount according to the type to be controlled and the initial control parameter includes:

[0025] Determining a preset state quantity from the current state according to the type to be controlled;

[0026] Determining a target state matrix and a target control matrix from the initial control parameters based on the type to be controlled;

[0027] A feedback gain matrix is ​​obtained by solving the target state matrix and the target control matrix;

[0028] The current control amount is calculated according to the preset state amount and the feedback gain matrix.

[0029] In one embodiment, before the step of obtaining the initial control parameters based on the preset control parameter model and the current state, the method further includes:

[0030] Obtain vehicle historical status data, preset weights, and historical control parameter data;

[0031] Based on the preset weights, a multi-layer neural network is trained using the vehicle historical state data and the historical control parameter data to construct a preset control parameter model.

[0032] In addition, to achieve the above-mentioned purpose, the present application also proposes a device for determining control parameters of an autonomous driving vehicle, the device for determining control parameters of an autonomous driving vehicle comprising:

[0033] An acquisition module is used to obtain the current state of the autonomous driving vehicle;

[0034] The acquisition module is used to obtain initial control parameters based on a preset control parameter model and the current state;

[0035] A determination module, configured to determine a current control amount based on the initial control parameter;

[0036] An updating module is used to update the initial control parameter according to the current control amount to obtain a target control parameter.

[0037] In addition, to achieve the above-mentioned purpose, the present application also proposes a vehicle, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the computer program is configured to implement the steps of the method for determining the control parameters of an autonomous driving vehicle as described above.

[0038] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the method for determining the control parameters of the autonomous driving vehicle as described above are implemented.

[0039] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the method for determining the control parameters of the autonomous driving vehicle as described above.

[0040] One or more technical solutions proposed in this application obtain the current state of the autonomous driving vehicle; obtain the initial control parameters based on the preset control parameter model and the current state; determine the current control amount based on the initial control parameters; and update the initial control parameters according to the current control amount to obtain the target control parameters. The preset control parameter model is constructed and trained through deep learning, and the control parameters can be directly output through the preset control parameter model, and the initial control parameters can be adjusted online through the current control amount, so as to continuously update the control parameters through the model to obtain the optimal control parameters to adapt to different control scenarios. Different vehicle states can use different control parameters, which are applicable to a wider control range and more refined control. At the same time, the expression of the control parameters is interpretable, and more artificial constraints can be added to make the control parameters easier to adjust. Compared with relying on manual experience to adjust the control parameters, the difficulty of adjusting the control parameters is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0042] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0043] Figure 1 A flowchart of the first embodiment of the method for determining control parameters of an autonomous driving vehicle of the present application;

[0044] Figure 2 A flowchart diagram of the second embodiment of the method for determining control parameters of an autonomous driving vehicle of the present application;

[0045] Figure 3 A flowchart diagram of the third embodiment of the method for determining control parameters of an autonomous driving vehicle of the present application;

[0046] Figure 4 A flowchart diagram of a fourth embodiment of the method for determining control parameters of an autonomous driving vehicle of the present application;

[0047] Figure 5 This is a schematic diagram of a simplified flow chart of an embodiment of a method for determining control parameters of an autonomous driving vehicle;

[0048] Figure 6 This is a schematic diagram of the module structure of the device for determining control parameters of an autonomous driving vehicle according to an embodiment of the present application;

[0049] Figure 7A schematic diagram of the vehicle structure of the hardware operating environment involved in the method for determining control parameters of an autonomous driving vehicle in an embodiment of the present application.

[0050] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0051] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0052] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0053] The main solution of the embodiment of the present application is: obtain the current state of the autonomous driving vehicle; obtain initial control parameters based on a preset control parameter model and the current state; determine the current control quantity based on the initial control parameters; update the initial control parameters according to the current control quantity to obtain the target control parameters.

[0054] Since there are many control parameters that need to be adjusted in the control process of autonomous driving vehicles in the existing technology, there are currently two main ways to express control parameters. The first method uses an end-to-end method to implicitly express control parameters. The input is the current vehicle information, and the output is the desired state. The error between the desired state and the reference state is used to train the implicitly expressed parameters, which is not interpretable. The second method uses an explicitly calibrated control model to explicitly express control parameters. The adjustment of parameters depends on manual experience, and the adjustment of parameters is difficult.

[0055] This application provides a solution that explicitly expresses the control parameters through a control parameter model, and uses the vehicle's current state, expected physical state, etc. to guide training, thereby obtaining the target control parameters. Its advantages are that different control parameters can be used for different vehicle states, with a wider range of applicable control and more refined control; the expression of control parameters is interpretable, and more artificial constraints can be added to make the control parameters easier to adjust.

[0056] It should be noted that the execution subject of this embodiment may be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a controller inside a vehicle, etc. The following takes the controller inside a vehicle as an example to illustrate this embodiment and the following embodiments.

[0057] Based on this, the present application embodiment provides a method for determining control parameters of an autonomous driving vehicle, referring to Figure 1 , Figure 1This is a flow chart of the first embodiment of the method for determining control parameters of an autonomous driving vehicle of the present application.

[0058] In this embodiment, the method for determining the control parameters of the autonomous driving vehicle includes steps S10 to S40:

[0059] Step S10: Obtain the current state of the autonomous driving vehicle.

[0060] It should be noted that the main application scenario of this embodiment is the determination of control parameters in the process of controlling an autonomous vehicle. Different vehicle states can use different control parameters. Therefore, when the state of the autonomous vehicle is different, the determined control parameters are also different.

[0061] It can be understood that the current state of the autonomous driving vehicle can be detected by sensors arranged at various positions on the autonomous driving vehicle. The current state may include the vehicle's lateral position, longitudinal position, vehicle angle (such as vehicle heading angle, yaw angle, attitude angle, etc.), vehicle offset value and other data of the vehicle during driving, and may also include other state data, such as vehicle speed, acceleration and other data, which is not limited in this embodiment.

[0062] Step S20: obtaining initial control parameters based on a preset control parameter model and the current state.

[0063] It should be noted that the preset control parameter model is a deep learning control parameter model that is trained and constructed in advance. The current state of the autonomous driving vehicle can be input into the preset control parameter model, and the corresponding control parameters can be directly output online through the preset control parameter model.

[0064] The preset control parameter model represents the correspondence between the state of the autonomous driving vehicle and the control parameters. Therefore, the initial control parameters can be obtained according to the preset control parameter model and the current state. The initial control parameters may not be the optimal parameters for vehicle control. The initial control parameters can be adjusted and updated to obtain the optimal control parameters. The initial control parameters may include lateral control parameters, longitudinal control parameters, vehicle state parameters, etc. For example, the longitudinal control parameters may include control parameters such as vehicle speed and acceleration, and the lateral control parameters may include control parameters such as steering, and may also include other control parameters, which can be determined according to specific control requirements and control scenarios.

[0065] In a feasible implementation manner, step S20 may include steps S11 to S12:

[0066] Step S11: Obtain vehicle historical status data, preset weights and historical control parameter data.

[0067] It should be noted that the vehicle historical status data is the driving data of different vehicles in different driving environments collected in advance. The preset weight can be set according to the needs. The preset weight is an initial value. During training, the weight will be continuously adjusted according to the training process.

[0068] The historical control parameter data is the control data of different vehicles in different driving processes collected in advance.

[0069] Step S12: Based on the preset weights, a multi-layer neural network is trained using the vehicle historical state data and the historical control parameter data to construct a preset control parameter model.

[0070] In a specific implementation, a multi-layer neural network may be trained according to preset weights, historical vehicle status data, and historical control parameter data to construct a preset control parameter model.

[0071] In specific implementation, the multi-layer neural network MLP can be constructed by selecting the appropriate number of layers and the number of neurons in each layer. The selection of the number of layers depends on the complexity of the task, while the number of neurons in each layer affects the learning ability and generalization ability of the model.

[0072] By taking the historical vehicle status data X as input, an explicit expression of the control parameters is constructed. For example, the model can be expressed as: Q = h1(X), R = h2(X), where h1 and h2 are explicit expressions of Q and R, and the h model can be expressed using a multi-layer neural network with weights W. M = [Q, R] can be expressed as: M = MLP(X) according to the neural network MLP, which can adjust the parameters of Q and R online according to the input state X to adapt to different control scenarios.

[0073] The state matrix Q is used to measure the "cost" or "importance" of state variables. The elements of the Q matrix include:

[0074] matrix_q_(0,0)=2: control parameter of lateral deviation.

[0075] matrix_q_(1,1)=1: control parameter of the lateral deviation rate.

[0076] matrix_q_(2,2)=0.1: control parameter of heading angle deviation.

[0077] matrix_q_(3,3)=0.1: Control parameter of heading angle deviation rate.

[0078] The control matrix R is used to measure the "cost" of the control input. The elements of the R matrix are:

[0079] matrix_r_(0,0)=10: control parameter of the control quantity (such as the front wheel steering angle). The type of matrix in the model can also be adjusted according to the actual control requirements.

[0080] Through the above historical data, deep learning training is performed until the model converges to obtain a preset control parameter model. In a specific implementation, the weights corresponding to the control parameters can be trained through the deep learning model, and the weights corresponding to the control parameters at the end of the training can be reversely solved to obtain the optimal control parameters.

[0081] Step S30: Determine the current control amount based on the initial control parameter.

[0082] In a specific implementation, the current control quantity is the optimal control input quantity during the automatic driving process. In different control processes, the corresponding current control quantity is different.

[0083] For example, the control process is to perform lateral control on the autonomous driving vehicle, and the current control amount may be the desired front wheel steering angle, or other control amounts, which are not limited in this implementation.

[0084] Step S40: updating the initial control parameter according to the current control amount to obtain a target control parameter.

[0085] It should be noted that the autonomous driving vehicle can be controlled according to the current control quantity, so that the initial control parameters can be continuously adjusted to obtain the optimal control parameters, namely the target control parameters.

[0086] In a specific implementation, after obtaining the target control parameters, the autonomous driving vehicle can be controlled according to the target control parameters, and the actual driving conditions of the vehicle can be continuously acquired, and the status of the vehicle can be continuously updated according to the driving conditions, so as to adjust the control parameters in real time according to the updated status.

[0087] The target control parameters may include optimal speed control parameters, acceleration control parameters, steering control parameters and other parameters, and may also include other control parameters, which are not limited in this embodiment.

[0088] This embodiment provides a method for determining control parameters of an autonomous driving vehicle, which obtains the current state of the autonomous driving vehicle; obtains initial control parameters based on a preset control parameter model and the current state; determines the current control amount based on the initial control parameters; and updates the initial control parameters according to the current control amount to obtain target control parameters. The preset control parameter model is constructed and trained through deep learning, and the control parameters can be directly output through the preset control parameter model, and the initial control parameters can be adjusted online through the current control amount, so that the control parameters are continuously updated through the model to obtain the optimal control parameters to adapt to different control scenarios. Different vehicle states can use different control parameters, which are applicable to a wider control range and more refined control. At the same time, the expression of control parameters is interpretable, and more artificial constraints can be added to make the control parameters easier to adjust. Compared with relying on manual experience to adjust the control parameters, the difficulty of adjusting the control parameters is reduced.

[0089] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 2 , step S40 includes steps S401 to S403:

[0090] Step S401: Control the automatic driving vehicle according to the current control amount.

[0091] It should be noted that after obtaining the current control amount, the autonomous driving vehicle can be controlled by the current control amount, thereby adjusting the state of the autonomous driving vehicle.

[0092] Step S402: Obtain an updated status of the automatic driving vehicle under the control of the current control amount.

[0093] During the control process of the autonomous driving vehicle, the updated status of the vehicle under the current control quantity can be obtained, and the updated status can be used as the final control effect.

[0094] Step S403: updating the initial control parameters through the update state to obtain target control parameters.

[0095] In a specific implementation, the initial control parameters may be updated based on the update state, for example, by determining the difference data between the update state and the preset reference state, thereby adjusting the initial control parameters to obtain target control parameters, which are optimized control parameters.

[0096] This embodiment controls the autonomous driving vehicle according to the current control amount; obtains the updated state of the autonomous driving vehicle under the control of the current control amount; updates the initial control parameters through the updated state to obtain the target control parameters. By continuously obtaining the updated state of the vehicle under the current control amount and adjusting the control parameters accordingly, the control error can be gradually reduced, so that the vehicle's driving trajectory is closer to the expected path, and the control accuracy of the autonomous driving system is improved.

[0097] Based on the second embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those in the above-mentioned second embodiment can refer to the above introduction, and will not be repeated in the following. Figure 3 , step S403 includes steps A11 to A13:

[0098] Step A11: Calculate the cost information between the updated state and the preset state.

[0099] It should be noted that the cost information is the set cost function cost. By replacing the loss function with true value supervision with cost information, the problem that the intermediate control variables cannot be trained without ground truth can be solved, and the trainability and interpretability of the model can be enhanced.

[0100] The preset state is the optimal state of the vehicle or the target state that the vehicle needs to achieve as set by the user.

[0101] In a specific implementation, the cost information can be determined based on the difference data between the corrected state and the preset state. Therefore, step A11 includes: calculating the error data and the vehicle smoothness data between the updated state and the preset state; respectively obtaining the weight coefficients affecting the error data and the vehicle smoothness data; and calculating the cost information between the updated state and the preset state based on the weight coefficients, the error data and the vehicle smoothness data.

[0102] It should be noted that the cost information to be solved includes but is not limited to the lateral error, longitudinal error, angular error, vehicle smoothness and other data between the corrected state and the preset state of the vehicle. Therefore, the error data between the data in the corrected state and the data in the preset state and the vehicle smoothness data can be used. If the current control of the autonomous driving vehicle is lateral control, the error data may include lateral error and angular error. It is also possible to assign weight coefficients corresponding to each error data and smoothness data, so as to continuously solve and update the cost. For example, the weight coefficients are a1, a2, and a3. When the weight of a1 is larger, the lateral error has a greater impact. When the weight of a2 is larger, the angular error has a greater impact. When the error of a3 is larger, the vehicle smoothness has a greater impact.

[0103] The process of calculating the cost information is as follows:

[0104] cost=(a1*error_leteral 2 +a2*error_heading 2 +a3*jerk 2 )(Formula 1)

[0105] In the above formula 1, error_leteral is the lateral error, error_heading is the angular error, and jerk is the ride comfort data.

[0106] Step A12: adjusting the parameters of the preset control parameter model according to the cost information to obtain an updated preset control parameter model.

[0107] In a specific implementation, the parameters of the preset control parameter model may be adjusted based on the cost information, such as adjusting the weight of the preset control parameter model, so as to continuously iteratively update the preset control parameter model to obtain an updated preset control parameter model.

[0108] In a specific implementation, the minimum cost value can be solved according to the above formula 1 to obtain the optimal state of the corresponding vehicle, and the parameters of the multi-layer neural network model can be adjusted according to the cost cost so that the cost cost reaches the minimum value, thereby optimizing the control parameter M.

[0109] Step A13: Obtaining target control parameters through the updated preset control parameter model and the updated state.

[0110] In a specific implementation, the latest state of the autonomous driving vehicle can be input into an updated preset control parameter model, so that the control parameters can be output online through the updated preset control parameter model to obtain the target control parameters, that is, the optimal control parameters.

[0111] This embodiment calculates the cost information between the updated state and the preset state; adjusts the parameters of the preset control parameter model according to the cost information to obtain an updated preset control parameter model; obtains the target control parameter through the updated preset control parameter model and the updated state. By designing the cost information, the training of the preset control parameter model is optimized, replacing the loss function with true value supervision, improving the training efficiency and interpretability of the model, thereby obtaining the optimal control parameter model, and then obtaining the optimal control parameters.

[0112] Based on the first embodiment of the present application, in the fourth embodiment of the present application, the same or similar contents as those in the first embodiment can be referred to the above description, and will not be described in detail later. Figure 4 , step S30 includes steps S301 to S302:

[0113] Step S301: Obtain control requirements, and determine the type to be controlled according to the control requirements.

[0114] It should be noted that the control requirements may be control requirements for the autonomous driving vehicle, such as control requirements for maintaining center driving, changing lanes to the left, changing lanes to the right, turning, etc.

[0115] In a specific implementation, the type to be controlled can be determined according to specific control requirements, and the type to be controlled may include lateral control, longitudinal control, etc. of the vehicle. For example, if the control requirement is to control the vehicle to maintain centering, the corresponding type to be controlled is lateral control.

[0116] Step S302: determining the current control amount according to the type to be controlled and the initial control parameter.

[0117] In a specific implementation, the current control amount can be determined according to the type to be controlled and the initial control parameters, for example, the corresponding control model can be determined according to the type to be controlled, so as to input the initial control parameters into the control model to obtain the current control amount. For example, if the type to be controlled is lateral control, the control model can be a lateral control model LQR.

[0118] In a feasible implementation, step S302 may include steps B11 to B14:

[0119] Step B11: determining a preset state quantity from the current state according to the type to be controlled;

[0120] In a specific implementation, the current state includes states of the vehicle under different control conditions. When the type to be controlled is lateral control, a state related to lateral control can be selected from the current state, including lateral position, vehicle angle, etc., to obtain a preset state quantity. The preset state quantity is the state corresponding to the type to be controlled. For example, the preset state quantity includes:

[0121] Lateral deviation e1: The lateral distance between the actual path of the vehicle and the expected path.

[0122] Lateral deviation rate e1': the rate of change of lateral deviation.

[0123] Heading angle deviation e2: The angle between the actual heading of the vehicle and the expected heading.

[0124] Heading angle deviation rate e2': the rate of change of heading angle deviation.

[0125] Step B12: determining a target state matrix and a target control matrix from the initial control parameters based on the type to be controlled;

[0126] In a specific implementation, if the type to be controlled is lateral control, the target state matrix determined from the initial control parameters is the state matrix Q of the lateral control, and the target control matrix is ​​the control matrix R of the lateral control.

[0127] Step B13: Obtain a feedback gain matrix by solving the target state matrix and the target control matrix.

[0128] In a specific implementation, the feedback gain matrix can be obtained by solving the Riccati (Algebra Riccati Equation, ARE) equation through the target state matrix and the target control matrix.

[0129] The form of ARE is as follows:

[0130] A T P+PA-PBR -1 B T P+Q=0(Formula 2)

[0131] Among them, P is a positive definite matrix, which is the solution of ARE. A and B are the system matrix and input matrix respectively, which can be collected in advance. After calculating P, the feedback gain matrix K can be calculated.

[0132] After obtaining the state matrix Q and the control matrix R, ARE can be solved numerically or analytically to obtain the matrix P, and the feedback gain matrix K can be calculated using P. The calculation process is as follows:

[0133] K=R -1 B T P(Formula 3)

[0134] After the feedback gain matrix is ​​calculated by the above formula 3, the calculated K can be applied to the process of calculating the current control quantity.

[0135] In practical applications, ARE can also be solved using existing software tools, such as the care function in MATLAB or the scipy.1inalg.solve_continuous_are function in Python. These tools can efficiently solve ARE and calculate the feedback gain matrix K.

[0136] Step B14: Calculate the current control quantity according to the preset state quantity and the feedback gain matrix.

[0137] In a specific implementation, the current control amount u* can be calculated by the preset state amount and the feedback gain matrix, and the current control amount u*=-Kx, where x is the preset state amount and K is the feedback gain matrix. For example, if the control type is lateral control, the current control amount is the desired front wheel steering angle.

[0138] This embodiment obtains control requirements, determines the type to be controlled according to the control requirements, and determines the current control amount according to the type to be controlled and the initial control parameter. It can quickly identify control requirements and determine the type to be controlled, so as to quickly respond to external changes or internal instructions and improve the response speed of the system. According to different control requirements and types, the system can flexibly adjust the control strategy to enhance the adaptability to different working conditions.

[0139] For example, in order to help understand the implementation process of the method for determining the control parameters of an autonomous driving vehicle obtained by combining this embodiment with the above-mentioned embodiment 1, please refer to Figure 5 , Figure 5 A simplified flow chart of a method for determining control parameters of an autonomous driving vehicle is provided. Specifically, the vehicle state X (such as lateral position and vehicle angle) is input, and the initial control parameters (M=MLP(X)) are output through a multi-layer MLP (preset control parameter model), and the lateral control model (taking LQR as an example) is input to obtain the new vehicle state calculation jerk, lateral error, and angle error, and then the cost is reversely solved so that the control parameter M=[Q, R] reaches the optimal value, thereby obtaining the optimal control parameter. The preset control parameter model can be specifically constructed by taking the current state of the vehicle as input to construct an explicit expression of the control parameter. The model can be expressed as follows: Q=h1(X), R=h2(X), where h1 and h2 are explicit expressions of Q and R, and the h model can be expressed by a multi-layer neural network with weights W. M=[Q,R], then according to the neural network MLP, it can be expressed as: M=MLP(X), guide the model training through the current state of the vehicle and the expected physical state, build a preset control parameter model, establish the cost of the control parameters and the final control effect, and solve the minimum cost, so that the control parameter M=[Q,R] reaches the optimal value. Specifically, the control parameters are input into the corresponding control model. For example, the lateral control model takes the LRQ model as an example, and the current control amount u can be obtained. The control amount actuates the vehicle to obtain a new vehicle state as the final control effect. The final solution cost is set to include but not limited to the following factors: the lateral error between the vehicle and the reference state, the angular error between the vehicle and the reference state, the vehicle smoothness (jerk), etc., and the optimal control parameters are obtained by reverse solution.

[0140] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the method for determining the control parameters of the autonomous driving vehicle of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0141] This application also provides a device for determining control parameters of an autonomous driving vehicle, please refer to Figure 6 , the automatic driving vehicle control parameter determination device comprises:

[0142] The acquisition module 10 is used to obtain the current state of the autonomous driving vehicle.

[0143] The acquisition module 10 is used to obtain initial control parameters based on a preset control parameter model and the current state.

[0144] The determination module 20 is used to determine the current control amount based on the initial control parameter.

[0145] The updating module 30 is used to update the initial control parameter according to the current control amount to obtain the target control parameter.

[0146] The autonomous driving vehicle control parameter determination device provided by the present application adopts the autonomous driving vehicle control parameter determination method in the above-mentioned embodiment, which can solve the technical problems that the adjustment of control parameters depends on manual experience, the adjustment of parameters is difficult and lacks interpretability. Compared with the prior art, the beneficial effects of the autonomous driving vehicle control parameter determination device provided by the present application are the same as the beneficial effects of the autonomous driving vehicle control parameter determination method provided by the above-mentioned embodiment, and the other technical features in the autonomous driving vehicle control parameter determination device are the same as the features disclosed in the above-mentioned embodiment method, which will not be repeated here.

[0147] In one embodiment, the update module 30 is also used to control the autonomous driving vehicle according to the current control amount; obtain the updated state of the autonomous driving vehicle under the control of the current control amount; and update the initial control parameters through the updated state to obtain target control parameters.

[0148] In one embodiment, the update module 30 is also used to calculate the cost information between the updated state and the preset state; adjust the parameters of the preset control parameter model according to the cost information to obtain an updated preset control parameter model; and obtain the target control parameter through the updated preset control parameter model and the updated state.

[0149] In one embodiment, the update module 30 is also used to calculate the error data and vehicle smoothness data between the updated state and the preset state; respectively obtain the weight coefficients affecting the error data and the vehicle smoothness data; and calculate the cost information between the updated state and the preset state based on the weight coefficients, the error data and the vehicle smoothness data.

[0150] In one embodiment, the determination module 20 is further used to obtain a control requirement, determine the type to be controlled according to the control requirement, and determine the current control amount according to the type to be controlled and the initial control parameter.

[0151] In one embodiment, the determination module 20 is also used to determine a preset state quantity from the current state according to the type to be controlled; determine a target state matrix and a target control matrix from the initial control parameters based on the type to be controlled; obtain a feedback gain matrix by solving the target state matrix and the target control matrix; and calculate a current control quantity according to the preset state quantity and the feedback gain matrix.

[0152] In one embodiment, the acquisition module 10 is also used to acquire vehicle historical status data, preset weights and historical control parameter data; based on the preset weights, a multi-layer neural network is trained through the vehicle historical status data and the historical control parameter data to construct a preset control parameter model.

[0153] The present application provides a vehicle, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for determining control parameters of an autonomous driving vehicle in the above-mentioned embodiment one.

[0154] Reference below Figure 7 , which shows a schematic structural diagram of a vehicle suitable for implementing an embodiment of the present application. Figure 7 The vehicle shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0155] like Figure 7As shown, the vehicle may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for vehicle operation are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the vehicle to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a vehicle with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0156] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0157] The vehicle provided by this application adopts the method for determining the control parameters of the autonomous driving vehicle in the above embodiment, which can solve the technical problem that the adjustment of the control parameters depends on manual experience, the adjustment is difficult and lacks interpretability. Compared with the prior art, the beneficial effects of the vehicle provided by this application are the same as the beneficial effects of the method for determining the control parameters of the autonomous driving vehicle provided by the above embodiment, and the other technical features in the vehicle are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0158] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0159] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0160] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the method for determining the control parameters of the autonomous driving vehicle in the above-mentioned embodiment.

[0161] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0162] The computer-readable storage medium may be included in the vehicle, or may exist independently without being installed in the vehicle.

[0163] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the vehicle, the vehicle: obtains the current state of the autonomous driving vehicle; obtains initial control parameters based on a preset control parameter model and the current state; determines the current control amount based on the initial control parameters; updates the initial control parameters according to the current control amount to obtain the target control parameters.

[0164] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0165] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0166] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0167] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned method for determining control parameters of an autonomous driving vehicle, and can solve the technical problem that the adjustment of control parameters depends on manual experience, the adjustment of parameters is difficult, and there is no interpretability. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the method for determining control parameters of an autonomous driving vehicle provided in the above-mentioned embodiment, and will not be repeated here.

[0168] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the method for determining control parameters of an autonomous driving vehicle as described above.

[0169] The computer program product provided by this application can solve the technical problem that the adjustment of control parameters depends on manual experience, the adjustment is difficult and lacks interpretability. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the method for determining the control parameters of the autonomous driving vehicle provided by the above embodiment, and will not be elaborated here.

[0170] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for determining control parameters of an autonomous driving vehicle, characterized in that: The method for determining control parameters of an autonomous driving vehicle comprises: Get the current state of the autonomous vehicle; Obtaining initial control parameters based on a preset control parameter model and the current state; Determine a current control amount based on the initial control parameter; The initial control parameter is updated according to the current control amount to obtain a target control parameter.

2. The method according to claim 1, characterized in that The step of updating the initial control parameter according to the current control amount to obtain the target control parameter comprises: Controlling the autonomous driving vehicle according to the current control amount; Obtaining an updated state of the autonomous driving vehicle under control of the current control amount; The initial control parameter is updated through the update state to obtain the target control parameter.

3. The method according to claim 2, characterized in that The step of updating the initial control parameter by using the update state to obtain the target control parameter comprises: Calculating cost information between the updated state and the preset state; Adjusting the parameters of the preset control parameter model according to the cost information to obtain an updated preset control parameter model; The target control parameter is obtained through the updated preset control parameter model and the updated state.

4. The method according to claim 3, characterized in that The step of calculating the cost information between the updated state and the preset state comprises: Calculating error data between the updated state and the preset state, and vehicle ride comfort data; Respectively obtaining weight coefficients affecting the error data and the vehicle ride comfort data; Cost information between the updated state and a preset state is calculated based on the weight coefficient, the error data, and the vehicle ride comfort data.

5. The method according to claim 1, characterized in that The step of determining the current control amount based on the initial control parameter comprises: Obtaining control requirements, and determining a type to be controlled according to the control requirements; The current control amount is determined according to the type to be controlled and the initial control parameter.

6. The method according to claim 5, characterized in that The step of determining the current control amount according to the type to be controlled and the initial control parameter comprises: Determining a preset state quantity from the current state according to the type to be controlled; Determining a target state matrix and a target control matrix from the initial control parameters based on the type to be controlled; A feedback gain matrix is ​​obtained by solving the target state matrix and the target control matrix; The current control amount is calculated according to the preset state amount and the feedback gain matrix.

7. The method according to any one of claims 1 to 6, characterized in that Before the step of obtaining the initial control parameters based on the preset control parameter model and the current state, the method further includes: Obtain vehicle historical status data, preset weights, and historical control parameter data; Based on the preset weights, a multi-layer neural network is trained using the vehicle historical state data and the historical control parameter data to construct a preset control parameter model.

8. A device for determining control parameters of an autonomous driving vehicle, characterized in that: The device comprises: An acquisition module is used to obtain the current state of the autonomous driving vehicle; The acquisition module is used to obtain initial control parameters based on a preset control parameter model and the current state; A determination module, configured to determine a current control amount based on the initial control parameter; An updating module is used to update the initial control parameter according to the current control amount to obtain a target control parameter.

9. A vehicle, characterized in that: The vehicle includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the method for determining control parameters of an autonomous driving vehicle as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for determining control parameters of an autonomous driving vehicle as described in any one of claims 1 to 7 are implemented.

Images