Distributed driving electric vehicle stability control method for shared driving

Abstract

The invention discloses a shared driving-oriented distributed driving electric vehicle stability control method, which comprises the following steps of: firstly, establishing a human-vehicle dynamics model, collecting vehicle state information in real time by combining a vehicle-mounted sensor, and processing the uncertainty problem of a system by adopting a finite vertex polyhedron method; the method comprises the following steps: firstly, establishing an active steering and additional yawing moment distributed parallel control method based on the game theory, then establishing a robust compensation-based control strategy to reduce the influence of system disturbance and uncertainty on system control, and finally, establishing a robust compensation-based control strategy. A man-machine sharing control strategy based on a dynamic intervention penalty factor is designed to improve the tracking performance and the driving performance of a shared driving vehicle. According to the method, the auxiliary compensation control logic that an auxiliary control system is arranged on a bottom layer is broken, man-machine sharing control is selected to be arranged on the control bottom layer logic, man-machine conflicts are effectively weakened, and the driving performance of man-machine sharing driving is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of distributed electric vehicle intelligent interaction and vehicle advanced assisted driving, and in particular to a distributed drive electric vehicle stability control method for shared driving. Background Art

[0002] With the revolution of automobile electrification and intelligent technology, the vehicle chassis has ushered in a technological transformation from traditional chassis, electric chassis to intelligent chassis, realizing the deep integration of subsystems such as active steering, wire-controlled drive / brake and intelligent suspension. Distributed drive electric vehicle chassis technology directly drives the wheels by wheel hub motors, simplifies the chassis structure and increases the control freedom. It has the advantages of short transmission chain, high efficiency and flexible space layout, and can effectively improve the vehicle's dynamics and safety. However, due to the coupling relationship between the wheel hub motor drive / brake system, active suspension system and independent steering system in dynamics and kinematics, there are overlaps and interferences in the control areas, resulting in obvious subsystem coupling conflicts and difficult to achieve chassis multi-objective control. In addition, considering that the control inputs of the driver and the active safety control system have redundant and game characteristics, and the driver's control characteristics have strong randomness and uncertainty, it will greatly affect the vehicle's handling quality, making it difficult to ensure vehicle stability. Therefore, establishing a reasonable and effective real-time dynamic coordination control scheme for multiple actuators is very important for improving the active safety of distributed electric vehicles for shared driving and promoting the transformation of the automobile industry. In particular, the auxiliary control system is set as the logical upper layer, and the human-machine shared driving is set as the logical lower layer, which provides a theoretical basis and technical support for the research on distributed drive electric vehicles for shared driving in my country. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a coordinated control method that can dynamically coordinate the control input of distributed drive electric vehicles under human-machine shared driving, effectively eliminate the coupling conflicts and human-machine conflicts that may be caused by incomplete actuator information and overlapping interference of functions, and provide a distributed drive electric vehicle human-machine shared dynamic coordinated control method with high intelligence level and strong practicality.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0005] A distributed drive electric vehicle stability control method based on shared driving is provided, comprising the following steps

[0006] Step S1: according to the control requirements of distributed drive electric vehicles, a steering system dynamics model based on torque control and a vehicle dynamics model are established, and the steering system dynamics model and the vehicle dynamics model are combined to establish a human-vehicle system dynamics model;

[0007] Step S2: for the human-vehicle system dynamics model established in step S1, the vertex polyhedron method is used to deal with the uncertainty in the system, and the state equation of the human-vehicle system after linearization is established;

[0008] Step S3: establishing a distributed drive electric vehicle active steering and additional yaw moment control strategy based on terminal constraint cooperative game;

[0009] Step S4: In view of the problem that cooperative game is difficult to eliminate system disturbances, a robust adaptive compensation control strategy is established to achieve efficient and safe dynamic coordinated control of the distributed electric vehicle active steering system and the additional yaw moment control system;

[0010] Step S5: Establishing the driver preview time t p and a linear parameter model with variable parameters of the penalty factor κ;

[0011] Step S6: Establish a penalty factor κ adjustment method based on fuzzy rules and design a gain-scheduled robust active steering control strategy to alleviate human-machine execution conflicts, thereby better ensuring the vehicle's handling stability.

[0012] Preferably, step S1 constructs a dynamic model characterizing the steering system and the vehicle system, thereby establishing a human-vehicle system dynamic model, comprising the following steps:

[0013] Step S11: Establish the steering system dynamics equation:

[0014]

[0015] In the formula, J dam represents the inertia of the steering system, δ represents the steering wheel angle, is the steering wheel angular velocity, is the steering wheel angular acceleration, B inter represents the steering system damping, τ e represents the assist torque provided by the electric power steering system, τ h represents the hand torque input by the driver, τ tor It is composed of the equivalent friction torque of the steering system and the equivalent steering resistance generated by the lateral force of the tire. st is the steering system transmission ratio.

[0016] Step S12: The tire slip angle of the vehicle is set to be small. The vehicle dynamics model includes longitudinal motion, lateral motion and yaw motion. Based on the vehicle dynamics model and combined with step S11, the human-vehicle system dynamics model can be expressed as:

[0017]

[0018] In the formula, δ is the steering wheel angle, is the steering wheel angular velocity, β is the vehicle center of mass sideslip angle, is the vehicle yaw rate, M c represents the additional yaw torque generated by the wheel hub motor, w s =τ tor , A s is the vehicle state system matrix, B s1 is the steering control input system matrix, B s2 is the additional yaw moment control input system matrix, B sw is the system disturbance matrix.

[0019] Preferably, the step S2 uses a vertex polyhedron method to process the uncertainty in the system and establishes a linearized human-vehicle system state equation, including the following steps:

[0020] Step S21: Setting the upper and lower limits of the longitudinal speed of the distributed drive electric vehicle to and

[0021] Step S22: establishing a human-vehicle system model based on the TS fuzzy system;

[0022] The human-vehicle system model is reconstructed by using the linear combination of affine models, the human-vehicle system dynamics model is discretized, and the discrete vehicle dynamics model based on the polyhedron method is established as follows:

[0023]

[0024] In the formula,

[0025] Δx s (k+1) represents the discretized system state, Δx s (k) represents the vehicle state at time k, represents the system summation, Δτ e (k) is the auxiliary torque at time k, ΔM c (k) is the additional yaw moment at time k, w s (k) is the disturbance information at time k, where:

[0026] Preferably, step S3 establishes a distributed drive electric vehicle active steering and additional yaw moment control strategy based on terminal constraint cooperative game, comprising the following steps:

[0027] Step S31: Establishing the objective function of the active steering system agent and the additional yaw torque control system agent:

[0028] Before designing the dynamic game strategy, we first ignore the system disturbance and uncertainty and establish the dynamic expression of the control intention of the two agents:

[0029]

[0030] In the formula, the coefficient parameter matrix of the active steering agent and the additional yaw moment agent is: Δτ e,m =Δτ e , ΔM c,n =ΔM c , Δτ e,m (k) and ΔM c,n (k) is the change of assisting torque and yaw torque at time K after ignoring system disturbance and uncertainty;

[0031] Step S32: Establish the control objective function of the active steering agent and the additional yaw moment agent:

[0032]

[0033] In the formula, J ν (k) is the integrated control objective function of the active steering agent and the additional yaw moment agent, Δx s (t+k|k) is the active steering agent function, Δτ e (t+k|k) is the additional yaw moment agent function, N P is the prediction time domain, Δx s (N p +1+k|k) is the terminal cost function, and the weight coefficients are Q ν ,P ν and R ν , indicating the control priority of vehicle maneuverability and stability;

[0034] Considering the global performance index, the control objective function of the active steering agent and the additional yaw moment agent can be expressed as:

[0035]

[0036] In the formula, ||Δτ e (t+k|k)||≤Δτ e,max , J1(k) is the active steering agent control objective function;

[0037]

[0038] In the formula, ||ΔM C (t+k|k)||≤ΔM c,max , J2(k) is the additional yaw moment agent control objective function;

[0039] here and is the interaction level of the two agents. Such a design enables the agent to fully consider the control intentions of the adjacent agents, thus pursuing comprehensive performance;

[0040] Step S33: solving the active steering and additional yaw moment dynamic coordination control rate of the distributed drive electric vehicle based on cooperative game, which can be expressed as:

[0041]

[0042] In the formula, J m For each agent’s control goal,

[0043] , is the interaction model of the agents. By using the QP algorithm to solve, we can get the optimal solution for the two agents.

[0044] This design reflects the global optimization performance of the intelligent agent, executes the first element of the optimal result at each K+1 moment, and regards other elements as interactive model prediction information. The stability of the system is guaranteed by combining terminal constraints with terminal input and terminal cost function.

[0045] It should be noted that in the rolling optimization process of the cooperative game control algorithm, each agent only executes the first element of the optimal result at time k-1, and regards the other elements as the prediction information of the interaction model. Therefore, the terminal constraint combines the terminal input and the terminal cost function to ensure the stability of the system.

[0046] Preferably, step S4 aims at the problem that cooperative game is difficult to eliminate system disturbances, establishes a robust adaptive compensation control strategy, and realizes efficient and safe dynamic coordinated control of the distributed electric vehicle active steering system and the additional yaw moment control system, including the following contents:

[0047] In game theory, the uncertainty of the system is ignored, and a robust compensation controller is established to suppress system disturbances, thereby ensuring the stability of the system. The control system of each agent can be expressed as:

[0048]

[0049] In the formula, represents the robust compensation control input, Solve the matrix for the optimal control rate. Considering that the system states of the vehicle yaw rate, center of mass sideslip angle and steering wheel angle can be obtained through on-board sensors, a robust state feedback Then the closed-loop system can be rewritten as:

[0050]

[0051] In the formula, is the robust compensation coefficient matrix, C = I, y(k) and z(k) are the vehicle output and performance output of the system respectively, and the robust performance index is used to suppress the interference of the system on the performance output z(k)

[0052] ||z||2≤r c ||w||2 (Formula 12)

[0053] In the formula, r c represents the robust controller performance, by minimizing r c To reduce the impact of system disturbances,

[0054] Therefore, by combining dynamic cooperative game with robust compensation control, the control input of each agent can be obtained

[0055]

[0056] Preferably, establish a driver preview time t p And the linear parameter model with variable parameters of penalty factor κ:

[0057] The established human-vehicle-road state equation is:

[0058]

[0059] In the formula, the system state vector δ represents the vehicle steering wheel angle, is the steering wheel angular velocity, β is the vehicle center of mass sideslip angle, is the vehicle yaw rate, e yp is the road deviation of the preview point P, Δγ is the vehicle yaw angle deviation, u=τ e represents the steering torque input provided by AFS, w = [τ tor ρ], ρ represents the road curvature disturbance, z = [e yp Δγτ e ] T is the control output, A1 is the state matrix, including the time-varying road curvature parameters, B1 is the control input matrix, including the time-varying penalty factor parameters, H1 is the system disturbance matrix, including the time-varying road curvature, D1 is the output control input matrix, and D2 is the output system disturbance matrix.

[0060] Preferably, step S6 establishes a penalty factor adjustment method based on fuzzy rules and designs a gain-scheduled robust active steering control strategy to alleviate human-machine execution conflicts, thereby better ensuring the handling stability of the vehicle, including the following contents:

[0061] Step S61: Establish a penalty factor κ based on fuzzy rules,

[0062] Considering that the driver's steering torque and the lateral displacement deviation of the preview point can effectively determine the consistency of human-machine intentions, based on fuzzy theory, e yp and τ h 25 fuzzy rules are established to determine the penalty factor κ to alleviate human-machine conflict. The input of the fuzzy controller is e yp and τ h , the output is κ;

[0063] Step S62: Establishing a robust active steering control strategy.

[0064] In order to ensure the stability of the vehicle under human-machine collaborative control, a performance index function is established based on path tracking error, system stability and control output, which can be expressed as:

[0065]

[0066] In the formula, is the path tracking error, Δγ 2 is the vehicle yaw angle deviation, u 2 is the control output, ζ1, ζ2, ζ3 are the respective weight coefficients.

[0067] Solve by LMI, iterate and The solution X of the Lyapunov function and the feedback matrix gain K can be obtained i .

[0068] The above design fully considers the execution interaction between various intelligent agents and the needs of intelligent coordinated control of distributed electric vehicles. It can make real-time dynamic adjustments to the intelligent agents as needed, effectively ensuring the driving safety of the vehicle.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] The present invention establishes a dynamic coordinated control method for distributed drive electric vehicles for shared driving when coordinating the control of the distributed electric vehicle multi-execution system, establishes a human-vehicle dynamics model in combination with the vehicle state information collected in real time by the on-board sensor, linearizes the vehicle model based on the finite vertex polyhedron method, solves the vehicle nonlinearity problem caused by the time-varying longitudinal vehicle speed, designs a distributed parallel control method for active steering and additional yaw torque based on game theory, establishes a robust compensation control strategy to reduce the impact of system disturbances and uncertainties on system control, and designs a human-machine sharing control strategy based on dynamic intervention penalty factors to improve the tracking performance and driving performance of shared driving vehicles. This method breaks the auxiliary compensation control logic that places the auxiliary control system at the bottom layer, and chooses to place the human-machine sharing control at the control bottom layer logic, which effectively weakens the human-machine conflict and ensures the driving performance of human-machine shared driving. This design provides a new exploration direction for the modularization and extensibility requirements of the chassis integrated control of distributed drive electric vehicles.

[0071] In the distributed electric vehicle active steering and additional yaw moment coordinated control method based on cooperative game designed by the present invention, the active steering system agent and the additional yaw moment system agent can independently and controllably complete the corresponding command operations, wherein a single agent is the micro level of the multi-agent system, and has responsiveness, autonomy and flexibility; and the human-machine shared control strategy based on fuzzy rules, by real-time adjustment of the penalty factor to dynamically adjust the human-machine intervention level, has the ability to achieve higher flexibility, environmental adaptability and expandable comprehensive functions. This shared coordinated control idea is very suitable for the intelligent control of distributed electric vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 is a schematic diagram of the intelligent assisted steering system of the present invention;

[0073] Figure 2 is a diagram of the active steering system and the additional yaw moment coordinated control architecture of the present invention;

[0074] Figure 3 This is an architecture diagram of a distributed drive electric vehicle stability control method for shared driving of the present invention; DETAILED DESCRIPTION

[0075] The specific implementation of the present invention is described below in conjunction with the accompanying drawings so that those skilled in the art can better understand the present invention. It should be noted that in the description of the present invention, when the current known functions and actual detailed descriptions may dilute the main content of the present invention, these descriptions will be ignored here, and the distributed electric vehicle in this application refers to a distributed drive electric vehicle.

[0076] Figure 1The figure shows the principle diagram of the active steering and additional yaw moment steering system of this application. Figure 2 The figure shows the architecture diagram of the active steering system and the additional yaw moment collaborative control. The additional yaw moment control system (DYC) agent and the active steering control system (AFS) agent are considered as jointly working agents through interaction.

[0077] Figure 3 The figure shows the architecture of the distributed drive electric vehicle stability control method for shared driving. The architecture is divided into upper control and lower control layers. The upper layer uses the finite vertex polyhedron method to deal with the uncertainty of the system, establishes a distributed parallel control method of active steering and additional yaw torque based on game theory, and establishes a robust compensation control strategy; the lower layer adopts a human-machine shared control strategy based on a dynamic intervention penalty factor, establishes a robust linear parameter variation (LPV) controller, places human-machine shared control in the underlying control logic, weakens human-machine conflict, and ensures the driving performance of human-machine shared driving.

[0078] The present application provides a distributed drive electric vehicle stability control method for shared driving, comprising the following steps:

[0079] Step S1: construct a dynamic model that characterizes the steering system and the vehicle system, thereby establishing a human-vehicle system dynamic model, including the following steps:

[0080] See also Figure 2 It can be seen that intelligent driving is mainly controlled based on the steering angle of the electric power steering system. The driver cannot intervene in the steering control, which may easily lead to the driver losing control of the vehicle. Therefore, the required auxiliary steering torque is provided by the electric power steering device, so the steering system dynamic equation is established.

[0081] Step S11: Establish the steering system dynamics equation:

[0082]

[0083] In the formula, J dam represents the inertia of the steering system, δ represents the steering wheel angle, is the steering wheel angular velocity, is the steering wheel angular acceleration, B inter represents the steering system damping, τ e represents the assist torque provided by the electric power steering system, τ h represents the hand torque input by the driver, τ tor It is composed of the equivalent friction torque of the steering system and the equivalent steering resistance generated by the lateral force of the tire. st is the steering system transmission ratio.

[0084] Step S12: The tire slip angle of the vehicle is set to be small. The vehicle dynamics model includes longitudinal motion, lateral motion and yaw motion. Based on the vehicle dynamics model and combined with step S11, the human-vehicle system dynamics model can be expressed as:

[0085]

[0086] In the formula, δ is the steering wheel angle, is the steering wheel angular velocity, β is the vehicle center of mass sideslip angle, is the vehicle yaw rate, M c represents the additional yaw torque generated by the wheel hub motor, w s =τ tor , A s is the vehicle state system matrix, B s1 is the steering control input system matrix, B s2 is the additional yaw moment control input system matrix, B sw is the system disturbance matrix.

[0087] The additional yaw moment control system (DYC) and the active steering control system (AFS) are considered as co-operating agents. The AFS agent aims to track the reference yaw rate to improve the driver's handling performance, and the DYC agent ensures the stability of the vehicle by maintaining a small center of mass sideslip angle. At the same time, the two agents can interact to share control information, thereby achieving global performance optimization of vehicle control. The reference yaw rate and reference center of mass sideslip angle can be obtained through the steady-state dynamic response of the vehicle. Its reference state is expressed as in, Both represent the reference yaw rate of the vehicle, represent the ideal steering wheel angle and the ideal steering wheel angular speed respectively.

[0088] Distributed electric vehicle experimental data collection and experimental data processing:

[0089] (I) Experimental data collection. Using on-board sensors, the distributed electric vehicle execution end intelligent body collects experimental data in real time, including driver hand torque, steering control torque, steering wheel angle, steering wheel speed, wheel hub motor output torque, vehicle speed, center of mass sideslip angle, and yaw angular velocity;

[0090] (II) Experimental data processing. Since data is collected through different sensors, some of the collected data is not easy to understand and observe intuitively, so it is necessary to convert the data units, convert the steering wheel angle and speed from radians to degrees, and convert the speed from m / s to km / h. The data in each group of data is segmented, and each time period represents the manipulation behavior of the execution end within a period of time. For each data segment, an adaptive Kalman filter algorithm is used to remove the outliers in each data segment; when the prediction result is that the error increases, the Taylor series expansion algorithm under the current number of iterations is determined to be a divergent state, and the measurement value of the current state is regarded as a bad point and removed.

[0091] Step S2 uses the vertex polyhedron method to process the uncertainty in the system and establishes the state equation of the human-vehicle system after linearization, including the following steps:

[0092] Step S21: Setting the upper and lower limits of the longitudinal speed of the distributed drive electric vehicle to and

[0093] Step S22: establishing a human-vehicle system model based on the TS fuzzy system;

[0094] The human-vehicle system model is reconstructed by using the linear combination of affine models, the human-vehicle system dynamics model is discretized, and the discrete vehicle dynamics model based on the polyhedron method is established as follows:

[0095]

[0096] In the formula, Δx s (k+1) represents the discretized system state, Δx s (k) represents the vehicle state at time k, represents the system summation, Δτ e (k) is the auxiliary torque at time k, ΔM c (k) is the additional yaw moment at time k, w s (k) is the disturbance information at time k, where:

[0097] Step S3 establishes a distributed drive electric vehicle active steering and additional yaw moment control strategy based on terminal constraint cooperative game, including the following steps:

[0098] Step S31: Establishing the objective function of the active steering system agent and the additional yaw torque control system agent:

[0099] Before designing the dynamic game strategy, we first ignore the system disturbance and uncertainty and establish the dynamic expression of the control intention of the two agents:

[0100]

[0101] In the formula, the coefficient parameter matrix of the active steering agent and the additional yaw moment agent is: Δτ e,m =Δτ e , ΔM c,n =ΔM c , Δτ e,m (k) and ΔM c,n (k) is the change of assisting torque and yaw torque at time K after ignoring system disturbance and uncertainty.

[0102] Step S32: Establishing the control objective function of the active steering agent and the additional yaw moment agent:

[0103]

[0104] In the formula, J ν (k) is the integrated control objective function of the active steering agent and the additional yaw moment agent, Δx s (t+k|k) is the active steering agent function, Δτ e (t+k|k) is the additional yaw moment agent function, N P is the prediction time domain, Δx s (N p +1+k|k) is the terminal cost function, and the terminal cost function Δx is designed. s (N p +1+k|k) to ensure the stability of the system; the weight coefficients are Q ν ,P ν and R ν , indicating the control priority of vehicle handling and stability.

[0105] Considering the global performance index, the control objective function of the active steering agent and the additional yaw moment agent can be expressed as:

[0106]

[0107] In the formula, ||Δτ e (t+k|k)||≤Δτ e,max , J1(k) is the active steering agent control objective function.

[0108]

[0109] In the formula, ||ΔM C (t+k|k)||≤ΔM c,max , J2(k) is the additional yaw moment agent control objective function.

[0110] here and This design allows the agent to fully consider the control intentions of the adjacent agents and thus pursue comprehensive performance.

[0111] Step S33: solving the dynamic coordinated control rate of active steering and additional yaw moment of the distributed drive electric vehicle based on cooperative game,

[0112] See also Figure 2 It can be seen that the active steering system agent provides steering torque and additional yaw torque to ensure the vehicle's handling stability. The agent dynamic game process based on the reference model and the interaction model is established to solve the cooperative game control strategy and design a robust compensation controller to suppress the uncertainty of the system. Assume that represents the optimal input of the neighbors of agent m at time k-1. Considering that the control input of neighboring agents will also affect the system state, the interaction model between agents can be expressed as:

[0113]

[0114] It should be noted that in the rolling optimization process of the cooperative game control algorithm, each agent only executes the first element of the optimal result at time k-1, and regards the other elements as the prediction information of the interaction model. Therefore, the terminal constraint combines the terminal input and the terminal cost function to ensure the stability of the system.

[0115] The optimization problem can be expressed as:

[0116]

[0117] In the formula, J m For each agent’s control goal,

[0118] , is the interaction model of intelligent agents. By solving it using the QP algorithm, we can get the optimal solution for the two intelligent agents.

[0119] Step S4: In view of the problem that cooperative game is difficult to eliminate system disturbances, a robust adaptive compensation control strategy is established to achieve efficient and safe dynamic coordinated control of the distributed electric vehicle active steering system and the additional yaw moment control system, including the following contents:

[0120] In game theory, the uncertainty of the system is ignored, and a robust compensation controller is established to suppress system disturbances, thereby ensuring the stability of the system. The control system of each agent can be expressed as:

[0121]

[0122] In the formula, represents the robust compensation control input, Solve the matrix for the optimal control rate. Considering that the system states of vehicle yaw rate, center of mass sideslip angle and steering wheel angle can be obtained through on-board sensors, a robust state feedback is adopted. Then the closed-loop system can be rewritten as:

[0123]

[0124] In the formula, is the robust compensation coefficient matrix, C = I, y(k) and z(k) are the vehicle output and performance output of the system respectively, and the robust performance index is used to suppress the interference of the system on the performance output z(k)

[0125] ||z||2≤r c ||w||2 (Formula 12)

[0126] In the formula, r c represents the robust controller performance, by minimizing r c To reduce the impact of system disturbances,

[0127] Therefore, by combining dynamic cooperative game with robust compensation control, the control input of each agent can be obtained

[0128]

[0129] Step S5: Establishing the driver preview time t p And the linear parameter model with variable parameters of penalty factor κ:

[0130] The established human-vehicle-road state equation is:

[0131]

[0132] In the formula, the system state vector δ represents the vehicle steering wheel angle, For steering wheel

[0133] Angular velocity, β is the vehicle's center of mass sideslip angle, is the vehicle yaw rate, e yp is the road deviation of the preview point P, Δγ is the vehicle yaw angle deviation, u=τ erepresents the steering torque input provided by AFS, w = [τ tor ρ], ρ represents the road curvature disturbance, z = [e yp Δγτ e ] T is the control output, A1 is the state matrix, including the time-varying road curvature parameters, B1 is the control input matrix, including the time-varying penalty factor parameters, H1 is the system disturbance matrix, including the time-varying road curvature, D1 is the output control input matrix, and D2 is the output system disturbance matrix.

[0134] Step S6: Establish a penalty factor adjustment method based on fuzzy rules and design a gain-scheduled robust active steering control strategy to alleviate human-machine execution conflicts, thereby better ensuring the vehicle's handling stability, including the following contents:

[0135] Step S61: Establish a penalty factor κ based on fuzzy rules.

[0136] Considering that the driver's steering torque and the lateral displacement deviation of the preview point can effectively determine the consistency of human-machine intentions, based on fuzzy theory, e yp and τ h 25 fuzzy rules are established to determine the penalty factor κ to alleviate human-machine conflict. The input of the fuzzy controller is e yp and τ h , the output is κ. Set e yp ∈[-0.60.6], τ h ∈[-88],κ∈

[01] . yp and τ h The fuzzy subsets are {FB, FM, Z, ZM, ZB}, which represent e yp and τ h The five states of negative large, negative medium, zero, positive medium, and positive large. Let the fuzzy subset of κ be {Z, S, M, MB, B}, which represent the five states of penalty factor zero, small, medium, medium large, and large. The established fuzzy rules are shown in Table 5.2. The κ obtained by fuzzy rule calculation is shown. According to the fuzzy rules, the cooperation status between the driver and the distributed drive electric vehicle auxiliary driving system is divided into two categories:

[0137] State 1: Human-machine cooperation driving the vehicle together, when τ h and e yp When both are in the Z state, the human-machine collaboratively controls the vehicle;

[0138] State 2: Vehicle control dominated by intelligent assisted driving, when τ h and e yp When both are in the FB or ZB state, the auxiliary driving system plays a dominant role in the vehicle control process.

[0139] When human-machine cooperative driving is in state 1, it can be determined that the vehicle is in a normal mainline driving state. At this time, the penalty factor κ is in the S state, and only a small auxiliary control yaw torque is required to assist the vehicle in driving. When human-machine cooperative driving is in state 2, it can be determined that the vehicle is gradually deviating from the lane to the right or left, and the driver's operation will further aggravate the vehicle's deviation from the road. At this time, the penalty factor κ is in the B state in the hope of correcting the driver's behavior as much as possible and maintaining vehicle stability. When the lateral displacement of the vehicle is small, it is determined that the driver's behavior can correct the vehicle deviation. At this time, the penalty factor κ is reduced, thereby alleviating the conflict between the driver and the vehicle's auxiliary driving system and improving the driver's driving experience.

[0140] Step S62: Establishing a robust active steering control strategy.

[0141] In order to ensure the stability of the vehicle under human-machine collaborative control, a performance index function is established based on path tracking error, system stability and control output, which can be expressed as:

[0142]

[0143] In the formula, is the path tracking error, Δγ 2 is the vehicle yaw angle deviation, u 2 is the control output, ζ1, ζ2, ζ3 are weight coefficients.

[0144] Solve by LMI, iterate and The solution X of the Lyapunov function and the feedback matrix gain K can be obtained i . Complete dynamic coordination of multiple actuators of distributed electric vehicles for shared driving.

[0145] See also Figure 3It can be seen that this example designs a distributed drive electric vehicle stability control method for shared driving. First, a human-vehicle dynamics model is established, and the vehicle status information is collected in real time by combining on-board sensors. The finite vertex polyhedron method is used to deal with the uncertainty of the system. The active steering and additional yaw torque distributed parallel control method based on game theory is established. Then, a robust compensation control strategy is established to reduce the impact of system disturbance and uncertainty on system control. Finally, a human-machine sharing control strategy based on dynamic intervention penalty factor is designed to improve the tracking performance and driving performance of shared driving vehicles. This method breaks the auxiliary compensation control logic that places the auxiliary control system at the bottom layer, and chooses to place the human-machine sharing control at the control bottom logic, which effectively weakens the human-machine conflict and ensures the driving performance of human-machine shared driving. This design effectively avoids the problem of distributed electric vehicles being difficult to coordinate and control when the distributed drive electric vehicle execution system (driver system, steering system, drive system, braking system) is imperfect in interaction and each execution system is independent.

[0146] The advantages of this example are:

[0147] The dynamic coordinated control method of distributed drive electric vehicles for shared driving adopted by this method can effectively identify the execution intentions of each execution system and improve information interaction. It is of great significance to realize intelligent, efficient, safe and stable driving of distributed drive electric vehicles for shared driving. The coordinated control algorithm based on cooperative game has the advantages of intelligent agent interactive control coupling and convenient real-time control; the driver and active steering system shared control method based on penalty factor realizes efficient coordinated control of human-machine shared driving. The robust compensation control method is safer, more reliable and applicable to various working conditions than the traditional method, and can meet the dynamic coordinated control of distributed drive electric vehicles under human-machine shared driving.

[0148] The above description is only a preferred feasible implementation example of the present invention, and does not limit the scope of rights of the present invention. All equivalent structural changes made using the contents of the present invention and the drawings are included in the scope of rights of the present invention.

Claims

1. A distributed drive electric vehicle stability control method for shared driving, characterized in that: The following steps are involved: Step S1: according to the control requirements of the distributed drive electric vehicle, a steering system dynamics model and a vehicle dynamics model based on torque control are established, and the steering system dynamics model and the vehicle dynamics model are combined to establish a human-vehicle system dynamics model; Step S2: for the human-vehicle system dynamics model established in step S1, the vertex polyhedron method is used to deal with the uncertainty in the system, and the state equation of the human-vehicle system after linearization is established; Step S3: establishing a distributed drive electric vehicle active steering and additional yaw moment control strategy based on terminal constraint cooperative game; Step S4: In view of the problem that cooperative game is difficult to eliminate system disturbances, a robust adaptive compensation control strategy is established to achieve efficient and safe dynamic coordinated control of the distributed electric vehicle active steering system and the additional yaw moment control system; Step S5: Establishing the driver preview time t p and a linear parameter model with variable parameters of the penalty factor κ; Step S6: Establish a penalty factor κ adjustment method based on fuzzy rules and design a gain-scheduled robust active steering control strategy to alleviate human-machine execution conflicts, thereby better ensuring the vehicle's handling stability.

2. A distributed drive electric vehicle stability control method for shared driving according to claim 1, characterized in that: Step S1 constructs a dynamic model that characterizes the steering system and the vehicle system, thereby establishing a human-vehicle system dynamic model, including the following steps: Step S11: Establish the steering system dynamics equation: In the formula, J dam represents the inertia of the steering system, δ represents the steering wheel angle, is the steering wheel angular velocity, is the steering wheel angular acceleration, B inter represents the steering system damping, τ e represents the assist torque provided by the electric power steering system, τ h represents the hand torque input by the driver, τ tor It is composed of the equivalent friction torque of the steering system and the equivalent steering resistance generated by the lateral force of the tire. st is the steering system transmission ratio; Step S12: The tire slip angle of the vehicle is set to be small. The vehicle dynamics model includes longitudinal motion, lateral motion and yaw motion. Based on the vehicle dynamics model and combined with step S11, the human-vehicle system dynamics model can be expressed as: In the formula, δ is the steering wheel angle, is the steering wheel angular velocity, β is the vehicle center of mass sideslip angle, is the vehicle yaw rate, M c represents the additional yaw torque generated by the wheel hub motor, w s =τ tor , A s is the vehicle state system matrix, B s1 is the steering control input system matrix, B s2 is the additional yaw moment control input system matrix, B sw is the system disturbance matrix.

3. The method for stabilizing a distributed drive electric vehicle for shared driving according to claim 1, characterized in that: The step S2 uses the vertex polyhedron method to process the uncertainty in the system and establishes the state equation of the human-vehicle system after linearization, including the following steps: Step S21: Setting the upper and lower limits of the longitudinal speed of the distributed drive electric vehicle to and Step S22: establishing a human-vehicle system model based on the TS fuzzy system; The human-vehicle system model is reconstructed by using the linear combination of affine models, the human-vehicle system dynamics model is discretized, and the discrete vehicle dynamics model based on the polyhedron method is established as follows: In the formula, Δx s (k+1) represents the discretized system state, Δx s (k) represents the vehicle state at time k, represents the system summation, Δτ e (k) is the auxiliary torque at time k, ΔM c (k) is the additional yaw moment at time k, w s (k) is the disturbance information at time k, where:

4. The method for stabilizing a distributed drive electric vehicle for shared driving according to claim 3, characterized in that: Step S3 establishes a distributed drive electric vehicle active steering and additional yaw moment control strategy based on terminal constraint cooperative game, including the following steps: Step S31: Establishing the objective function of the active steering system agent and the additional yaw torque control system agent: Before designing the dynamic game strategy, we first ignore system disturbances and uncertainties and establish the dynamic expressions of the control intentions of the two agents: In the formula, the coefficient parameter matrix of the active steering agent and the additional yaw moment agent is: Δτ e,m =Δτ e , ΔM c,n =ΔM c , Δτ e,m (k) and ΔM c,n (k) is the change of assisting torque and yaw torque at time K after ignoring system disturbance and uncertainty; Step S32: Establish the control objective function of the active steering agent and the additional yaw moment agent: In the formula, J ν (k) is the integrated control objective function of the active steering agent and the additional yaw moment agent, Δx s (t+k|k) is the active steering agent function, Δτ e (t+k|k) is the additional yaw moment agent function, N P is the prediction time domain, Δx s (N p +1+k|k) is the terminal cost function, and the weight coefficients are Q ν ,P ν and R ν , indicating the control priority of vehicle maneuverability and stability; Considering the global performance index, the control objective function of the active steering agent and the additional yaw moment agent can be expressed as: In the formula, ‖Δτ e (t+k|k)‖≤Δτ e,max , J1(k) is the active steering agent control objective function; In the formula, ||ΔM C (t+k|k)||≤ΔM c,max , J2(k) is the additional yaw moment agent control objective function; and is the interaction level of the two agents; Step S33: solving the active steering and additional yaw moment dynamic coordination control rate of the distributed drive electric vehicle based on cooperative game, which can be expressed as: In the formula, J m For each agent’s control goal, is the interaction model of intelligent agents. By solving it using the QP algorithm, we can get the optimal solution for the two intelligent agents.

5. The method for stabilizing a distributed drive electric vehicle for shared driving according to claim 4, characterized in that: In step S4, in order to solve the problem that cooperative game is difficult to eliminate system disturbances, a robust adaptive compensation control strategy is established to realize efficient and safe dynamic coordinated control of the distributed electric vehicle active steering system and the additional yaw moment control system, including the following: In game theory, the uncertainty of the system is ignored, and a robust compensation controller is established to suppress system disturbances, thereby ensuring the stability of the system. The control system of each agent can be expressed as: In the formula, represents the robust compensation control input, Solve the matrix for the optimal control rate. Considering that the system states of vehicle yaw rate, center of mass sideslip angle and steering wheel angle can be obtained through on-board sensors, a robust state feedback is adopted. Then the closed-loop system can be rewritten as: In the formula, is the robust compensation coefficient matrix, C = I, y(k) and z(k) are the vehicle output and performance output of the system respectively, and the robust performance index is used to suppress the interference of the system on the performance output z(k) ||z||2≤r c ||w||2 (Formula 12) In the formula, r c represents the robust controller performance, by minimizing r c To reduce the impact of system disturbances, Therefore, by combining dynamic cooperative game with robust compensation control, the control input of each agent can be obtained 6. The method for stabilizing a distributed drive electric vehicle for shared driving according to claim 5, characterized in that: Establish the driver preview time t p And the linear parameter model with variable parameters of penalty factor κ: The established human-vehicle-road state equation is: In the formula, the system state vector δ represents the vehicle steering wheel angle, is the steering wheel angular velocity, β is the vehicle center of mass sideslip angle, is the vehicle yaw rate, e yp is the road deviation of the preview point P, Δγ is the vehicle yaw angle deviation, u=τ e represents the steering torque input provided by AFS, w = [τ tor ρ], ρ represents the road curvature disturbance, z = [e yp Δγτ e ] T is the control output, A1 is the state matrix, including the time-varying road curvature parameters, B1 is the control input matrix, including the time-varying penalty factor parameters, H1 is the system disturbance matrix, including the time-varying road curvature, D1 is the output control input matrix, and D2 is the output system disturbance matrix.

7. The method for stabilizing a distributed drive electric vehicle for shared driving according to claim 6, characterized in that: Step S6 establishes a penalty factor κ adjustment method based on fuzzy rules and designs a gain-scheduled robust active steering control strategy to alleviate human-machine execution conflicts, thereby better ensuring the vehicle's handling stability, including the following contents: Step S61: Establish a penalty factor κ based on fuzzy rules, Considering that the driver's steering torque and the lateral displacement deviation of the preview point can effectively determine the consistency of human-machine intentions, based on fuzzy theory, e yp and τ h 25 fuzzy rules are established to determine the penalty factor κ to alleviate human-machine conflict. The input of the fuzzy controller is e yp and τ h , the output is κ; Step S62: Establishing a robust active steering control strategy. In order to ensure the stability of the vehicle under human-machine collaborative control, a performance index function is established based on path tracking error, system stability and control output, which can be expressed as: In the formula, is the path tracking error, Δγ 2 is the vehicle yaw angle deviation, u 2 is the control output, ζ1, ζ2, ζ3 are the respective weight coefficients.

Images