A vehicle chassis control method and system based on a two-way master-slave game and a vehicle

Abstract

This invention discloses a bidirectional master-slave game-based vehicle chassis control method, system, and vehicle, relating to the field of vehicle dynamics control. The method includes: calculating the ideal yaw rate under normal operating conditions; setting the ideal centroid sideslip angle to 0; calculating the comprehensive situational risk index; comparing the comprehensive situational risk index with the extreme instability critical boundary and the linear stability recovery boundary to determine the current game topology state variables; determining the leader and follower based on the current game topology state variables; using the ideal yaw rate and ideal centroid sideslip angle under normal operating conditions to form the desired control states of the leader and follower; solving the bidirectional master-slave game between the leader and follower to obtain the rear wheel steering angle and additional yaw moment; sending the rear wheel steering angle to the rear wheel actuator; and establishing a torque optimization allocation function based on the additional yaw moment to allocate the desired driving torque to each wheel. This invention improves the maneuverability of corner module vehicles under normal operating conditions and their stability under extreme operating conditions.

Description

Technical Field

[0001] This invention relates to the field of vehicle dynamics control, specifically to a vehicle chassis control method, system, and vehicle based on a two-way master-slave game. Background Technology

[0002] With the rapid development of steer-by-wire technology in automotive chassis, modular vehicles have become a cutting-edge research direction in intelligent electric vehicle chassis. These vehicles integrate the drive motor, steering actuator, and braking system into a single wheel hub module, achieving completely independent control of four-wheel power and steering angle, greatly expanding the vehicle's maneuverability. However, this high degree of control flexibility also brings extremely complex system coupling problems, especially under conditions of high-speed driving, large-angle steering, and complex road surface adhesion, where functional overlap and conflicts between various execution subsystems become particularly prominent.

[0003] Currently, yaw stability control for corner-mounted vehicles primarily relies on Direct Yaw Moment Control (DYC) and Active Rear-Wheel Steering (ARS). DYC generates additional yaw moment by adjusting the longitudinal force differential of the four wheels, exhibiting strong corrective capabilities and fast response; however, frequent intervention can lead to speed loss, increased tire wear, and reduced ride comfort. ARS, on the other hand, directly alters the tire lateral force by adjusting the rear wheel steering angle, resulting in better handling and higher energy efficiency. Existing research often employs rule-based allocation or fixed weighting strategies to coordinate these two systems. However, in real-world operating conditions, the control effectiveness of both systems exhibits significant nonlinear characteristics due to the influence of road adhesion conditions.

[0004] On high-traction surfaces, tire lateral forces have a large linear margin, allowing ARS to efficiently adjust vehicle motion with minimal impact on longitudinal dynamics. However, on low-traction surfaces, tire lateral forces easily reach saturation. Changing the rear wheel steering angle at this point not only fails to provide the necessary yaw moment but may also push the tire into the nonlinear saturation region, leading to skidding or even loss of control. Under these extreme conditions, DYC, with its longitudinal force adjustment capabilities, can more effectively pull the vehicle back from the brink of instability.

[0005] Traditional control frameworks often fix DYC or ARS as the dominant player or adopt an equal-power Nash game approach, which ignores the fundamental physical constraints on the allocation of control. While the Stackelberg model in game theory can well describe decision-making processes with hierarchical structures, fixed-role Stackelberg games cannot adapt to dynamically changing road conditions.

[0006] Therefore, developing a control strategy that can adaptively switch game priorities in real time based on road surface adhesion margin and achieve DYC and ARS coordination is of urgent engineering and academic value for improving the robustness of corner module vehicles under extreme conditions. Summary of the Invention

[0007] This invention addresses the shortcomings of existing technologies by providing a vehicle chassis control method, system, and vehicle based on a two-way master-slave game.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] A vehicle chassis control method based on a two-way master-slave game includes the following steps: Calculate the ideal yaw rate under normal operating conditions; set the ideal center of mass sideslip angle to 0. Calculate the sideslip angle of each tire and the longitudinal slip ratio of the wheel, and input the sideslip angle of each tire and the longitudinal slip ratio of the wheel into the Magic Formula tire model to calculate the longitudinal force and lateral force of each wheel in real time; use the longitudinal force and lateral force of each wheel to calculate the tire adhesion utilization rate, and extract the maximum value of the tire adhesion utilization rate of the rear axle wheels. A transient stability function based on the sideslip angle and yaw rate is constructed, and the transient stability of the current control cycle is evaluated. By combining the maximum tire adhesion utilization rate of the rear axle wheels with the transient stability of the current control cycle, a comprehensive situational risk index is obtained. By comparing the comprehensive situation risk index with the critical boundary of extreme instability and the linear stability recovery boundary, the current game topology state variables are determined. The dominant and follower are determined based on the values ​​of the current game topology state variables. A game topology state variable of 0 indicates that ARS is the dominant and DYC is the follower; a game topology state variable of 1 indicates that DYC is the dominant and ARS is the follower. The ideal yaw rate and ideal centroid sideslip angle under normal operating conditions are used to form the expected control states of the dominant and follower. The bidirectional master-slave game between the dominant and follower is solved using a distributed model predictive control algorithm to obtain the rear wheel steering angle and additional yaw moment. The rear wheel steering angle is sent to the rear wheel actuator; considering the constraints of the motor peak torque, additional yaw moment, total desired drive torque and road adhesion coefficient, a torque optimization allocation function is established, and the desired drive torque of each wheel is allocated based on the torque optimization allocation function.

[0010] To optimize the above technical solution, the specific measures also include: Furthermore, the calculation of the ideal yaw rate under normal operating conditions is specifically as follows: The ideal yaw rate is calculated using a linear two-degree-of-freedom vehicle dynamics model, and the amplitude of the ideal yaw rate is constrained, as shown in the following formula: (6); In the formula, It is the first under normal working conditions Ideal yaw rate for each control cycle The road surface adhesion coefficient, It is the acceleration due to gravity. For longitudinal velocity, Wheelbase , The distance from the center of gravity to the front axle. The distance from the center of gravity to the rear axle. For the steering angle of the vehicle's front wheels, For vehicle stability factors, , For the overall vehicle quality, and These are the front axle lateral stiffness and the rear axle lateral stiffness, respectively. It is a symbolic function.

[0011] Furthermore, the calculation of each tire slip angle and wheel longitudinal slip ratio specifically involves: The formula for calculating the tire slip angle of the four wheels is as follows: (8); In the formula, For the first Tire slip angle of each wheel; These represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. For the first The actual wheel angle of each wheel; The wheelbase of the vehicle. The distance from the center of gravity to the front axle. The distance from the center of gravity to the rear axle. For longitudinal velocity, For lateral velocity, This refers to the yaw rate; Based on the rotational angular velocity of each wheel and the longitudinal velocity of the wheel center, the longitudinal slip ratio of each wheel is calculated. longitudinal slip ratio of each wheel The expression is as follows: (9); In the formula, The first value measured by the wheel speed sensor The rotational angular velocity of each wheel; The effective rolling radius of the wheel, This represents the longitudinal velocity.

[0012] Furthermore, the step of inputting the slip angles of each tire and the longitudinal slip ratio of the wheel into the Magic Formula tire model to calculate the longitudinal and lateral forces of each wheel in real time is specifically as follows: (10); In the formula, For the first The longitudinal force of each wheel For the first Lateral force on each wheel; These represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. It is the first The longitudinal slip ratio of each wheel For the first Tire slip angle of each wheel The longitudinal fitting coefficients of the magic formula are denoted as stiffness factor, shape factor, peak factor, and curvature factor, respectively. The lateral fitting coefficients of the magic formula; The calculation of tire adhesion utilization rate using the longitudinal and lateral forces of each wheel, and the extraction of the maximum tire adhesion utilization rate of the rear axle wheels, specifically involves: Based on the friction circle theory, the tire adhesion utilization rate of each wheel is calculated in real time, using the following formula: (11); In the formula, For the first Tire adhesion utilization rate of each wheel; For the first The longitudinal force of each wheel For the first Lateral force on each wheel; For the first The current vertical load on each wheel; The road surface adhesion coefficient; The maximum tire adhesion utilization rate of the rear axle wheels is expressed by the following formula: (12); In the formula, It is the maximum tire adhesion utilization rate of the rear axle wheels. It is the tire adhesion utilization rate of the left rear wheel. It is the tire adhesion utilization rate of the right rear wheel; The construction of the transient stability function based on the sideslip angle and yaw rate, and the evaluation of the transient stability during the current control cycle, are specifically as follows: The expression for the transient stability function based on the sideslip angle and yaw rate is as follows: (13); In the formula, This is the sequence number of the control cycle. For the first The transient stability of each control cycle, if This indicates that the vehicle is within a stable envelope region. This indicates that the vehicle is in an unstable state; For the first Vehicle center of gravity sideslip angle per control cycle; For the first The actual yaw rate of the vehicle in each control cycle; The physical limit of steady-state yaw rate. ; To ensure the physical limit of the center of mass sideslip angle, The road surface adhesion coefficient, For longitudinal velocity, It is the acceleration due to gravity; The comprehensive situational risk index is obtained by combining the maximum tire adhesion utilization rate of the rear axle wheels and the transient stability of the current control cycle. (14); In the formula, As a comprehensive situation risk index, This is the tire limit weighting coefficient. This is the weighting coefficient for overall vehicle stability. It is the maximum tire adhesion utilization rate of the rear axle wheels. For the first Transient stability of each control cycle.

[0013] Furthermore, the specific steps for determining the current game topology state variables by comparing the comprehensive situation risk index with the extreme instability critical boundary and the linear stability recovery boundary are as follows: The overall situation risk index is The critical boundary of the ultimate instability is The linear stable recovery boundary is ; when When a situation is deemed to have a high risk of instability or when the rear wheel lateral force is nearing saturation, the current game topology state variables are adjusted accordingly. ; when When the game is in a linear safety margin region, let the current game topology state variable be determined. ; when When this happens, it enters the disturbance-resistant memory region, maintaining the state of the previous control cycle, and makes... .

[0014] Furthermore, the specific steps for solving the bidirectional master-slave game between the leader and followers using the distributed model predictive control algorithm are as follows: when At that time, DYC was the dominant force, and ARS was the follower. These are the current game topology state variables; When the dominant controller (DYC) is optimizing, the follower (ARS) will maintain the optimal control input sequence from the previous control cycle. DYC is defined in the first... The expected output sequence for each control cycle is: , ,in, for An identity matrix of dimension 1 To predict the step size, This is the desired control state of DYC, the current dominant controller in the control cycle. This represents the Kronecker product, used to extend the single-step desired state to the entire prediction time domain; ,in, It is the first under normal working conditions Ideal yaw rate for each control cycle It is the first Ideal centroid sideslip angle for each control cycle; Establish the cost function for DYC: (19); In the formula, It is the first The cost of one control cycle DYC It is the first The predicted output of DYC for each control cycle. DYC is in the The expected output sequence for each control cycle. The DYC outputs a weighted matrix for the tracking error. This is the control input sequence for DYC. The weighting matrix for the DYC control input; Transform equation (19) into standard least squares form: (25); Define the weighted prediction input matrix of DYC The weighted expectation vector of DYC ; In the formula, The DYC outputs a weighted matrix for the tracking error. The weighting matrix for the DYC control input. The output matrix of DYC at the prediction step size The block diagonal prediction matrix obtained by inner expansion; The discretized rear wheel steering angle input matrix is ​​in the prediction step size. The block diagonal prediction matrix obtained by inner expansion, It is the independent prediction error term of DYC; Based on equation (25), the least squares problem is to minimize the cost of DYC, as follows: (26); In each control cycle Within, the least squares problem is solved iteratively, and equation (26) is in the first... The solution for the next iteration is as follows: (27); In the formula, It is DYC's number The control input sequence for the next iteration, expanded, yields: (28); In the formula, The DYC outputs a weighted matrix for the tracking error. The weighting matrix for the DYC control input. The output matrix of DYC at the prediction step size The block diagonal prediction matrix obtained by inner expansion; The discretized rear wheel steering angle input matrix is ​​in the prediction step size. The block diagonal prediction matrix obtained by inner expansion, It is the independent prediction error term of DYC; Follower ARS obtained the first result calculated by leader DYC. The control input sequence for the next iteration Then, it is substituted into the prediction model (15). (15); In the formula, This is the state sequence vector within the prediction step; The discrete state input matrix is ​​used in the prediction step. The block diagonal prediction matrix obtained by inner expansion; This is the initial state of the current control cycle. The input matrix for the discretized rear wheel steering angle is in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; This is the control input sequence for DYC. The input matrix for the discretized additional yaw moment is at the prediction step size. The block diagonal prediction matrix obtained by inner expansion; This is the control input sequence for ARS. The discrete external disturbance input matrix is ​​used in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; It is an external disturbance sequence; The prediction output of ARS is obtained based on the state sequence vector within the prediction step. : (31); In the formula, The output matrix of ARS at the prediction step size The block diagonal prediction matrix obtained by inner expansion; Establish the cost function for ARS: (29) In the formula, It is the first The cost of ARS per control cycle It is the first The predicted output of ARS for each control cycle. It is ARS in the The expected output sequence for each control cycle. Output the weighted matrix of tracking error for ARS. The weighting matrix for ARS control inputs. This is the control input sequence for ARS; Using the leader DYC The control input sequence for the next iteration Update the independent prediction error term for the follower ARS : (30); In the formula, It is ARS in the The expected output sequence for each control cycle. The output matrix of ARS at the prediction step size The block diagonal prediction matrix obtained by inner expansion; The discrete state input matrix is ​​used in the prediction step. The block diagonal prediction matrix obtained by inner expansion; This is the initial state of the current control cycle. The input matrix for the discretized rear wheel steering angle is in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; The discrete external disturbance input matrix is ​​used in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; It is an external disturbance sequence; Output error for: (32); In the formula, It is the first The predicted output of ARS for each control cycle. It is ARS in the The expected output sequence for each control cycle. The output matrix of ARS at the prediction step size The block diagonal prediction matrix obtained by inner expansion, The input matrix for the discretized additional yaw moment is at the prediction step size. The block diagonal prediction matrix obtained by inner expansion; This is the control input sequence for ARS. This is the independent prediction error term for ARS; Therefore, equation (29) can be expressed as: (33); Rewritten in standard least squares form: (34); In the formula, the weighted prediction input matrix of ARS The weighted expectation vector of ARS , Output the weighted matrix of tracking error for ARS. The weighting matrix for the ARS control input; Solve for the control input sequence that minimizes the cost of the ARS, and obtain the first ARS. The control input sequence for the next iteration : (35); Expanding, we get: (36); Calculate the infinite norm error of the control input sequence between the two iterations, when the following convergence condition is met: (37); Or the number of iterations Reaching the set limit When the iteration terminates, the optimal control input sequence of DYC for the current control cycle is output. The optimal control input sequence of ARS Otherwise, p+1, proceed to the next iteration; where, It is the convergence threshold; Extract the first control element vector from the optimal control input sequence: (38); In the formula, It is the optimal rear wheel steering angle. It is the optimal additional yaw moment, used as the additional yaw moment for establishing the torque optimization distribution function. .

[0015] Furthermore, the specific steps for solving the bidirectional master-slave game between the leader and followers using the distributed model predictive control algorithm are as follows: when At that time, ARS was the leader and DYC was the follower. These are the current game topology state variables; The dominant ARS is based on the optimal control input sequence of the previous control cycle of DYC. First, optimize to obtain the first ARS. The control input sequence for the next iteration : (35); Follower DYC acquires After optimization, the first DYC is obtained. The control input sequence for the next iteration : (27); Calculate the infinite norm error of the control input sequence between the two iterations, when the following convergence condition is met: (37); Or the number of iterations Reaching the set limit When the iteration terminates, the optimal control input sequence of DYC for the current control cycle is output. The optimal control input sequence of ARS Otherwise, p+1, proceed to the next iteration; For DYC's The control input sequence for each iteration; It is the ARS's The control input sequence for each iteration; Extract the first control element vector from the optimal control input sequence: (38); In the formula, It is the optimal rear wheel steering angle. It is the optimal additional yaw moment, used as the additional yaw moment for establishing the torque optimization distribution function. .

[0016] Furthermore, the torque optimization allocation function is specifically as follows: (39); In the formula, This represents the peak torque of the motor. Vertical load for a single tire; The road surface adhesion coefficient, Tire radius; This represents the total desired driving torque. For the desired driving torque of a single wheel, These represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. and These are the front wheel track and the rear wheel track, respectively. To add yaw moment; The desired driving torque is allocated to each wheel based on the torque optimization allocation function. .

[0017] This invention also proposes a two-way master-slave game-based vehicle chassis control system, comprising: The vehicle state reference value calculation unit is used to calculate the ideal yaw rate under normal operating conditions, and sets the ideal center of gravity sideslip angle to 0. The risk assessment module calculates the sideslip angle and longitudinal slip ratio of each tire and inputs these values ​​into the Magic Formula tire model to calculate the longitudinal and lateral forces of each wheel in real time. It then uses these forces to calculate tire adhesion utilization and extracts the maximum tire adhesion utilization of the rear axle wheels. A transient stability function based on the center-of-gravity sideslip angle and yaw rate is constructed, and the transient stability of the current control cycle is evaluated. Finally, the module integrates the maximum tire adhesion utilization of the rear axle wheels and the transient stability of the current control cycle to obtain a comprehensive situational risk index. The role assignment module compares the comprehensive situation risk index with the extreme instability critical boundary and the linear stability recovery boundary to determine the current game topology state variables. Based on the value of the current game topology state variables, the dominant player and the follower are determined. A game topology state variable of 0 indicates that ARS is the dominant player and DYC is the follower. A game topology state variable of 1 indicates that DYC is the dominant player and ARS is the follower. The two-way master-slave game module is used to form the desired control states of the leader and the follower under normal working conditions by using the ideal yaw rate and the ideal centroid sideslip angle. The distributed model predictive control algorithm is used to solve the two-way master-slave game between the leader and the follower to obtain the rear wheel steering angle and the additional yaw moment. Rear wheel actuator, used to drive rear wheel steering based on rear wheel rotation angle; The driving torque distribution unit is used to consider the constraints of motor peak torque, additional yaw moment, total desired driving torque and road adhesion coefficient, establish a torque optimization distribution function, and distribute the desired driving torque of each wheel based on the torque optimization distribution function.

[0018] The present invention also proposes a vehicle including a vehicle chassis control system based on the bidirectional master-slave game described above.

[0019] The beneficial effects of this invention are as follows: By constructing a bidirectional master-slave game architecture based on distributed model predictive control, this invention achieves dynamic coordination between Direct Yaw Moment Control (DYC) and Active Rear Steering (ARS). Unlike traditional fixed master-slave structures, this invention dynamically reconstructs the master-slave game topology based on tire adhesion limits and vehicle instability using a hysteretic state machine: under extreme conditions, DYC is given decision priority to ensure yaw stability, while in the linear safety margin region, ARS takes the lead to improve vehicle steering agility and smoothness. This dynamic game strategy based on physical boundaries effectively avoids high-frequency oscillations in control power, breaks through the computational bottleneck of centralized control, and aligns with the nonlinear evolution of the vehicle's underlying dynamics, providing more precise and robust active chassis safety assurance for distributed drive vehicles in complex driving environments. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the vehicle chassis control method based on bidirectional master-slave game proposed in this invention. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0022] Example 1: This invention proposes a two-way master-slave game-based vehicle chassis control method, the principle of which is as follows: Figure 1 As shown.

[0023] This invention is based on a discrete vehicle model, and the process of building the discrete vehicle model is as follows: Considering the effects of additional yaw moment and rear wheel steering angle, we first establish a two-degree-of-freedom model for the vehicle's lateral and yaw directions: (1); In the formula, This is the vehicle state vector; for The first derivative, external disturbance ; Input the vehicle state matrix; The input matrix is ​​the external disturbance. For the overall vehicle weight; The distance from the center of gravity to the front axle. This is the distance from the center of mass to the rear axle; For the whole vehicle to be around Moment of inertia of the shaft; These are the front and rear axle lateral stiffness, respectively. This refers to the yaw rate; Longitudinal velocity; It is the centroid sideslip angle; This refers to the steering angle of the vehicle's front wheels.

[0024] Considering the effects of rear wheel steering angle and additional yaw moment on vehicle state, equation (1) can be rewritten as: (2); In the formula, Input matrix for rear wheel steering angle state; Input matrix for additional yaw moment state; This is the input for rear wheel steering angle control; For additional yaw moment control input; The rear wheel steering angle; To add yaw moment.

[0025] Equation (2) is discretized using an approximate discretization method: (3); In the formula, The input matrix is ​​the discretized vehicle state. The input matrix is ​​the discretized rear wheel steering angle state; The input matrix is ​​the discrete additional yaw moment state. The external disturbance input matrix is ​​the discretized form. For discretization time. for Real-time vehicle status. for Real-time vehicle status. for Rear wheel angle at moment for A yaw moment is constantly applied. for External disturbances are constant.

[0026] The steps of the vehicle chassis control method based on a two-way master-slave game based on a discrete vehicle model are as follows: S1: The ideal yaw rate is calculated using a linear two-degree-of-freedom vehicle dynamics model. The formula is as follows: (4); In the formula, For vehicle stability factors, , Wheelbase .

[0027] Meanwhile, considering road surface adhesion limitations, the desired yaw rate is constrained: (5); In the formula, The road surface adhesion coefficient, This is the acceleration due to gravity.

[0028] Ideal yaw rate under normal operating conditions for: (6); Ideal centroid sideslip angle Set to 0: (7).

[0029] S2: The corner module vehicle has four-wheel independent steering capability. Based on the velocity vector relationship of each wheel center in the vehicle coordinate system, the tire slip angle of the four wheels is calculated as follows: (8); In the formula, For the first Tire slip angle of each wheel; These represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. For the first The actual wheel angle of each wheel; The wheelbase of the vehicle. The distance from the center of gravity to the front axle. The distance from the center of gravity to the rear axle. For longitudinal velocity, For lateral velocity, This refers to the yaw rate; Based on the rotational angular velocity of each wheel and the longitudinal velocity of the wheel center, the longitudinal slip ratio of each wheel is calculated. longitudinal slip ratio of each wheel The expression is as follows: (9); In the formula, The first value measured by the wheel speed sensor The rotational angular velocity of each wheel; The effective rolling radius of the wheel, This represents the longitudinal velocity.

[0030] By inputting the slip angles of each tire and the longitudinal slip ratio of the wheel into the Magic Formula tire model, the longitudinal and lateral forces of each wheel are calculated in real time: (10); In the formula, For the first The longitudinal force of each wheel For the first Lateral force on each wheel; These represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. It is the first The longitudinal slip ratio of each wheel For the first Tire slip angle of each wheel The longitudinal fitting coefficients of the magic formula are denoted as stiffness factor, shape factor, peak factor, and curvature factor, respectively. These are the lateral fitting coefficients for the magic formula. All of these coefficients are obtained in advance from bench tests of specific tires under current vertical load and road adhesion conditions and stored in the controller.

[0031] Based on the friction circle theory, the tire adhesion utilization rate is calculated using the longitudinal and lateral forces of each wheel. However, considering that the control authority and lateral stiffness of the active rear wheel steering (ARS) system are limited by the physical state of the rear axle tires, the maximum value of the tire adhesion utilization rate of the rear axle wheels is extracted as an evaluation index of ARS control effectiveness.

[0032] The formula for calculating the tire adhesion utilization rate of each wheel is as follows: (11); In the formula, For the first Tire adhesion utilization rate of each wheel; For the first The longitudinal force of each wheel For the first Lateral force on each wheel; For the first The current vertical load on each wheel; The road surface adhesion coefficient; The maximum tire adhesion utilization rate of the rear axle wheels is expressed by the following formula: (12); In the formula, It is the maximum tire adhesion utilization rate of the rear axle wheels. It is the tire adhesion utilization rate of the left rear wheel. This refers to the tire adhesion utilization rate of the right rear wheel. When When this occurs, it indicates that at least one wheel on the rear axle is about to enter the nonlinear saturation region. If the rear wheel angle is still adjusted by ARS at this time, the rear wheel will lock up completely or slip, resulting in a serious fishtailing and loss of control of the vehicle.

[0033] A transient stability function based on the sideslip angle and yaw rate is constructed, and the transient stability during the current control cycle is evaluated. The expression for the transient stability function based on the sideslip angle and yaw rate is as follows: (13); In the formula, This is the sequence number of the control cycle. For the first The transient stability of each control cycle, if This indicates that the vehicle is within a stable envelope region. This indicates that the vehicle is in an unstable state; For the first Vehicle center of gravity sideslip angle per control cycle; For the first The actual yaw rate of the vehicle in each control cycle; The physical limit of steady-state yaw rate. ; To ensure the physical limit of the center of mass sideslip angle, The road surface adhesion coefficient, For longitudinal velocity, This is the acceleration due to gravity.

[0034] By combining the maximum tire adhesion utilization rate of the rear axle wheels with the transient stability during the current control cycle, a comprehensive situational risk index is obtained, as shown in the following formula: (14); In the formula, As a comprehensive situation risk index, This is the tire limit weighting coefficient. The overall vehicle stability weighting coefficient satisfies... . It is the maximum tire adhesion utilization rate of the rear axle wheels. For the first Transient stability of each control cycle. Through dynamic fusion. and This enables dual monitoring of ARS actuator limitations and vehicle yaw instability.

[0035] By comparing the comprehensive situation risk index with the critical boundary of extreme instability and the linear stability recovery boundary, the current game topology state variables are determined.

[0036] Define discrete state variables To characterize the master-slave topology of the game in the system. The characterization is "ARS as the dominant force and DYC as the follower"; The system is characterized as "DYC as the dominant entity and ARS as the follower." To prevent high-frequency jitter in control due to signal noise at the attachment limit boundary, the following asymmetric dead-zone hysteresis logic is designed: when When a situation is deemed to have a high risk of instability or the rear wheel lateral force is nearing saturation, a takeover mechanism is forcibly triggered, causing the current game topology state variables to change. ; when When the game is in a linear safety margin zone, control is smoothly transferred, and the current game topology state variables are determined to be within this range. ; when When this happens, it enters the disturbance-resistant memory region, maintaining the state of the previous control cycle, and makes... .

[0037] This represents the critical boundary for ultimate instability. This is the linearly stable recovery boundary.

[0038] S3: The ideal yaw rate and ideal centroid sideslip angle under normal operating conditions are used to form the expected control states of the leader and follower. The distributed model predictive control algorithm is used to solve the bidirectional master-slave game between the leader and follower to obtain the rear wheel steering angle and additional yaw moment.

[0039] Based on distributed model predictive control theory, in the... For each control cycle, the standard form of the centralized multi-step predictive state equation of the system is: (15); In the formula: ; ; ; ; ; ; ; ; To predict the step size; This is the sequence number of the control cycle; This is a vector of state sequences in the time domain for prediction. This represents the initial state of the current control cycle; Indicates the first The first control cycle The predicted vehicle status of the step, The control input sequence for DYC; Indicates the first The first control cycle rear wheel angle This is the control input sequence for ARS; Indicates the first The first control cycle Additional yaw moment of the step This is an external perturbation sequence. Indicates the first External disturbances during each control cycle; The discrete state input matrix is ​​used in the prediction step. The block diagonal prediction matrix obtained by inner expansion; The input matrix for the discretized rear wheel steering angle is in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; The input matrix for the discretized additional yaw moment is at the prediction step size. The block diagonal prediction matrix obtained by inner expansion; The discrete external disturbance input matrix is ​​used in the prediction step size. The block diagonal prediction matrix obtained by inner expansion.

[0040] Define the desired control states of DYC and ARS at the current time as follows: and To achieve joint control of vehicle stability, the desired states of both are consistent, namely: (16); in, It is the first under normal working conditions Ideal yaw rate for each control cycle; For the first The ideal sideslip angle of the vehicle's center of gravity during each control cycle.

[0041] Based on this, DYC and ARS are defined in terms of prediction step size. Expected output reference sequence within and for: (17); In the formula, for An identity matrix of dimension 1 This represents the Kronecker product, used to extend the single-step desired state to the entire prediction time domain.

[0042] Define participants The predicted output is: (18); In the formula, DYC represents It represents ARS. For participants The output matrix in the prediction time domain The block diagonal prediction matrix obtained by inner expansion; ; The output matrix contains the output matrix of DYC. The output matrix of ARS , for An identity matrix of dimension 1.

[0043] when At that time, DYC was the dominant force, and ARS was the follower. These are the current game topology state variables; When the dominant controller (DYC) is optimizing, the follower (ARS) will maintain the optimal control input sequence from the previous control cycle. DYC is defined in the first... The expected output sequence for each control cycle is: , ,in, for An identity matrix of dimension 1 To predict the step size, This is the desired control state of DYC, the current dominant controller in the control cycle. This represents the Kronecker product, used to extend the single-step desired state to the entire prediction time domain; ,in, It is the first under normal working conditions Ideal yaw rate for each control cycle It is the first Ideal centroid sideslip angle for each control cycle; Establish the cost function for DYC: (19); In the formula, It is the first The cost of one control cycle DYC It is the first The predicted output of DYC for each control cycle. DYC is in the The expected output sequence for each control cycle. The DYC outputs a weighted matrix for the tracking error. This is the control input sequence for DYC. The weighting matrix for the DYC control input; To transform the optimization problem into a standard least squares solution, we extract and Irrelevant independent prediction error terms : (20); Substituting equation (17) into equation (20), we obtain the independent prediction error term for DYC: (twenty one); According to equation (18), we can obtain: (twenty two); Therefore, equation (19) can be expressed as: (twenty four); Rewrite equation (24) in standard least squares form: (25); Define the weighted prediction input matrix of DYC The weighted expectation vector of DYC ; In the formula, The DYC outputs a weighted matrix for the tracking error. The weighting matrix for the DYC control input. The output matrix of DYC at the prediction step size The block diagonal prediction matrix obtained by inner expansion; The discretized rear wheel steering angle input matrix is ​​in the prediction step size. The block diagonal prediction matrix obtained by inner expansion, It is the independent prediction error term of DYC; Based on equation (25), the least squares problem is to minimize the cost of DYC, as follows: (26); In each control cycle Within this framework, the least squares problem is solved iteratively, with an initial number of iterations. Equation (26) in the first The solution for the next iteration is as follows: (27); In the formula, It is DYC's number The control input sequence for the next iteration, expanded, yields: (28); In the formula, The DYC outputs a weighted matrix for the tracking error. The weighting matrix for the DYC control input. The output matrix of DYC at the prediction step size The block diagonal prediction matrix obtained by inner expansion; The discretized rear wheel steering angle input matrix is ​​in the prediction step size. The block diagonal prediction matrix obtained by inner expansion, It is the independent prediction error term of DYC; Follower ARS obtained the first result calculated by leader DYC. The control input sequence for the next iteration Then, it is substituted into the prediction model (15). (15); In the formula, This is the state sequence vector within the prediction step; The discrete state input matrix is ​​used in the prediction step. The block diagonal prediction matrix obtained by inner expansion; This is the initial state of the current control cycle. The input matrix for the discretized rear wheel steering angle is in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; This is the control input sequence for DYC. The input matrix for the discretized additional yaw moment is at the prediction step size. The block diagonal prediction matrix obtained by inner expansion; This is the control input sequence for ARS. The discrete external disturbance input matrix is ​​used in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; It is an external disturbance sequence; Establish the cost function for ARS: (29) In the formula, It is the first The cost of ARS per control cycle It is the first The predicted output of ARS for each control cycle. It is ARS in the The expected output sequence for each control cycle. Output the weighted matrix of tracking error for ARS. The weighting matrix for ARS control inputs. This is the control input sequence for ARS; Using the leader DYC The control input sequence for the next iteration Update the independent prediction error term for the follower ARS : (30); In the formula, It is ARS in the The expected output sequence for each control cycle. The output matrix of ARS at the prediction step size The block diagonal prediction matrix obtained by inner expansion; The discrete state input matrix is ​​used in the prediction step. The block diagonal prediction matrix obtained by inner expansion; This is the initial state of the current control cycle. The input matrix for the discretized rear wheel steering angle is in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; The discrete external disturbance input matrix is ​​used in the prediction step size. The block diagonal prediction matrix obtained by inner expansion; It is an external disturbance sequence; The prediction output of ARS is obtained based on the state sequence vector within the prediction step. : (31); In the formula, The output matrix of ARS at the prediction step size The block diagonal prediction matrix obtained by inner expansion; Output error for: (32); In the formula, It is the first The predicted output of ARS for each control cycle. It is ARS in the The expected output sequence for each control cycle. The output matrix of ARS at the prediction step size The block diagonal prediction matrix obtained by inner expansion, The input matrix for the discretized additional yaw moment is at the prediction step size. The block diagonal prediction matrix obtained by inner expansion; This is the control input sequence for ARS. This is the independent prediction error term for ARS; Therefore, equation (29) can be expressed as: (33); Rewritten in standard least squares form: (34); In the formula, the weighted prediction input matrix of ARS The weighted expectation vector of ARS , Output the weighted matrix of tracking error for ARS. The weighting matrix for the ARS control input; Solve for the control input sequence that minimizes the cost of the ARS, and obtain the first ARS. The control input sequence for the next iteration : (35); Expanding, we get: (36); Calculate the infinite norm error of the control input sequence between the two iterations, when the following convergence condition is met: (37); Or the number of iterations Reaching the set limit When the iteration terminates, the optimal control input sequence of DYC for the current control cycle is output. The optimal control input sequence of ARS Otherwise, p+1, proceed to the next iteration; where, It is the convergence threshold; in the next iteration, the leader DYC first uses the control input sequence of the follower ARS from the previous iteration. Substituting these values ​​into the formula for solving the independent prediction error term, the updated control input sequence of the dominant force is obtained. Subsequently, the follower ARS received the latest... Substituting the result into its own cost function, the updated control input sequence of the follower is obtained. .

[0044] Extract the first control element vector from the optimal control input sequence: (38); In the formula, It is the optimal rear wheel steering angle. It is the optimal additional yaw moment, used as the additional yaw moment for establishing the torque optimization distribution function. .

[0045] when At that time, ARS was the leader and DYC was the follower. These are the current game topology state variables; The dominant ARS is based on the optimal control input sequence of the previous control cycle of DYC. First, optimize to obtain the first ARS. The control input sequence for the next iteration : (35); Follower DYC acquires After optimization, the first DYC is obtained. The control input sequence for the next iteration : (27); Calculate the infinite norm error of the control input sequence between the two iterations, when the following convergence condition is met: (37); Or the number of iterations Reaching the set limit When the iteration terminates, the optimal control input sequence of DYC for the current control cycle is output. The optimal control input sequence of ARS Otherwise, p+1, proceed to the next iteration; For DYC's The control input sequence for each iteration; It is the ARS's The control input sequence for each iteration; It is the convergence threshold; Extract the first control element vector from the optimal control input sequence: (38); In the formula, It is the optimal rear wheel steering angle. It is the optimal additional yaw moment, used as the additional yaw moment for establishing the torque optimization distribution function. .

[0046] S4: Send the rear wheel steering angle to the rear wheel actuator; considering the constraints of motor peak torque, additional yaw moment, total desired drive torque and road adhesion coefficient, establish a torque optimization allocation function, and allocate the desired drive torque of each wheel based on the torque optimization allocation function.

[0047] The torque optimization allocation function is as follows: (39); In the formula, This represents the peak torque of the motor. Vertical load for a single tire; The road surface adhesion coefficient, Tire radius; This represents the total desired driving torque. For the desired driving torque of a single wheel, These represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. and These are the front wheel track and the rear wheel track, respectively. To add yaw moment; The desired driving torque is allocated to each wheel based on the torque optimization allocation function. .

[0048] Example 2: This invention proposes a bidirectional master-slave game-based vehicle chassis control system corresponding to the method in Embodiment 1, comprising: The vehicle state reference value calculation unit is used to calculate the ideal yaw rate under normal operating conditions, and sets the ideal center of gravity sideslip angle to 0. The risk assessment module calculates the sideslip angle and longitudinal slip ratio of each tire and inputs these values ​​into the Magic Formula tire model to calculate the longitudinal and lateral forces of each wheel in real time. It then uses these forces to calculate tire adhesion utilization and extracts the maximum tire adhesion utilization of the rear axle wheels. A transient stability function based on the center-of-gravity sideslip angle and yaw rate is constructed, and the transient stability of the current control cycle is evaluated. Finally, the module integrates the maximum tire adhesion utilization of the rear axle wheels and the transient stability of the current control cycle to obtain a comprehensive situational risk index. The role assignment module compares the comprehensive situation risk index with the extreme instability critical boundary and the linear stability recovery boundary to determine the current game topology state variables. Based on the value of the current game topology state variables, the dominant player and the follower are determined. A game topology state variable of 0 indicates that ARS is the dominant player and DYC is the follower. A game topology state variable of 1 indicates that DYC is the dominant player and ARS is the follower. The two-way master-slave game module is used to form the desired control states of the leader and the follower under normal working conditions by using the ideal yaw rate and the ideal centroid sideslip angle. The distributed model predictive control algorithm is used to solve the two-way master-slave game between the leader and the follower to obtain the rear wheel steering angle and the additional yaw moment. Rear wheel actuator, used to drive rear wheel steering based on rear wheel rotation angle; The driving torque distribution unit is used to consider the constraints of motor peak torque, additional yaw moment, total desired driving torque and road adhesion coefficient, establish a torque optimization distribution function, and distribute the desired driving torque of each wheel based on the torque optimization distribution function.

[0049] The implementation methods of each module and its function in the system are completely consistent with the steps of the method in Implementation Example 1, so they will not be repeated here.

[0050] Example 3: The present invention proposes a vehicle that includes the bidirectional master-slave game-based vehicle chassis control system of Embodiment 2.

[0051] Example 4: The present invention proposes a computer-readable storage medium storing a computer program that enables a computer to execute a vehicle chassis control method based on a two-way master-slave game, as described in Example 1.

[0052] In the embodiments disclosed in this application, a computer storage medium may be a tangible medium that may contain or store programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0053] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0054] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A vehicle chassis control method based on a two-way master-slave game, characterized in that, Includes the following steps: Calculate the ideal yaw rate under normal operating conditions; set the ideal centroid sideslip angle to 0; the calculation of the ideal yaw rate under normal operating conditions specifically involves: The ideal yaw rate is calculated using a linear two-degree-of-freedom vehicle dynamics model, and the amplitude of the ideal yaw rate is constrained, as shown in the following formula: （6）； In the formula, It is the first under normal working conditions Ideal yaw rate for each control cycle The road surface adhesion coefficient, It is the acceleration due to gravity. For longitudinal velocity, Wheelbase , The distance from the center of gravity to the front axle. The distance from the center of gravity to the rear axle. For the steering angle of the vehicle's front wheels, For vehicle stability factors, , For the overall vehicle quality, and These are the front axle lateral stiffness and the rear axle lateral stiffness, respectively. It is a symbolic function; Calculate the sideslip angle of each tire and the longitudinal slip ratio of the wheel, and input the sideslip angle of each tire and the longitudinal slip ratio of the wheel into the Magic Formula tire model to calculate the longitudinal force and lateral force of each wheel in real time; use the longitudinal force and lateral force of each wheel to calculate the tire adhesion utilization rate, and extract the maximum value of the tire adhesion utilization rate of the rear axle wheels. A transient stability function based on the sideslip angle and yaw rate is constructed, and the transient stability of the current control cycle is evaluated. By combining the maximum tire adhesion utilization rate of the rear axle wheels with the transient stability of the current control cycle, a comprehensive situational risk index is obtained. By comparing the comprehensive situation risk index with the critical boundary of extreme instability and the linear stability recovery boundary, the current game topology state variables are determined. The dominant and follower are determined based on the values ​​of the current game topology state variables. A game topology state variable of 0 indicates that ARS is the dominant and DYC is the follower; a game topology state variable of 1 indicates that DYC is the dominant and ARS is the follower. The ideal yaw rate and ideal centroid sideslip angle under normal operating conditions are used to form the expected control states of the dominant and follower. The bidirectional master-slave game between the dominant and follower is solved using a distributed model predictive control algorithm to obtain the rear wheel steering angle and additional yaw moment. The rear wheel steering angle is sent to the rear wheel actuator; considering the constraints of the motor peak torque, additional yaw moment, total desired drive torque and road adhesion coefficient, a torque optimization allocation function is established, and the desired drive torque of each wheel is allocated based on the torque optimization allocation function.

2. The vehicle chassis control method based on bidirectional master-slave game theory as described in claim 1, characterized in that, The calculation of each tire slip angle and wheel longitudinal slip ratio is specifically as follows: The formula for calculating the tire slip angle of the four wheels is as follows: （8）； In the formula, For the first Tire slip angle of each wheel; These represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. For the first The actual wheel angle of each wheel; The wheelbase of the vehicle. The distance from the center of gravity to the front axle. The distance from the center of gravity to the rear axle. For longitudinal velocity, For lateral velocity, This refers to the yaw rate; Based on the rotational angular velocity of each wheel and the longitudinal velocity of the wheel center, the longitudinal slip ratio of each wheel is calculated. longitudinal slip ratio of each wheel The expression is as follows: （9）； In the formula, The first value measured by the wheel speed sensor The rotational angular velocity of each wheel; The effective rolling radius of the wheel, This represents the longitudinal velocity.

3. The vehicle chassis control method based on bidirectional master-slave game theory as described in claim 1, characterized in that, The construction of the transient stability function based on the sideslip angle and yaw rate, and the evaluation of the transient stability during the current control cycle, are specifically as follows: The expression for the transient stability function based on the sideslip angle and yaw rate is as follows: （13）； In the formula, This is the sequence number of the control cycle. For the first The transient stability of each control cycle, if This indicates that the vehicle is within a stable envelope region. This indicates that the vehicle is in an unstable state; For the first Vehicle center of gravity sideslip angle per control cycle; For the first The actual yaw rate of the vehicle in each control cycle; The physical limit of steady-state yaw rate. ; To ensure the physical limit of the center of mass sideslip angle, The road surface adhesion coefficient, For longitudinal velocity, It is the acceleration due to gravity; The comprehensive situational risk index is obtained by combining the maximum tire adhesion utilization rate of the rear axle wheels and the transient stability of the current control cycle. （14）； In the formula, As a comprehensive situation risk index, This is the tire limit weighting coefficient. This is the weighting coefficient for overall vehicle stability. It is the maximum tire adhesion utilization rate of the rear axle wheels. For the first Transient stability of each control cycle.

4. The vehicle chassis control method based on bidirectional master-slave game theory as described in claim 1, characterized in that, The specific steps for determining the current game topology state variables by comparing the comprehensive situation risk index with the extreme instability critical boundary and the linear stability recovery boundary are as follows: The overall situation risk index is The critical boundary of the ultimate instability is The linear stable recovery boundary is , It is the sequence number of the control cycle; when When a situation is deemed to have a high risk of instability or when the rear wheel lateral force is nearing saturation, the current game topology state variables are adjusted accordingly. ; when When the game is in a linear safety margin region, let the current game topology state variable be determined. ; when When this happens, it enters the disturbance-resistant memory region, maintaining the state of the previous control cycle, and makes... .

5. The vehicle chassis control method based on bidirectional master-slave game theory as described in claim 1, characterized in that, The specific steps for solving the two-way master-slave game between the leader and the follower using a distributed model predictive control algorithm are as follows: when At that time, DYC was the dominant force, and ARS was the follower. These are the current game topology state variables; When the dominant controller (DYC) is optimizing, the follower (ARS) will maintain the optimal control input sequence from the previous control cycle. The least squares problem for the dominant DYC during optimization is to minimize the cost of DYC, as follows: （26）； In the formula, This is the control input sequence for DYC. It is the weighted prediction input matrix of DYC. It is the weighted expectation vector of DYC; In each control cycle Within, the least squares problem is solved iteratively, and equation (26) is in the first... The solution for the next iteration is as follows: （27）； In the formula, It is DYC's number The control input sequence for each iteration; Follower ARS obtains the first [number] calculated by leader DYC. The control input sequence for the next iteration Then, the cost function of ARS in standard least squares form is established: （34）； In the formula, It is the first The cost of ARS per control cycle It is the weighted prediction input matrix of ARS. It is the weighted expectation vector of ARS. This is the control input sequence for ARS; Solve for the control input sequence that minimizes the cost of the ARS, and obtain the first ARS sequence. The control input sequence for the next iteration : （35）； Calculate the infinite norm error of the control input sequence between the two iterations. When the infinite norm error converges or the number of iterations reaches zero, the result is obtained. Reaching the set limit When the iteration terminates, the optimal control input sequence of DYC for the current control cycle is output. The optimal control input sequence of ARS Otherwise, p+1, proceed to the next iteration; Extract the first control element vector from the optimal control input sequence: （38）； In the formula, It is the optimal rear wheel steering angle. It is the optimal additional yaw moment, used as the additional yaw moment for establishing the torque optimization distribution function. .

6. The vehicle chassis control method based on bidirectional master-slave game theory as described in claim 1, characterized in that, The specific steps for solving the two-way master-slave game between the leader and the follower using a distributed model predictive control algorithm are as follows: when At that time, ARS was the leader and DYC was the follower. These are the current game topology state variables; The dominant ARS is based on the optimal control input sequence of the previous control cycle of DYC. First, optimize to obtain the first ARS. The control input sequence for the next iteration : （35）； In the formula, It is the weighted prediction input matrix of ARS. It is the weighted expectation vector of ARS. It is the sequence number of the control cycle; Follower DYC acquires After optimization, the first DYC is obtained. The control input sequence for the next iteration : （27）； In the formula, It is the weighted prediction input matrix of DYC. It is the weighted expectation vector of DYC; Calculate the infinite norm error of the control input sequence between the two iterations, when the following convergence condition is met: （37）； Or the number of iterations Reaching the set limit When the iteration terminates, the optimal control input sequence of DYC for the current control cycle is output. The optimal control input sequence of ARS Otherwise, p+1, proceed to the next iteration; For DYC's The control input sequence for each iteration; It is the ARS's The control input sequence for each iteration; It is the convergence threshold; Extract the first control element vector from the optimal control input sequence: （38）； In the formula, It is the optimal rear wheel steering angle. It is the optimal additional yaw moment, used as the additional yaw moment for establishing the torque optimization distribution function. .

7. The vehicle chassis control method based on bidirectional master-slave game theory as described in claim 1, characterized in that, The torque optimization allocation function is specifically as follows: （39）； In the formula, This represents the peak torque of the motor. Vertical load for a single tire; The road surface adhesion coefficient, Tire radius; This represents the total desired driving torque. For the desired driving torque of a single wheel, These represent the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively. and These are the front track width and the rear track width, respectively. To add yaw moment; The desired driving torque is allocated to each wheel based on the torque optimization allocation function. .

8. A vehicle chassis control system based on a two-way master-slave game, characterized in that, include: The vehicle state reference value calculation unit is used to calculate the ideal yaw rate under normal operating conditions, setting the ideal center-of-gravity sideslip angle to 0; the calculation of the ideal yaw rate under normal operating conditions specifically involves: The ideal yaw rate is calculated using a linear two-degree-of-freedom vehicle dynamics model, and the amplitude of the ideal yaw rate is constrained, as shown in the following formula: （6）； In the formula, It is the first under normal working conditions Ideal yaw rate for each control cycle The road surface adhesion coefficient, It is the acceleration due to gravity. For longitudinal velocity, Wheelbase , The distance from the center of gravity to the front axle. The distance from the center of gravity to the rear axle. For the steering angle of the vehicle's front wheels, For vehicle stability factors, , For the overall vehicle quality, and These are the front axle lateral stiffness and the rear axle lateral stiffness, respectively. It is a symbolic function; The risk assessment module calculates the sideslip angle and longitudinal slip ratio of each tire and inputs these values ​​into the Magic Formula tire model to calculate the longitudinal and lateral forces of each wheel in real time. It then uses these forces to calculate tire adhesion utilization and extracts the maximum tire adhesion utilization of the rear axle wheels. A transient stability function based on the center-of-gravity sideslip angle and yaw rate is constructed, and the transient stability of the current control cycle is evaluated. Finally, the module integrates the maximum tire adhesion utilization of the rear axle wheels and the transient stability of the current control cycle to obtain a comprehensive situational risk index. The role assignment module compares the comprehensive situation risk index with the extreme instability critical boundary and the linear stability recovery boundary to determine the current game topology state variables. Based on the value of the current game topology state variables, the dominant player and the follower are determined. A game topology state variable of 0 indicates that ARS is the dominant player and DYC is the follower. A game topology state variable of 1 indicates that DYC is the dominant player and ARS is the follower. The two-way master-slave game module is used to form the desired control states of the leader and the follower under normal working conditions by using the ideal yaw rate and the ideal centroid sideslip angle. The distributed model predictive control algorithm is used to solve the two-way master-slave game between the leader and the follower to obtain the rear wheel steering angle and the additional yaw moment. Rear wheel actuator, used to drive rear wheel steering based on rear wheel rotation angle; The driving torque distribution unit is used to consider the constraints of motor peak torque, additional yaw moment, total desired driving torque and road adhesion coefficient, establish a torque optimization distribution function, and distribute the desired driving torque of each wheel based on the torque optimization distribution function.

9. A vehicle, characterized in that, The vehicle chassis control system includes the two-way master-slave game as described in claim 8.

Images