Vehicle stability control method and device, storage medium, controller and vehicle

Abstract

The invention discloses a vehicle stability control method and device, a storage medium, a controller and a vehicle, and relates to the technical field of vehicles. The vehicle stability control method comprises the steps that a working condition recognition signal and a vehicle body deviation signal of a vehicle are obtained; and determining a target control quantity based on the deep learning model, the working condition identification signal and the vehicle body deviation signal, and controlling the vehicle based on the target control quantity. Therefore, by using the strong learning ability and generalization ability of the deep learning model, the accurate target control quantity can be obtained according to the working condition identification signal and the vehicle body deviation signal, so that the accurate control and optimization improvement of the vehicle control stability can be realized, and the control stability of the vehicle under various complex working conditions is improved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicles, and in particular, to a control method, device, storage medium, controller, and vehicle for vehicle stability. Background Art

[0002] In modern vehicle engineering, the handling stability of a vehicle is one of the key factors ensuring driving safety and experience. With the continuous development of vehicle technology and the increasing complexity of driving conditions, traditional stability control systems gradually show their limitations when dealing with vehicle state deviations caused by various uncertain factors. For example, changes in road surface conditions, uneven distribution of vehicle loads, wear of components, and measurement errors of sensors may all cause the actual driving state of the vehicle to deviate from the ideal state, thus affecting the handling stability of the vehicle. Existing stability control methods often rely on fixed models and preset parameters, and it is difficult to adapt to these dynamically changing deviation situations in real time, accurately, and comprehensively, resulting in unsatisfactory control effects and unable to fully meet the strict requirements for handling stability under various complex conditions of the vehicle. Summary of the Invention

[0003] Therefore, the purpose of the present application is to propose a control method, device, storage medium, controller, and vehicle for vehicle stability to improve the handling stability of the vehicle under different conditions.

[0004] In a first aspect, an embodiment of the present application proposes a control method for vehicle stability, the method including: obtaining a working condition identification signal and a body deviation signal of the vehicle; determining a target control quantity based on a deep learning model, the working condition identification signal, and the body deviation signal, and controlling the vehicle based on the target control quantity.

[0005] In a second aspect, an embodiment of the present application proposes a control device for vehicle stability, the device including: an identification module for obtaining a working condition identification signal and a body deviation signal of the vehicle; a control module for determining a target control quantity based on a deep learning model, the working condition identification signal, and the body deviation signal; and an execution module for controlling the vehicle according to the target control quantity.

[0006] In a third aspect, an embodiment of the present application proposes a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method described in the first aspect above is implemented.

[0007] In a fourth aspect, an embodiment of the present application proposes a controller, including a memory, a processor, and a computer program stored on the memory, and when the computer program is executed by the processor, the method described in the first aspect above is implemented.

[0008] Fifth aspect, an embodiment of the present application provides a vehicle, including: the control device for vehicle stability described in the second aspect above, and / or, the controller described in the fourth aspect above.

[0009] For the vehicle stability control method, device, storage medium, controller, and vehicle according to the embodiments of the present application, first, a vehicle condition identification signal and a body deviation signal of the vehicle are obtained; then, by utilizing the powerful learning ability and generalization ability of the deep learning model, an accurate target control quantity can be obtained according to the vehicle condition identification signal and the body deviation signal; finally, the vehicle can be effectively controlled for stability according to the target control quantity, so as to achieve precise control and optimization improvement of the vehicle handling stability, and improve the vehicle handling stability under various complex working conditions.

[0010] The additional aspects and advantages of the present application will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present application. Description of the Drawings

[0011] Figure 1 is a flowchart of the vehicle stability control method according to the embodiment of the present application;

[0012] Figure 2 is a schematic diagram of an identification module according to an embodiment of the present application;

[0013] Figure 3 is a schematic diagram of a control module according to an embodiment of the present application;

[0014] Figure 4 is a schematic diagram of a supervised learning module according to an embodiment of the present application;

[0015] Figure 5 is a schematic diagram of a reinforcement self-learning module according to an embodiment of the present application;

[0016] Figure 6 is a schematic structural diagram of a deep learning model according to an embodiment of the present application;

[0017] Figure 7 is a schematic diagram of an execution module according to an embodiment of the present application;

[0018] Figure 8 is a schematic diagram of the execution process of the vehicle stability control method according to an embodiment of the present application;

[0019] Figure 9 is a structural block diagram of the vehicle stability control device according to the embodiment of the present application;

[0020] Figure 10 is a structural block diagram of the controller according to the embodiment of the present application. Detailed Embodiments

[0021] Embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present application, and should not be construed as a limitation of the present application.

[0022] The following describes a method, device, storage medium, controller, and vehicle for controlling vehicle stability according to embodiments of the present application with reference to the accompanying drawings.

[0023] Figure 1 It is a flowchart of the method for controlling vehicle stability according to an embodiment of the present application.

[0024] As Figure 1 shown, the method for controlling vehicle stability includes:

[0025] S11, obtaining a vehicle condition identification signal and a body deviation signal of the vehicle.

[0026] Among them, the vehicle condition identification signal and the body deviation signal may be signals characterizing vehicle stability. The vehicle condition identification signal can be obtained based on at least one of a wheel speed signal, accelerations in N directions, a yaw angle, a tire pressure signal, a suspension height, a suspension stiffness, a suspension damping signal, a motor torque, a motor speed, a motor resolver signal, and a terrain identification signal. The body deviation signal may include the difference between the actual pose and the desired pose of the vehicle, where the N directions include at least one of a lateral direction, a longitudinal direction, and a vertical direction.

[0027] Specifically, as Figure 2 shown, the vehicle condition identification signal and the body deviation signal of the vehicle can be obtained through an identification module. The signals used to determine the vehicle condition identification signal may include: the wheel speed signal sent by a wheel speed sensor, the vehicle instantaneous state signals such as longitudinal acceleration, lateral acceleration, vertical acceleration, and yaw angle sent by an IMU (Inertial Measurement Unit), the vehicle instantaneous tire pressure signal sent by a high-frequency tire pressure sensor, the suspension height, stiffness, and damping signals sent by a suspension controller, the actual torque, actual speed, resolver signal, etc. sent by a motor controller, and the terrain identification signal (which can be a number, for example, the number 1 represents an urban terrain, the number 2 represents a highway terrain, the number 3 represents a sandy land, etc.) generated by the point cloud signal collected by vision and lidar through an image recognition method. The collected signals are cross-checked and the vehicle condition identification signal is output, for example, a mountain acceleration condition signal, a city deceleration condition signal, a highway condition signal, etc.

[0028] Among them, the cross-checking method may include multi-source signal redundancy check, multi-source signal fusion, multi-algorithm check, etc. For example, the resolver signal and the wheel speed signal can be selected according to different working conditions. When the working condition is identified as rainy or snowy weather, the resolver signal is used for working condition identification instead of the wheel speed signal. By cross-checking multiple signals for identifying working conditions, errors, noises or anomalies in the signals can be detected, ensuring the quality and consistency of the signals, thereby improving the reliability and accuracy of subsequent signal processing.

[0029] The vehicle body deviation signal refers to the difference between the desired pose and the actual pose. Among them, the desired pose can be the pose calculated based on the vehicle dynamics model (such as an eight-degree-of-freedom model) and the parameters corresponding to the working condition table at the previous moment, and the actual pose is the pose at the current moment calculated through GPS (Global Positioning System) signals and IMU signals. The difference between the two is the vehicle body deviation signal, which can be a vector v = [dx, dy, dz, dα, dβ, dγ]’, where dx, dy, and dz are the deviation amounts of the longitudinal, lateral, and vertical displacements respectively, and dα, dβ, and dγ are the deviation amounts of the longitudinal, lateral, and vertical rotation angles respectively.

[0030] S12. Determine the target control quantity based on the deep learning model, the working condition identification signal, and the vehicle body deviation signal, and control the vehicle based on the target control quantity.

[0031] In some embodiments of the present application, the working condition identification signal and the vehicle body deviation signal are used as the inputs of the deep learning model, and the target control quantity is used as the output of the deep learning model. The deep learning model can directly generate the target control quantity based on the working condition identification signal and the vehicle body deviation signal.

[0032] In some other embodiments of the present application, the working condition identification signal and the vehicle body deviation signal are used as the inputs of the deep learning model, and the predicted adjustment quantity is used as the output of the deep learning model. First, the preliminary control quantity can be obtained according to the working condition identification signal, and the predicted adjustment quantity can be obtained by using the deep learning model based on the working condition identification signal and the vehicle body deviation signal. Then, the preliminary control quantity is adjusted by using the predicted adjustment quantity to obtain the target control quantity.

[0033] Among them, the target control quantity includes control quantities in N directions, and the control object includes at least one of the vehicle's suspension, motor, engine, and steering wheel.

[0034] Exemplarily, obtaining the preliminary control quantity according to the working condition identification signal includes: searching for a preset corresponding relationship according to the working condition identification signal to obtain the preliminary control quantity.

[0035] Among them, the preset corresponding relationship, that is, the corresponding relationship between the working condition identification signal and the control quantity, can be stored in the form of a table.

[0036] Specifically, as Figure 3 shown, step S12 can be completed by the control module. The working condition recognition signal is input into the calibration table based on the preset model, and the preliminary control quantity is output by looking up the table; the working condition recognition signal and the vehicle body deviation signal are input into the trained deep learning model, and the predicted adjustment quantity is output; the sum of the preliminary control quantity and the predicted adjustment quantity is the target control quantity. The target control quantity can include control quantities in the longitudinal, lateral, and vertical directions. The lateral control quantity includes the target height of the suspension, the target stiffness of the suspension, the target damping of the suspension, etc. The longitudinal control quantity includes the target torque of the motor, the target speed of the motor, the target torque of the engine, the target braking force at the wheel end, etc. The measurement control quantity includes the target steering angle of the steering wheel, the target speed of the steering wheel, etc.

[0037] Exemplarily, the deep learning model can adopt a deep neural network structure and can be trained by the supervised learning module.

[0038] As Figure 4 shown, the supervised learning module takes the working condition recognition signal and the vehicle body deviation signal as inputs, takes the predicted control quantity as the output, and uses the predicted adjustment quantity and the manually labeled adjustment control quantity to train the deep learning network. The trained model can be directly used in the actual vehicle.

[0039] In some embodiments of the present application, the method for controlling vehicle stability further includes: obtaining the calculated adjustment quantity according to the vehicle body deviation signal by using the optimization model; optimizing the parameters of the deep learning model according to the calculated adjustment quantity and the predicted adjustment quantity.

[0040] Exemplarily, the optimization model is represented by the following formula:

[0041]

[0042] where, Δu t represents the calculated adjustment quantity at time t, represents the vehicle dynamics model (such as an eight-degree-of-freedom model), x represents the pose, u represents the control quantity, represents the actual pose at time t, represents the desired pose at time t, u t-1 represents the control quantity at time t-1.

[0043] Specifically, the deep learning model can be reinforced and self-learned by the reinforcement self-learning module during vehicle operation. As Figure 5As shown, the enhanced self-learning module realizes enhanced self-learning by inputting the vehicle body deviation signal into a preset optimization model during vehicle operation, automatically calculating the calculation adjustment amount, and using the calculation adjustment amount and the corresponding predicted adjustment amount to optimize the parameters of the deep learning model. It should be noted that since the optimization model in the enhanced self-learning module has high requirements for computing power, in actual use, the vehicle body deviation signal and the working condition recognition signal can be uploaded to the cloud for calculation, and the updated results are downloaded from the cloud to the vehicle terminal.

[0044] In some embodiments of the present application, as Figure 6 shown, the deep learning model includes a plurality of PINN modules ( Figure 6 illustrated by taking 3 PINN modules as an example in

[0045] connected in sequence) and a first fully connected layer. The PINN module includes a second fully connected layer, a Sin activation layer, a third fully connected layer, a first Tanh activation layer, a fourth fully connected layer, and a second Tanh activation layer connected in sequence. Specifically, the fully connected layer can integrate features, introduce nonlinearity, map outputs, etc., to enhance the expression ability and task adaptability of the deep learning model. Tanh activation refers to activation using the inverse tangent function. The inverse tangent function can enhance the expression ability of the deep learning model by introducing nonlinearity, normalizing the output range, and providing strong gradient characteristics; Sin activation refers to activation using the sine function. The sine function can enhance the expression ability of the deep learning model when processing periodic data by introducing periodic nonlinear characteristics. This PINN module has significant advantages in fitting and generalization, with high generalization ability. At the same time, combined with enhanced self-learning, the deep learning model can enhance the generalization ability for various uncertain situations.

[0046] Exemplarily, both the input and output of the deep learning model can be vectors.

[0047] After obtaining the target control quantity, the execution module can control the vehicle according to the target control quantity. Among them, the target control quantity can be sent to the corresponding electronic control units of the vehicle, such as the brake control unit, the steering control unit, the motor control unit, the suspension control unit, etc. in the form of a control instruction through the system integration and interaction module, and they perform specific control operations.

[0048] In some embodiments of the present application, controlling the vehicle based on the target control quantity includes: verifying the target control quantity to obtain the actual control quantity; controlling the vehicle using the actual control quantity.

[0049] In some examples, verifying the target control quantity to obtain the actual control quantity includes: performing line-following control on the target control quantity to obtain the actual control quantity.

[0050] Specifically, the follow-the-line control refers to the P control in the PID (Proportional Integral Derivative) control. For example, if the difference between the target control quantity and the actual control quantity at the current moment is e, then the adjustment quantity of the actual control quantity at the next moment is: Kp*e, where Kp is the calibrated quantity. In other words, the actual control quantity at the current moment is obtained based on the target control quantity and the actual control quantity at the previous moment. Through the follow-the-line control, fast response of the control can be achieved, and the dynamic performance can be improved.

[0051] In some examples, verifying the target control quantity to obtain the actual control quantity includes: comparing the target control quantity with a preset maximum allowable control quantity; if the target control quantity is greater than the preset maximum allowable control quantity, then using the preset maximum allowable control quantity as the actual control quantity; if the target control quantity is less than or equal to the preset maximum allowable control quantity, then using the target control quantity as the actual control quantity.

[0052] By comparing the target control quantity with the preset maximum allowable control quantity and taking the smaller value thereof for actual control, vehicle stability control can be performed within the control ability of the vehicle, avoiding vehicle instability or device damage. Of course, since the target control quantity may include multiple sub-control quantities for controlling different objects (such as the above-mentioned suspension target height, suspension target stiffness, suspension target damping, motor target torque, motor target speed, engine target torque, wheel-end target braking force, steering wheel target angle, steering wheel target speed, etc.), when verifying, each sub-control quantity needs to be verified separately.

[0053] Exemplarily, during verification, the follow-the-line control can also be performed first, and then the control quantity obtained by the follow-the-line control is compared with the corresponding preset maximum allowable control quantity, and the smaller value is taken as the actual control quantity for control.

[0054] The following combines Figure 8 to illustrate the execution process of the vehicle stability control method according to the embodiments of the present application:

[0055] First, after the vehicle is ignited and started, the vehicle stability control method starts to run accordingly and performs initialization operations, including: self-checking and calibrating each sensor in the recognition module to ensure that it can accurately collect the initial state information of the vehicle; the supervised learning module loads the pre-trained deep learning model parameters and sets them to the initial working state; the control module initializes relevant control variables and policy parameters, prepares to receive information from other modules and generate control instructions; the execution module establishes a communication connection with other electronic control units of the vehicle to ensure the smoothness of the data transmission channel.

[0056] Then, during the normal driving of the vehicle, the multi-source sensor integration unit of the recognition module collects various state information of the vehicle in real time at a preset frequency, including: the GPS sensor obtains the position and speed information of the vehicle, the IMU measures the acceleration and attitude changes of the vehicle, the wheel speed sensor monitors the rotational speed of the wheels, the IMU sensor detects the yaw angle and acceleration of the vehicle body, etc.; it also collects information related to the vehicle driving environment, such as road surface conditions (which can be obtained through road surface sensors or other existing systems of the vehicle), etc. These information are cross-checked and deviation calculated to obtain the working condition recognition signal and the vehicle body deviation signal, and are transmitted to the control module in real time, such as Figure 8 shown

[0057] After that, referring to Figure 8 , after the control module receives the working condition recognition signal and the vehicle body deviation signal transmitted by the recognition module, on the one hand, it inputs the working condition recognition signal into the preset calibration table, and on the other hand, it inputs the working condition recognition signal and the vehicle body deviation signal into the established deep learning model. The preset calibration table outputs the preliminary control quantity according to the working condition recognition signal; the deep learning model calculates through forward propagation, and according to the relationship between the deviation, working condition signal and adjustment quantity it has learned, predicts the adjustment quantity for the current deviation situation to obtain the predicted adjustment quantity. During the prediction process of the deep learning model, the deep learning model can also continuously perform self-learning and adjustment. Specifically, according to the new data and feedback information in the actual driving process of the vehicle, the reinforcement learning method can be used to online update and optimize parameters such as the weights and biases of the deep learning model, so as to make more accurate predictions and control adjustments when encountering similar situations in the future.

[0058] After obtaining the preliminary control quantity and the predicted adjustment quantity, the predicted adjustment quantity can be used to adjust the preliminary control quantity, such as adding the two, so as to formulate a specific handling stability control strategy (i.e., the target control quantity), including the control instructions required for each control object. For example, if the predicted adjustment quantity indicates that the vehicle is about to skid, the control strategy may be to appropriately increase the braking force of a certain wheel, and at the same time adjust the power assist size and steering wheel angle of the steering system, and optimize the output torque of the power system to maintain the stable driving of the vehicle.

[0059] Finally, to ensure the reliability and safety of the control, the execution module can verify the target control quantity to obtain the actual control quantity, and then use the actual control quantity to control the vehicle.

[0060] It should be noted that the above execution process also involves the operation monitoring of each module, the interactive monitoring between modules and between modules and other electronic control units, which can be realized by the monitoring unit, specifically as follows:

[0061] 1) Monitor whether the data communication between various modules is unobstructed, so as to ensure the real-time transmission of vehicle status deviation data, predicted control adjustment amounts, handling stability control strategy instructions, and other relevant information;

[0062] 2) Monitor the operating status of each module, the working conditions of sensors, and the accuracy of data transmission, etc. Once a failure or abnormal situation is detected, such as sensor failure, communication interruption, or calculation abnormality of the deep learning module, an alarm signal will be immediately sent to alert the driver and take corresponding emergency measures, such as switching to the standby control mode or restricting certain functions of the vehicle, to ensure the driving safety of the vehicle; at the same time, the fault information can also be recorded for subsequent maintenance and analysis;

[0063] 3) Coordinate and interact with other electronic control units of the vehicle to achieve the collaborative work of the entire vehicle system;

[0064] 4) Provide the driver with an intuitive display interface for the handling stability status information of the vehicle. Through the dashboard display screen or in-vehicle multimedia system, important information such as the current deviation of the vehicle, the execution effect of the control strategy, and the operating status of the system is shown to the driver; thus, the driver can understand the handling stability status of the vehicle through the prompts and information on the interface, and perform some simple operations and settings on the handling stability control system according to the actual situation, such as selecting different driving modes (if the system supports) or adjusting the sensitivity of certain control parameters.

[0065] The vehicle stability control method according to the embodiments of the present application can monitor the status deviation of the vehicle in real time and accurately. Utilizing the powerful learning ability and generalization ability of deep learning, it can adaptively and accurately predict the control adjustment amount, and accordingly generate an effective handling stability control strategy to achieve precise control and optimization improvement of the vehicle handling stability, thereby significantly improving the vehicle handling stability under various complex working conditions.

[0066] Figure 9 It is the structural block diagram of the vehicle stability control device according to the embodiments of the present application.

[0067] As Figure 9 shown, the vehicle stability control device 100 includes: an identification module 10, a control module 20, and an execution module 30. Among them, the identification module 10 is used to obtain the vehicle condition identification signal and the body deviation signal; the control module 20 is used to determine the target control amount based on the deep learning model, the vehicle condition identification signal, and the body deviation signal; the execution module 30 is used to control the vehicle according to the target control amount.

[0068] In some embodiments of the present application, determining a target control amount based on a deep learning model, a working condition recognition signal, and a vehicle body deviation signal includes: obtaining a preliminary control amount according to the working condition recognition signal; using the deep learning model to obtain a predicted adjustment amount according to the working condition recognition signal and the vehicle body deviation signal, and using the predicted adjustment amount to adjust the preliminary control amount to obtain the target control amount.

[0069] In some embodiments of the present application, the working condition recognition signal is obtained according to at least one of a wheel speed signal, accelerations in N directions, a yaw angle, a tire pressure signal, a suspension height, a suspension stiffness, a suspension damping signal, a motor torque, a motor speed, a motor resolver signal, and a terrain recognition signal, and the vehicle body deviation signal includes the difference between the actual pose and the desired pose of the vehicle, where the N directions include at least one of a lateral direction, a longitudinal direction, and a vertical direction.

[0070] In some embodiments of the present application, the deep learning model includes a plurality of PINN modules and a first fully connected layer connected in sequence, and the PINN module includes a second fully connected layer, a Sin activation layer, a third fully connected layer, a first Tanh activation layer, a fourth fully connected layer, and a second Tanh activation layer connected in sequence.

[0071] In some embodiments of the present application, obtaining a preliminary control amount according to the working condition recognition signal includes: looking up a preset corresponding relationship according to the working condition recognition signal to obtain the preliminary control amount.

[0072] In some embodiments of the present application, the control module 20 is further configured to: use an optimization model to obtain a calculated adjustment amount according to the vehicle body deviation signal; optimize the parameters of the deep learning model according to the calculated adjustment amount and the predicted adjustment amount.

[0073] In some embodiments of the present application, the optimization model is represented by the following formula:

[0074]

[0075] where, Δu t represents the calculated adjustment amount at time t, represents the vehicle dynamics model, x represents the pose, u represents the control amount, represents the actual pose at time t, represents the desired pose at time t, u t-1 represents the control amount at time t-1.

[0076] In some embodiments of the present application, controlling the vehicle according to the target control amount includes: verifying the target control amount to obtain an actual control amount; controlling the vehicle using the actual control amount.

[0077] In some embodiments of the present application, verifying the target control amount to obtain the actual control amount includes: performing line following control on the target control amount to obtain the actual control amount.

[0078] In some embodiments of the present application, verifying the target control amount to obtain the actual control amount includes: comparing the target control amount with a preset maximum allowable control amount; if the target control amount is greater than the preset maximum allowable control amount, then using the preset maximum allowable control amount as the actual control amount; if the target control amount is less than or equal to the preset maximum allowable control amount, then using the target control amount as the actual control amount.

[0079] In some embodiments of the present application, the target control amount includes control amounts in N directions, and the control object includes at least one of a vehicle's suspension, motor, engine, and steering wheel.

[0080] It should be noted that for other specific implementation manners of the vehicle stability control device in the embodiments of the present application, reference may be made to the specific implementation manners of the vehicle stability control method in the above embodiments.

[0081] Based on the vehicle stability control method in the above embodiments, the present application proposes a computer-readable storage medium.

[0082] In this embodiment, a computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the vehicle stability control method in the above embodiments is implemented.

[0083] Figure 10 It is a structural block diagram of the controller in the embodiments of the present application.

[0084] As Figure 10 shown, the controller 500 includes: a processor 501 and a memory 503. Among them, the processor 501 and the memory 503 are connected, such as connected through a bus 502. Optionally, the controller 500 may further include a transceiver 504. It should be noted that in actual applications, the transceiver 504 is not limited to one, and the structure of the controller 500 does not constitute a limitation to the embodiments of the present application.

[0085] The processor 501 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logic blocks, modules, and circuits described in connection with the disclosure of this application. The processor 501 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0086] The bus 502 may include a path for transmitting information between the above components. The bus 502 may be a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus 502 may be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 10 only a thick line is shown herein, but it does not mean that there is only one bus or one type of bus.

[0087] The memory 503 is used to store a computer program corresponding to the vehicle stability control method of the foregoing embodiments of this application, and the computer program is controlled and executed by the processor 501. The processor 501 is used to execute the computer program stored in the memory 503 to implement the content shown in the foregoing method embodiments.

[0088] Among them, the controller 500 includes, but is not limited to: a domain controller of the vehicle, a vehicle controller, etc. Figure 10 The illustrated controller 500 is only an example and should not impose any limitations on the functions and usage scope of the embodiments of this application.

[0089] This application also proposes a vehicle, including: the vehicle stability control device 200 of the foregoing embodiments, and / or, the controller 500 of the foregoing embodiments.

[0090] Note that the logic and / or steps represented in the flowchart or described otherwise herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or used in conjunction with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer diskette case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other appropriate processing as necessary, and then stored in a computer memory.

[0091] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0092] In the description of this specification, the description referring to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0093] In the description of the present application, it should be understood that the orientation or positional relationships indicated by the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. are based on the orientation or positional relationships shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the present application.

[0094] In addition, the terms "first" and "second" are only used for descriptive purposes and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of the present application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise specifically and clearly defined.

[0095] In the present application, unless otherwise clearly specified and limited, the terms such as "mounted", "connected", "connected to", "fixed", etc. should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or integrated; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the internal communication of two elements or the interaction relationship between two elements, unless otherwise clearly limited. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0096] In the present application, unless otherwise clearly specified and limited, the first feature being "on" or "under" the second feature may be that the first and second features are in direct contact, or the first and second features are indirectly in contact through an intermediate medium. Moreover, the first feature being "above", "over" and "on top of" the second feature may be that the first feature is directly above or obliquely above the second feature, or merely indicates that the first feature has a higher horizontal height than the second feature. The first feature being "under", "beneath" and "underneath" the second feature may be that the first feature is directly below or obliquely below the second feature, or merely indicates that the first feature has a lower horizontal height than the second feature.

[0097] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as a limitation to the present application. Those of ordinary skill in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present application.

Claims

1. A method for controlling vehicle stability, characterized in that, The method includes: Obtaining a working condition identification signal and a body deviation signal of the vehicle; Determining a target control quantity based on a deep learning model, the working condition identification signal, and the body deviation signal, and controlling the vehicle based on the target control quantity.

2. The control method for vehicle stability according to claim 1, characterized in that The determining the target control quantity based on the deep learning model, the working condition identification signal, and the body deviation signal includes: Obtaining a preliminary control quantity according to the working condition identification signal; Using the deep learning model to obtain a predicted adjustment quantity according to the working condition identification signal and the body deviation signal, and using the predicted adjustment quantity to adjust the preliminary control quantity to obtain the target control quantity.

3. The control method for vehicle stability according to claim 1, wherein The working condition identification signal is obtained according to at least one of a wheel speed signal, accelerations in N directions, a yaw angle, a tire pressure signal, a suspension height, a suspension stiffness, a suspension damping signal, a motor torque, a motor speed, a motor resolver signal, and a terrain identification signal. The body deviation signal includes a difference between an actual pose and a desired pose of the vehicle, where the N directions include at least one of a lateral direction, a longitudinal direction, and a vertical direction.

4. The control method for vehicle stability according to claim 1, characterized in that, The deep learning model includes a plurality of PINN modules and a first fully connected layer connected in sequence. The PINN module includes a second fully connected layer, a Sin activation layer, a third fully connected layer, a first Tanh activation layer, a fourth fully connected layer, and a second Tanh activation layer connected in sequence.

5. The control method for vehicle stability according to claim 2, wherein The obtaining the preliminary control quantity according to the working condition identification signal includes: Looking up a preset corresponding relationship according to the working condition identification signal to obtain the preliminary control quantity.

6. The control method for vehicle stability according to claim 2, characterized in that, The method further includes: Obtaining a calculated adjustment quantity according to the body deviation signal by using an optimization model; Optimizing parameters of the deep learning model according to the calculated adjustment quantity and the predicted adjustment quantity.

7. The control method for vehicle stability according to claim 6, characterized in that, The optimization model is represented by the following formula: where, Δu t represents the calculated adjustment amount at time t, represents the vehicle dynamics model, x represents the pose, u represents the control quantity, represents the actual pose at time t, represents the desired pose at time t, u t-1 represents the control quantity at time t - 1.

8. The control method for vehicle stability according to any one of claims 1-7, characterized in that, The controlling the vehicle based on the target control quantity includes: Verifying the target control quantity to obtain an actual control quantity; Controlling the vehicle by using the actual control quantity.

9. The control method for vehicle stability according to claim 8, wherein, The verifying the target control quantity to obtain the actual control quantity includes: Performing line following control on the target control quantity to obtain the actual control quantity.

10. The control method for vehicle stability according to claim 8, wherein, The verifying the target control quantity to obtain the actual control quantity includes: Comparing the target control quantity with a preset maximum allowable control quantity; If the target control quantity is greater than the preset maximum allowable control quantity, then using the preset maximum allowable control quantity as the actual control quantity; If the target control quantity is less than or equal to the preset maximum allowable control quantity, then using the target control quantity as the actual control quantity.

11. The control method for vehicle stability according to any one of claims 1-7, characterized in that The target control quantity includes control quantities in N directions, and the control objects include at least one of a suspension, a motor, an engine, and a steering wheel of the vehicle.

12. A control device for vehicle stability, characterized in that, The device includes: An identification module, configured to obtain a working condition identification signal and a body deviation signal of the vehicle; A control module, configured to determine a target control quantity based on a deep learning model, the working condition identification signal, and the body deviation signal; An execution module, configured to control the vehicle according to the target control quantity.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method according to any one of claims 1-11.

14. A controller, comprising a memory, a processor, and a computer program stored on the memory, characterized in that, When the computer program is executed by the processor, it implements the method according to any one of claims 1-11.

15. A vehicle, characterized in that, Comprising: A control device for vehicle stability according to claim 12, and / or, a controller according to claim 14.

Images