Intelligent electric vehicle path tracking control method based on coevolution genetic algorithm
By decomposing the path tracking system of smart electric vehicles into input and output subsystems and zero dynamic subsystems, and using multi-agent adaptive sliding mode control and co-evolution genetic algorithm to optimize controller parameters, the intelligent electric vehicles are solved at high speed and large The problems of path tracking accuracy and dynamic stability during curvature driving achieve rapid convergence and efficient control of the system.
Patent Information
- Application Number
- CN202410971311.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-08
AI Technical Summary
When smart electric vehicles are driving at high speed and high curvature, the lateral nonlinear factors of the tire are enhanced, resulting in a decrease in path tracking control accuracy and is difficult to ensure dynamic stability under extreme operating conditions. There are few existing active steering and DYC coordinated control methods. Consider the trade-off between tire slip and computational efficiency.
By redefining the system output, the intelligent electric vehicle path tracking system with four-wheel independent driving and four-wheel steering is decomposed into an input and output subsystem and a zero-dynamic subsystem, and the adaptive sliding mode control method of the input and output subsystem of multiple agents is adopted. In addition, the controller parameters are optimized in combination with coevolutionary genetic algorithms to achieve rapid convergence and dynamic stability of the system.
It effectively improves path tracking accuracy and dynamic stability, simplifies controller design, improves the real-time and computing efficiency of the algorithm, and is suitable for a wide range of application scenarios.
Smart Images

Figure CN118915745A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent electric vehicle path tracking control, and mainly relates to a path tracking control method for a four-wheel steering and four-wheel independent drive intelligent electric vehicle based on a co-evolutionary genetic algorithm. Background Art
[0002] Path tracking control is a key technology for realizing the intelligentization of automobiles. It controls the size of the front wheel steering angle in real time through vehicle status information and pre-planned desired paths, aiming to minimize the lateral offset and heading error between the vehicle and the target path, and ensure that the vehicle travels along the expected path. At present, intelligent vehicles can achieve good path tracking effects under most working conditions. However, when the car is driving at high speed and large curvature, the lateral nonlinear factors of the tires are enhanced, and the path tracking accuracy under steering control will be greatly affected. The front wheel steering angle determined by the path tracking controller is difficult to meet the control requirements; on high-speed and low-adhesion roads, the car is prone to side slip, tail swing and other dangerous conditions when turning. Since there is only one control quantity, the front wheel steering angle, under extreme conditions, it is difficult for the steering controller to ensure the dynamic stability of the car while tracking the path.
[0003] The steering system directly generates lateral acceleration of the vehicle through the lateral force of the tire, which directly affects the lateral displacement, while active yaw moment control (DYC) indirectly affects the lateral response of the vehicle by affecting the yaw angular velocity. The coordinated control of DYC and active steering can effectively improve the dynamic performance of the vehicle in the process of path tracking. The driving / braking torque of the four wheels of a four-wheel independent drive electric vehicle is independently controllable, and the driving / braking torque control of a single wheel and the torque distribution control between axles and wheels can be accurately and quickly realized within the capacity of the motor, providing additional degrees of freedom for the lateral control of the vehicle, which can significantly improve the maneuverability and stability of the vehicle. The control methods commonly used for the coordination mechanism of active steering and DYC in four-wheel independent drive intelligent electric vehicles include: optimal control, sliding mode control, adaptive inversion control, model predictive control, etc. Reference 1 [Liang J, Lu Y, Yin G, et al. A Distributed Integrated Control Architecture of AFS and DYC based on MAS for Distributed Drive Electric Vehicles [J]. IEEE Transactions on Vehicular Technology, 2021, (99): 1-1.] regards the AFS system and the DYC system as multi-agents, and proposes a new distributed framework for integrated control of the AFS+DYC system based on the MAS idea. The collaborative control strategy of the two agents is obtained through the Pareto optimal theory to improve the lateral stability of the vehicle and reduce the amount of calculation for the driver during path tracking. Reference 2 [Hang P, Chen X, Luo F. LPV / H-infinity Controller Design for Path Tracking of Autonomous Ground Vehicles Through Four-Wheel Steering and Direct Yaw-Moment Control [J]. International Journal of Automotive Technology, 2019 (4): 20.] A four-wheel steering (4WS) and direct yaw moment control (DYC) system is used to design the path tracking control of an autonomous ground vehicle (AGV). The linear variable parameter (LPV) / H∞ control is used as the upper-level controller, and the weighted least squares (WLS) allocation algorithm is used at the lower level for torque distribution.Reference 3 [Mashadi, B., Ahmadizadeh, P., Majidi, M. et al. Integrated robust controller for vehicle path following. Multibody Syst Dyn 33, 207–228 (2015).] For the integrated 4WS+DYC control system, a robust controller is designed based on the μ synthesis method to realize the path tracking of AGV. This method has the powerful ability to enable the vehicle to track the required path in the presence of parameter uncertainty. The above active steering and DYC coordinated control schemes both utilize the redundant control freedom brought by the active yaw torque to improve the path tracking capability of the four-wheel drive smart electric vehicle while taking into account the dynamic stability. However, the above methods rarely consider tire slip, as well as the effective trade-off between computational efficiency and system dynamic performance during implementation.
[0004] Since multi-agent systems can achieve complex intelligence through the mutual cooperation of each agent, while reducing the complexity of system modeling, the robustness, reliability and flexibility of the system are improved. They have been widely used in the fields of aircraft formation, sensor network, data fusion, parallel computing, traffic vehicle control, etc. The main idea of sliding mode control is to design a dynamic, continuous or discrete control strategy so that the state of the system can slide on the preset sliding surface to achieve the desired performance. Therefore, the present invention redefines the system output, decomposes the intelligent electric vehicle path tracking system with four-wheel independent drive and four-wheel steering into an input-output subsystem and a zero dynamic subsystem, and proposes an adaptive sliding mode control method for the input-output subsystem based on multi-agents, so as to select the optimization function and constraint conditions with the goal of meeting the rapid convergence of the zero dynamic subsystem. On this basis, a controller parameter design method based on co-evolution genetic algorithm optimization is proposed to effectively balance the path tracking accuracy and dynamic stability. Summary of the invention
[0005] In order to improve the path tracking capability of intelligent electric vehicles with four-wheel independent drive and four-wheel steering and ensure their dynamic stability under extreme working conditions, the present invention aims to reduce the design difficulty of the entire system controller and improve the real-time performance of the algorithm. It weighs the performance indicators and computational efficiency of the system, adopts the method of redefining the system output, and proposes an intelligent electric vehicle path tracking control method based on a coevolutionary genetic algorithm. While ensuring good tracking control effects under different working conditions, the dynamic stability of the system is improved.
[0006] The technical solution adopted by the present invention to solve the technical problem is as follows:
[0007] A path tracking control method for an intelligent electric vehicle based on a co-evolutionary genetic algorithm comprises the following steps:
[0008] Step 1: Based on the kinematics and dynamics of the vehicle, a path tracking model of the four-wheel steering and four-wheel independent drive smart electric vehicle is established;
[0009] Step 2: Based on the path tracking model obtained in step 1, redefine the output of the system, and decompose the path tracking model into an input-output subsystem and a zero dynamic subsystem by input-output linearization;
[0010] Step 3: Consider the 6 degrees of freedom equations in the input-output subsystem as 6 intelligent agents. According to the internal working communication of the 6 intelligent agents, determine the topological structure of the 6 intelligent agents in the input-output subsystem. Based on the graph theory method, obtain the adjacency matrix A, in-degree matrix D, Laplace matrix L and correlation matrix B used for information exchange between the 6 intelligent agents.
[0011] Step 4: Use the CarSim car model to obtain the real-time parameters of the vehicle: Y-axis coordinates, X-axis coordinates, heading angle, lateral speed and acceleration, longitudinal speed and acceleration, heading angular velocity, rolling angular velocity of each wheel, vertical force of each wheel, etc.
[0012] Step 5: According to the Y-axis coordinates, X-axis coordinates, heading angle, lateral velocity and acceleration, longitudinal velocity and acceleration, heading angular velocity, rolling angular velocity of each wheel, vertical force of each wheel, etc. output by the CarSim car model in step 4, with the goal of rapid convergence of the zero dynamic subsystem state obtained in step 2, the controller design parameters are obtained through a controller parameter design module based on co-evolution genetic algorithm optimization;
[0013] Step 6. Based on the given reference trajectory, the input-output subsystem obtained in step 2, the adjacency matrix, in-degree matrix, Laplace matrix and correlation matrix for information exchange between the six agents obtained in step 3, the Y-axis coordinates, X-axis coordinates, heading angle, lateral velocity and acceleration, longitudinal velocity and acceleration, heading angular velocity, rolling angular velocity of each wheel, vertical force of each wheel, etc. output by the CarSim car model in step 4, and the controller parameters obtained in step 5, the four-wheel steering angle and four-wheel drive torque are obtained through the adaptive sliding mode control module of the input-output subsystem based on the multi-agent, and they are input into the CarSim car model.
[0014] The beneficial effects of the present invention are as follows:
[0015] 1) According to the dynamic motion mechanism of the vehicle, the present invention establishes a mathematical model of the path tracking system, considers the slip state of the four wheels, trajectory tracking error and yaw stability, redefines the output of the system, and decomposes the path tracking system into an input-output subsystem and a zero dynamic subsystem through input-output linearization, thereby achieving model dimensionality reduction and control input decoupling.
[0016] 2) The present invention regards the 6-degree-of-freedom equations in the input-output subsystem as 6 intelligent agents, establishes a dimensionality reduction control model of the input-output subsystem based on graph theory, and simplifies the model dimension.
[0017] 3) The present invention proposes an adaptive sliding mode control strategy based on multi-agent for the input-output subsystem, so that the state of the input-output subsystem: the slip state of each wheel Δω σ , the lateral motion of the vehicle Δγ and Δv y It converges to zero in a finite time, preventing excessive wheel slip while controlling the vehicle's handling stability, effectively balancing the system's performance indicators and the computing power requirements of the algorithm in engineering applications.
[0018] 4) The present invention aims to quickly stabilize the heading angle following error and lateral displacement following error of the zero dynamic subsystem near the equilibrium point, and proposes a controller design parameter selection method based on a co-evolutionary genetic algorithm, so as to ensure the asymptotic stability of the trajectory tracking control system of the intelligent electric vehicle.
[0019] 5) The method proposed in the present invention is simple, easy to implement, and has a wide range of applications, and is suitable for widespread promotion and application. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 It is a schematic diagram of the topological structure of the input and output subsystem based on graph theory of the present invention.
[0021] Figure 2 It is a principle block diagram of the intelligent electric vehicle path tracking control method based on the co-evolutionary genetic algorithm of the present invention.
[0022] Figure 3 It is a principle block diagram of controller parameter design based on co-evolutionary genetic algorithm optimization of the present invention. DETAILED DESCRIPTION
[0023] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0024] like Figure 1 As shown, in the intelligent electric vehicle path tracking control method based on the coevolutionary genetic algorithm of the present invention, the six degree-of-freedom equations in the input-output subsystem are regarded as six intelligent agents, and according to the communication topology of the six subsystems, a topological structure diagram of the six intelligent agents of the input-output subsystem based on graph theory is constructed.
[0025] like Figure 2As shown, the intelligent electric vehicle path tracking control method based on the co-evolutionary genetic algorithm of the present invention includes: step 1, according to the kinematic and dynamic mechanism of the vehicle, a path tracking model of a four-wheel steering and four-wheel independent drive intelligent electric vehicle is established; step 2, according to the path tracking model obtained in step 1, the output of the system is redefined, and the path tracking model is decomposed into an input-output subsystem and a zero dynamic subsystem through input-output linearization; step 3, the 6 degrees of freedom equations in the input-output subsystem are regarded as 6 intelligent agents, and the topological structure of the 6 intelligent agents in the input-output subsystem is determined according to the internal working communication of the 6 intelligent agents, and the adjacency matrix A, the in-degree matrix D, the Laplace matrix L and the correlation matrix B for information exchange between the 6 intelligent agents are obtained based on the graph theory method; step 4, the real-time parameters of the vehicle are obtained using the CarSim car model: the coordinates under the Y axis, the coordinates under the X axis, the heading angle, the lateral speed and acceleration, the longitudinal speed and acceleration, the heading angular velocity, the rolling angular velocity of each wheel, the vertical direction of each wheel force, etc.; step 5. According to the Y-axis coordinate, X-axis coordinate, heading angle, lateral velocity and acceleration, longitudinal velocity and acceleration, heading angular velocity, rolling angular velocity of each wheel, vertical force of each wheel, etc. output by the CarSim car model in step 4, with the goal of rapid convergence of the zero dynamic subsystem state obtained in step 2, the controller design parameters are obtained through the controller parameter design module based on co-evolutionary genetic algorithm optimization; step 6. According to the given reference trajectory, the input and output subsystem obtained in step 2, the adjacency matrix, in-degree matrix, Laplace matrix and correlation matrix for information exchange between the 6 intelligent agents obtained in step 3, the Y-axis coordinate, X-axis coordinate, heading angle, lateral velocity and acceleration, longitudinal velocity and acceleration, heading angular velocity, rolling angular velocity of each wheel, vertical force of each wheel, etc. output by the CarSim car model in step 4, and the controller parameters obtained in step 5, the four-wheel steering angle and four-wheel drive torque are obtained through the adaptive sliding mode control module of the input and output subsystem based on multi-agents.
[0026] like Figure 3 As shown, in the intelligent electric vehicle path tracking control method based on the coevolutionary genetic algorithm of the present invention, the controller parameter design process based on the coevolutionary genetic algorithm optimization includes: (1) population and parameter initialization; (2) termination criterion 1 judgment; (3) termination criterion 2 judgment; (4) genetic evolution operation of controller parameters; (5) genetic evolution operation of penalty factor.
[0027] The present invention is based on the intelligent electric vehicle path tracking control method of the co-evolutionary genetic algorithm, and the specific implementation steps are as follows:
[0028] Step 1: Based on the vehicle's dynamics mechanism, a path tracking model for a four-wheel drive smart electric vehicle is established. The specific implementation is as follows:
[0029] 1.1 Vehicle dynamics model
[0030] In the design of the controller, the handling dynamics model of the four-wheel independent drive smart electric vehicle can be described as:
[0031]
[0032] Among them, v x and v y are the longitudinal and lateral velocities of the vehicle, respectively. r is the yaw rate. m and I z is the vehicle mass and yaw inertia. F yf and F yr is the lateral force of the front and rear tires. f and δ r is the steering angle of the front and rear wheels. f and l r ΔM is the distance from the center of mass of the vehicle to the front axle and the rear axle respectively. z The additional yaw moment generated by the longitudinal force difference between the left and right tires is:
[0033]
[0034] Among them, T qσ is the driving torque of each wheel, σ=fl, fr, rl, rr, representing the left front wheel, right front wheel, left rear wheel, right rear wheel, t wσ Satisfy t wfl =t wfr =t wf , t wrl =t wrr =t wr , t wf ,t wr are the wheelbases of the front and rear wheels of the vehicle, R e is the wheel rolling radius.
[0035] When the lateral acceleration is less than 0.4g (g is the acceleration of gravity), the slip angle does not exceed 4° to 5°. The relationship between the tire lateral force and the slip angle can be expressed as a linear relationship, which can be written as:
[0036] F yf =-k f α f , F yr =-k r α r (4)
[0037] Among them, k f and k r are the cornering stiffness of the front and rear tires respectively. f and α r is the front and rear tire slip angle, defined as:
[0038]
[0039] Assuming that the steering angles of the front and rear wheels are small, cosδ in equations (1) and (2) is f ≈cosδ r ≈1. Substituting equations (4) and (5), the vehicle model described by equations (1) and (2) can be simplified to the following linear dynamic equation:
[0040]
[0041] Wheel dynamics equation:
[0042]
[0043] In the formula, ω σ is the rolling angular velocity of each wheel, I ω is the wheel moment of inertia, F zσ The vertical force on each wheel during the vehicle's roll and pitch motion is not considered, and μ is the road friction coefficient.
[0044]
[0045] In the formula, m ω is the sprung mass, and h is the height of the sprung center of mass.
[0046] The desired wheel rolling angular velocity ω is calculated from the wheel center velocity dσ :
[0047]
[0048] Where η wfl =η wrl =-1,η wfr =η wrr =1.
[0049] Desired wheel rolling angular velocity ω dσ Derivatives:
[0050]
[0051] In order to control the tire slip rate, the desired wheel angular velocity is set to the angular velocity of the wheel when it is in pure rolling, so the desired tire slip state is Δω σ =ω σ -ω dσ =0, indicating that the tire has no slip.
[0052]
[0053] 1.2 Path tracking model
[0054] In order to enable the vehicle to accurately track the target path, a path tracking control strategy is designed to reduce the heading angle error The lateral position error Δy is minimized. Δy is the lateral deviation between the center of mass of the vehicle and the closest point d on the reference path, R is the arc length along the reference path, and are the desired heading angle on the reference path and the actual heading angle of the car, respectively.
[0055] Actual heading angle and target heading angle The difference is the heading angle error, which is
[0056]
[0057] Based on the Serret-Frenet coordinate system and linearized by the small angle assumption, we can obtain:
[0058]
[0059] Where ρ is the curvature corresponding to the reference path point.
[0060] The lateral position error satisfies the following equation:
[0061]
[0062] Under ideal steady-state conditions, Therefore, according to equations (13) and (14), γ d =v x ρ; v yd = 0, let Δγ = γ - γ d , Δv y =v y -v yd is the desired yaw rate γ d Tracking deviation, expected lateral velocity v yd tracking deviation.
[0063]
[0064]
[0065] make According to equations (11), (13)-(16), the path tracking model is:
[0066]
[0067] u z =[δ β ,δγ ,T fl ,T fr ,T rl ,T rr ] T ,
[0068]
[0069]
[0070]
[0071] Step 2: Based on the path tracking model obtained in step 1, redefine the output of the system, and decompose the path tracking model into an input-output subsystem and a zero dynamic subsystem through input-output linearization. The specific implementation is as follows:
[0072] In order to control the wheel slip rate while controlling the vehicle's handling stability and prevent excessive wheel slip, the wheel slip state Δω is calculated. σ , the lateral motion of the vehicle Δγ and Δv y As a state variable, we control z1 =Δv y ,x z2 =Δγ,x z3 =Δω fl ,x z4 =Δω fr ,x z5 =Δω rl ,x z6 =Δω rr , considering the influence of the slip state of the four wheels on the trajectory tracking error and yaw stability, the output of the system is redefined as:
[0073]
[0074] In the formula, the design parameter k i1 , k i2 , k i3 is a constant to be determined, i = 1, 2, ..., 6. By selecting appropriate design parameters k i1 , k i2 , k i3 , which can ensure the internal dynamic stability of the system.
[0075] By taking the derivative of equation (18), according to the path tracking model (17), the heading angle error (12) and the lateral position error (13), the input-output subsystem can be obtained:
[0076]
[0077] In the formula,
[0078]
[0079] K z3 =[k 12 ,k 22 ,k 32 ,k 42 ,k 52 ,k 62 ] T
[0080]
[0081] Since the trajectory tracking model (17) is 8-dimensional and the input-output subsystem (19) is only 6-dimensional, the remaining 2-dimensional system states, namely the heading angle following error (13) and the lateral displacement following error (14), constitute the internal subsystem of the trajectory tracking control system. It can be seen that by adopting the method of redefining the system output (18), the path tracking control system (17) is decomposed into the input-output subsystem (19) and the internal subsystems (13)-(14).
[0082] When a specific control input u z , so that when the output z of the input-output subsystem (19) is zero, the internal subsystems (13) and (14) are zero dynamic subsystems. Ensuring the local stability of the zero dynamic subsystem can ensure the local asymptotic stability of the entire closed-loop system. Therefore, the present invention proposes an adaptive sliding mode control strategy based on multi-agents for the input-output subsystem, so that the input-output subsystem converges to zero in a finite time; for the internal subsystem, when a specific control input u z When the output of the input-output subsystem is zero, the internal subsystem is the zero dynamic subsystem. By selecting reasonable controller design parameters, the zero dynamic subsystem is asymptotically stable near the equilibrium point, thereby ensuring the asymptotic stability of the intelligent electric vehicle trajectory tracking control system.
[0083] Step 3: Consider the 6 degrees of freedom equations in the input-output subsystem as 6 intelligent agents, determine the 6-agent communication topology of the input-output subsystem according to the internal working communication of the 6 intelligent agents, and use graph theory methods to obtain the adjacency matrix A, in-degree matrix D, Laplace matrix L and correlation matrix B used for information exchange between the 6 intelligent agents. The specific implementation method is as follows:
[0084] Let K B =K z1 B z , K A =K z1 A z +K z2 According to formula (19), we can know that K z1 Bz is a diagonal matrix. Rewrite the input-output subsystem (19) as follows:
[0085]
[0086] In the formula, K Aii Denotes the matrix K A The element in the i-th row and i-th column, K Aij Denotes the matrix K A The element in row i and column j, K Bii Denotes the matrix K B The element at the i-th row and i-th column.
[0087] Drawing on the theory of multi-agents, each degree of freedom equation in the input-output subsystem (20) is regarded as an agent. For the i-th agent (20), a multi-agent-based adaptive sliding mode control strategy is proposed. Through the network interconnection of 6 agents, sharing resources and mutual communication, the task of the entire input-output subsystem is completed together. From the comparison of equations (19) and (20), it can be seen that the 6-dimensional input-output subsystem is organized into 6 1-dimensional agents that collaborate to complete the task, achieving model dimensionality reduction.
[0088] According to graph theory, any multi-agent system can be represented as G = (V, E, A), where V = {1, 2, ..., 6} represents a set of nodes. represents the edge set, A=[a ij ] is the adjacency matrix of G, a ij Used to describe the relationship between nodes and edges. If (i, j) ∈ E then a ij =1, otherwise a ij = 0. The correlation matrix B = diag(b1, b2, ..., b6), b i >0 means that the i-th agent can receive the communication information sent by the leader; b i =0 means that the i-th agent cannot receive the communication information sent by the leader. The in-degree matrix D = diag(d1, d2, ..., d6), Laplace matrix L = DA, L = [l ij ], each element is defined as:
[0089]
[0090] According to the communication topology of the six agents (20), the topological structure diagram of each agent in the input and output subsystem can be obtained, as shown in Figure 1 shown. Figure 1In the equation, agents 1-6 are the 6 degrees of freedom equations of the input and output subsystems respectively. Based on graph theory, the adjacency matrix A, in-degree matrix D, Laplace matrix L and correlation matrix B between the 6 agents can be obtained:
[0091]
[0092] Step 4: Use the CarSim car model to obtain the real-time parameters of the vehicle: Y-axis coordinates, X-axis coordinates, heading angle, lateral speed and acceleration, longitudinal speed and acceleration, heading angular velocity, rolling angular velocity of each wheel, vertical force of each wheel, etc.
[0093] Step 5: Based on the Y-axis coordinates, X-axis coordinates, heading angle, lateral velocity and acceleration, longitudinal velocity and acceleration, heading angular velocity, rolling angular velocity of each wheel, vertical force of each wheel, etc. output by the CarSim car model, with the goal of rapid convergence of the zero dynamic subsystem state obtained in step 2, the controller design parameters are obtained through the controller parameter design module based on co-evolutionary genetic algorithm optimization, and sent to the input-output subsystem adaptive sliding mode control module based on multi-agent. The specific implementation method is as follows:
[0094] The 2D state variables of the internal subsystem (heading angle following error The lateral displacement following error Δy) is derivatized twice so that the control u z Appear:
[0095]
[0096]
[0097] General Substituting (22) and (23), we get
[0098]
[0099] In the formula,
[0100]
[0101]
[0102] From the input and output subsystem, we can see that when the system state converges to zero in a finite time, that is, When u z =-(K z1 B z ) -1 [(K z1 A z +K z2 )x z +D z], and bring it into the internal subsystem (24), we get the zero dynamic subsystem:
[0103]
[0104] make:
[0105]
[0106] The following equation can be obtained:
[0107]
[0108] Substituting equation (26) into (25), we can obtain
[0109]
[0110] From formula (27), we can see that if [A ο -B ο (K z1 B z ) -1 K N2 K N3 ] is strictly guaranteed to be in the left half plane of the complex plane, then the equilibrium point of the original nonlinear system is asymptotically stable. Therefore, we select a suitable parameter k to redefine the system output. i1 , k i2 , k i3 , i = 1, 2, ..., 6, so that the matrix [A ο -B ο (K z1 B z ) -1 K N2 K N3 ] are all negative, then the zero dynamic subsystem (27) will be at the equilibrium point The path tracking control system is asymptotically stable.
[0111] In order to better realize the rapid stabilization of the path tracking control system, the present invention adopts a genetic algorithm based on co-evolution to adjust the parameter k i1 , k i2 , k i3 , i=1,2,…,6 for optimization design, select the characterization design parameter k i1 , k i2 , k i3 There are two types of populations with optimal solutions and penalty factors. The first type of population G_1 contains Size1 individuals, each of which is determined by the parameter k i1 , k i2 , k i3 Represents the genetic composition, and uses genetic algorithm to i1 , ki2 , k i3 The second population G_2 contains Size2 individuals, each of which is composed of genes represented by the penalty factor in the penalty function. The penalty factor is optimized using a genetic algorithm. Through the genetic algorithm under the coevolution model, the penalty factor is continuously adjusted, and finally k that meets the constraints is obtained. i1 , k i2 , k i3 The optimal solution. The specific implementation steps are as follows:
[0112] (1) Population and parameter initialization
[0113] Initialize populations G_1 and G_2, set the evolutionary generations of populations G_1 and G_2 to l1 and l2 respectively, and the maximum evolutionary generations to l1max and l1max respectively, set l1=0, l2=0, and generate the initial value of the penalty factor randomly.
[0114] (2) Termination Criteria 1
[0115] If l1=l 1max , then the iteration ends and the optimized controller parameter k is output i1 , k i2 , k i3 ; Otherwise, go to step (3).
[0116] (3) Termination Criteria 2
[0117] If l2=l 2max , the iteration ends and the optimized penalty factor is output to step (4); otherwise, jump to step (5).
[0118] (4) Genetic evolution operation of controller parameters
[0119] Select objective function: Use the quadratic performance index related to the heading control error and its rate of change as the optimization objective function:
[0120]
[0121] Where Θ∈R 4×4 is a positive symmetric matrix.
[0122] Constraints: Each individual in the population must satisfy [A ο -B ο (K z1 B z ) -1 K N2 K N3 ](27)The real part of the eigenvalue is less than zero:
[0123] I Γ =Re[λ Γ (Ao )]<0 (29)
[0124] Where Γ is an integer, 1≤Γ≤4. Γ (A o ) represents the matrix A o The Γth eigenvalue of .
[0125] According to the optimization objective function (28) and the inequality constraint (29), the following penalty function (30) is selected as the fitness function:
[0126]
[0127] In the formula, β J (l1) is the penalty factor, l1 is the parameter k i1 , k i2 , k i3 The evolutionary algebra of .
[0128] According to the penalty factor obtained, the fitness values of all individuals in the G_1 population are calculated using formula (30), and then the selection, crossover and mutation operators of the genetic algorithm are used to perform genetic evolution operations on the population G_1 to generate a new generation of G_1 population, setting l1=l1+1, l2=0, and returning to step (2).
[0129] (5) Genetic evolution operation of penalty factor
[0130] According to equations (29) and (30), the evaluation function of G_2 is selected as:
[0131]
[0132] In the formula, ∑F feasible Represents the sum of fitness values of all feasible solutions, num feasible represents the number of feasible solutions, max(F Ξ ) represents the maximum fitness value of individuals in the population, Ξ is an integer, 0<Ξ<size2, size2 is the population size of G_2, ∑sum viol The total number of individuals violating constraint (29), ∑mum viol The total number of individuals violating constraint (29).
[0133] Using the first individual G_2(1) of population G_2 to the Size2th individual G_2(Size2) of population G_2 as penalty factors, the G_1 population is evolved using a genetic algorithm. The fitness function value of the evolved new population is calculated and sorted according to the fitness function value to generate a new generation of population G_2. Let l2=l2+1 and return to step (3).
[0134] Step 6: Based on the given reference trajectory, the input and output subsystem obtained in step 2, the adjacency matrix, in-degree matrix, Laplace matrix and correlation matrix for information exchange between the six agents obtained in step 3, the Y-axis coordinates, X-axis coordinates, heading angle, lateral velocity and acceleration, longitudinal velocity and acceleration, heading angular velocity, rolling angular velocity of each wheel, vertical force of each wheel, etc. output by the CarSim car model in step 4, and the controller parameters obtained in step 5, the four-wheel steering angle and four-wheel drive torque are obtained through the adaptive sliding mode control module of the input and output subsystem based on multi-agents, and input them into the CarSim car model. The specific implementation is as follows:
[0135] According to the basic knowledge of multi-agent graph theory and the topological structure diagram of each agent in the input and output subsystem, the deviation equation of the i-th agent following the leader is defined as follows:
[0136]
[0137] Where z0 is the state variable of the leader, satisfying z0=0,l ij is the element of the Laplace matrix L.
[0138] Let τ i =∫υ i dt, for the i-th system agent of the input-output subsystem, the following sliding surface is introduced:
[0139]
[0140] Where c1, c2>0, 0<λ τ <1.
[0141] Let β si is the sliding mode switching coefficient, β si satisfy:
[0142]
[0143] When the state parameters of other agents are unknown due to communication or other reasons, the parameter β si Unknown. β si The estimated value of , for the i-th agent (20) of the input-output subsystem, the following control strategy (35) and Adaptation rate (36):
[0144]
[0145] In the formula, β ui >0.
[0146] The adaptive rate is designed as follows:
[0147]
[0148] In the formula, ρ βi is the design parameter, and ρ βi >0.
Claims
1. A path tracking control method for intelligent electric vehicles based on co-evolutionary genetic algorithm, characterized in that: The method comprises the following steps: Step 1: Based on the kinematics and dynamics of the vehicle, a path tracking model of the four-wheel steering and four-wheel independent drive smart electric vehicle is established; Step 2: Based on the path tracking model obtained in step 1, redefine the output of the system, and decompose the path tracking model into an input-output subsystem and a zero dynamic subsystem by input-output linearization; Step 3: Consider the 6 degrees of freedom equations in the input-output subsystem as 6 intelligent agents. According to the internal working communication of the 6 intelligent agents, determine the topological structure of the 6 intelligent agents in the input-output subsystem. Based on the graph theory method, obtain the adjacency matrix A, in-degree matrix D, Laplace matrix L and correlation matrix B used for information exchange between the 6 intelligent agents. Step 4: Use the CarSim car model to obtain the real-time parameters of the vehicle: Y-axis coordinates, X-axis coordinates, heading angle, lateral speed and acceleration, longitudinal speed and acceleration, heading angular velocity, rolling angular velocity of each wheel, and vertical force of each wheel; Step 5: According to the Y-axis coordinate, X-axis coordinate, heading angle, lateral velocity and acceleration, longitudinal velocity and acceleration, heading angular velocity, rolling angular velocity of each wheel, and vertical force of each wheel output by the CarSim car model in step 4, with the goal of fast convergence of the zero dynamic subsystem state obtained in step 2, the controller design parameters are obtained through a controller parameter design module based on co-evolution genetic algorithm optimization; Step 6. Based on the given reference trajectory, the input-output subsystem obtained in step 2, the adjacency matrix, in-degree matrix, Laplace matrix and correlation matrix for information exchange between the six agents obtained in step 3, the Y-axis coordinates, X-axis coordinates, heading angle, lateral velocity and acceleration, longitudinal velocity and acceleration, heading angular velocity, rolling angular velocity of each wheel, and vertical force of each wheel output by the CarSim car model in step 4, and the controller parameters obtained in step 5, the four-wheel steering angle and four-wheel drive torque are obtained through the adaptive sliding mode control module of the input-output subsystem based on the multi-agent, and they are input into the CarSim car model.
2. The intelligent electric vehicle path tracking control method based on coevolutionary genetic algorithm as claimed in claim 1, characterized in that: The implementation process of the path tracking model of the four-wheel steering and four-wheel independent drive intelligent electric vehicle described in step 1 is as follows: 1.1 Vehicle dynamics model In the design of the controller, the handling dynamics model of the four-wheel independent drive smart electric vehicle can be described as: Among them, v x and v y are the longitudinal and lateral velocities of the vehicle, respectively; r is the yaw rate; m and I z is the vehicle mass and yaw inertia; F yf and F yr is the lateral force of the front and rear tires; δ f and δ r is the front and rear wheel steering angle; l f and l r are the distances from the center of mass of the vehicle to the front axle and the rear axle respectively; ΔM z The additional yaw moment generated by the longitudinal force difference between the left and right tires is: Among them, T qσ is the driving torque of each wheel, σ=fl, fr, rl, rr, representing the left front wheel, right front wheel, left rear wheel, right rear wheel, t wσ Satisfy t wfl =t wfr =t wf , t wrl =t wrr =t wr , t wf ,t wr are the wheelbases of the front and rear wheels of the vehicle, R e is the wheel rolling radius; When the lateral acceleration is less than 0.4g (g is the acceleration of gravity), the slip angle does not exceed 4° to 5°. The relationship between the tire lateral force and the slip angle can be expressed as a linear relationship, written as: F yf =-k f a f ,F yr =-k r a r (4) Among them, k f and k r are the cornering stiffness of the front and rear tires respectively; α f and α r is the front and rear tire slip angle, defined as: Assuming that the steering angles of the front and rear wheels are small, cosδ in equations (1) and (2) is f ≈cosδ r ≈1; Substituting equations (4) and (5), the vehicle model described by equations (1) and (2) can be simplified to the following linear dynamic equation: Wheel dynamics equation: In the formula, ω σ is the rolling angular velocity of each wheel, I ω is the wheel moment of inertia, F zσ is the vertical force of each wheel when the vehicle rolls and pitches, μ is the road friction coefficient; In the formula, m ω is the sprung mass, h is the height of the sprung mass center; The desired wheel rolling angular velocity ω is calculated from the wheel center velocity dσ : where η wfl = η wrl = -1, η wfr = η wrr = 1; Desired wheel rolling angular velocity ω dσ Derivatives: In order to control the tire slip rate, the desired wheel angular velocity is set to the angular velocity of the wheel when it is in pure rolling, so the desired tire slip state is Δω σ =ω σ -ω dσ =0, indicating that the tire has no slip; 1.2 Path tracking model In order to enable the vehicle to accurately track the target path, a path tracking control strategy is designed to reduce the heading angle error The lateral position error Δy is minimized; Δy is the lateral deviation between the center of mass of the vehicle and the nearest point d on the reference path, R is the arc length along the reference path, and They are the desired heading angle on the reference path and the actual heading angle of the car; Actual heading angle and target heading angle The difference is the heading angle error, which is Based on the Serret-Frenet coordinate system and linearized by the small angle assumption, we can obtain: Where ρ is the curvature corresponding to the reference path point; The lateral position error satisfies the following equation: Under ideal steady-state conditions, Therefore, according to equations (13) and (14), γ d =v x ρ; v yd = 0, let Δγ = γ - γ d , Δv y =v y -v yd is the desired yaw rate γ d Tracking deviation, expected lateral velocity v yd tracking deviation; make The path tracking model is obtained from equations (11), (13)-(16): u z =[δ β ,δ γ ,T fl ,T fr ,T rl ,T rr ] T , 3. The intelligent electric vehicle path tracking control method based on coevolutionary genetic algorithm as claimed in claim 1, characterized in that: The implementation process of the input-output subsystem and zero dynamic subsystem model in step 2 is as follows: In order to control the wheel slip rate while controlling the vehicle's handling stability and prevent excessive wheel slip, the wheel slip state Δω is calculated. σ , the lateral motion of the vehicle Δγ and Δv y As a state variable, control; let x z1 =Δv y ,x z2 =Δγ,x z3 =Δω fl ,x z4 =Δω fr ,x z5 =Δω rl ,x z6 =Δω rr , considering the influence of the slip state of the four wheels on the trajectory tracking error and yaw stability, the output of the system is redefined as: In the formula, the design parameter k i1 , k i2 , k i3 is an undetermined constant, i=1,2,…,6; by selecting a suitable design parameter k i1 , k i2 , k i3 , which can ensure the internal dynamic stability of the system; By taking the derivative of equation (18), according to the path tracking model (17), the heading angle error (12) and the lateral position error (13), the input-output subsystem can be obtained: In the formula, K z3 =[k 12 ,k 22 ,k 32 ,k 42 ,k 52 ,k 62 ] T Since the trajectory tracking model (17) is 8-dimensional and the input-output subsystem (19) is only 6-dimensional, the remaining 2-dimensional system states, namely the heading angle following error (13) and the lateral displacement following error (14), constitute the internal subsystem of the trajectory tracking control system. It can be seen that by adopting the method of redefining the system output (18), the path tracking control system (17) is decomposed into the input-output subsystem (19) and the internal subsystem (13)-(14); When a specific control input u z , so that when the output z of the input-output subsystem (19) is zero, the internal subsystems (13) and (14) are zero dynamic subsystems.
4. The intelligent electric vehicle path tracking control method based on coevolutionary genetic algorithm as claimed in claim 1, characterized in that: The implementation process of the adjacency matrix, in-degree matrix, Laplacian matrix and incidence matrix described in step 3 is as follows: Let K B =K z1 B z , K A =K z1 A z +K z2 According to formula (19), we can know that K z1 B z is a diagonal matrix; rewrite the input-output subsystem (19) into the following form: In the formula, K Aii Denotes the matrix K A The element in the i-th row and i-th column, K Aij Denotes the matrix K A The element in row i and column j, K Bii Denotes the matrix K B The element in row i and column i; According to graph theory, any multi-agent system can be represented as G = (V, E, A), where V = {1, 2, ..., 6} represents a set of nodes. represents the edge set, A=[a ij ] is the adjacency matrix of G, a ij Used to describe the relationship between each node and edge; if (i, j) ∈ E then a ij =1, otherwise a ij =0; the correlation matrix B = diag(b1,b2,...,b6), b i >0 means that the i-th agent can receive the communication information sent by the leader; b i =0 means that the i-th agent cannot receive the communication information sent by the leader; the in-degree matrix D = diag(d1, d2, ..., d6), Laplace matrix L = DA, L = [l ij ], each element is defined as: According to the communication topology of the six agents (20), the topological structure diagram of each agent in the input and output subsystem can be obtained. In the figure, agents 1-6 are the six degree-of-freedom equations of the input and output subsystems respectively. Based on graph theory, the adjacency matrix A, in-degree matrix D, Laplace matrix L and correlation matrix B between the six agents can be obtained:
5. The intelligent electric vehicle path tracking control method based on coevolutionary genetic algorithm as claimed in claim 1, characterized in that: The implementation process of the controller parameter design method based on co-evolutionary genetic algorithm optimization described in step 5 is as follows: The 2D state variables of the internal subsystem (heading angle following error The lateral displacement following error Δy) is derivatized twice so that the control u z Appear: General Substituting (22) and (23), we can obtain In the formula, From the input and output subsystem, we can see that when the system state converges to zero in a finite time, that is, When u z =-(K z1 B z ) -1 [(K z1 A z +K z2 )x z +D z ], and bring it into the internal subsystem (24), we get the zero dynamic subsystem: make: The following equation can be obtained: Substituting equation (26) into (25), we can obtain From formula (27), we can see that if [A ο -B ο (K z1 B z ) -1 K N2 K N3 ] is strictly guaranteed to be in the left half plane of the complex plane, then the equilibrium point of the original nonlinear system is asymptotically stable; therefore, select an appropriate parameter k to redefine the system output i1 , k i2 , k i3 , i = 1, 2, ..., 6, so that the matrix [A ο -B ο (K z1 B z ) -1 K N2 K N3 ] are all negative, then the zero dynamic subsystem (27) will be at the equilibrium point The path tracking control system is asymptotically stable. In order to better realize the rapid stabilization of the path tracking control system, a genetic algorithm based on co-evolution is used to adjust the parameter k i1 , k i2 , k i3 , i=1,2,…,6 for optimization design, select the characterization design parameter k i1 , k i2 , k i3 There are two types of populations with optimal solutions and penalty factors. The first type of population G_1 contains Size1 individuals, each of which is determined by the parameter k i1 , k i2 , k i3 The genetic composition of k is represented by genetic algorithm. i1 , k i2 , k i3 The second population G_2 contains Size2 individuals, each of which is composed of genes represented by the penalty factor in the penalty function. The penalty factor is optimized using a genetic algorithm. The penalty factor is continuously adjusted using a genetic algorithm under the coevolution model, and finally k that meet the constraints are obtained. i1 , k i2 , k i3 The optimal solution; the specific implementation steps are as follows: (1) Population and parameter initialization Initialize populations G_1 and G_2, set the evolutionary generations of populations G_1 and G_2 to be l1 and l2 respectively, and the maximum evolutionary generations to be l1max and l1max respectively, set l1=0, l2=0, and generate the initial value of the penalty factor randomly; (2) Termination Criteria 1 If l1=l 1max , then the iteration ends and the optimized controller parameter k is output i1 , k i2 , k i3 ; Otherwise, go to step (3); (3) Termination Criteria 2 If l2=l 2max , then the iteration ends and the optimized penalty factor is output to step (4); otherwise, jump to step (5); (4) Genetic evolution operation of controller parameters Select objective function: Use the quadratic performance index related to the heading control error and its rate of change as the optimization objective function: Where Θ∈R 4×4 is a positive symmetric matrix. Constraints: Each individual in the population must satisfy [A ο -B ο (K z1 B z ) -1 K N2 K N3 ](27)The real part of the eigenvalue is less than zero: I Γ =Re[λ Γ (A o )]<0 (29) Where Γ is an integer, 1≤Γ≤4; λ Γ (A o ) represents the matrix A o The Γth eigenvalue of ; According to the optimization objective function (28) and the inequality constraint (29), the following penalty function (30) is selected as the fitness function: In the formula, β J (l1) is the penalty factor, l1 is the parameter k i1 , k i2 , k i3 The evolutionary algebra of According to the penalty factor obtained, the fitness values of all individuals in the G_1 population are calculated using formula (30), and then the selection, crossover and mutation operators of the genetic algorithm are used to perform genetic evolution operations on the population G_1 to generate a new generation of the G_1 population, and l1 = l1 + 1, l2 = 0, and return to step (2); (5) Genetic evolution operation of penalty factor According to equations (29) and (30), the evaluation function of G_2 is selected as: In the formula, ΣF feasible Represents the sum of fitness values of all feasible solutions, num feasible represents the number of feasible solutions, max(F Ξ ) represents the maximum fitness value of individuals in the population, Ξ is an integer, 0<Ξ<size2, size2 is the population size of G_2, ∑sum viol The total number of individuals violating constraint (29), Σmum viol The total number of individuals violating the constraint (29); Using the first individual G_2(1) of population G_2 to the Size2th individual G_2(Size2) of population G_2 as penalty factors, the G_1 population is evolved using a genetic algorithm. The fitness function value of the evolved new population is calculated and sorted according to the fitness function value to generate a new generation of population G_2. Let l2=l2+1 and return to step (3).
6. The intelligent electric vehicle path tracking control method based on coevolutionary genetic algorithm as claimed in claim 1, characterized in that: The implementation process of the adaptive sliding mode control method of the input-output subsystem based on multi-agent described in step 6 is as follows: According to the basic knowledge of multi-agent graph theory and the topological structure diagram of each agent in the input and output subsystem, the deviation equation of the i-th agent following the leader is defined as follows: Where z0 is the state variable of the leader, satisfying z0=0,l ij is the element of Laplace matrix L; Let τ i =∫υ i dt, for the i-th system agent of the input-output subsystem, the following sliding surface is introduced: In the formula, c1, c2>0, 0<λ τ <1; Let β si is the sliding mode switching coefficient, β si satisfy: When the state parameters of other agents are unknown due to communication or other reasons, the parameter β si Unknown; Order β si The estimated value of , for the i-th agent (20) of the input-output subsystem, the following control strategy (35) and Adaptation rate (36): In the formula, β ui >0; The adaptive rate is designed as follows: In the formula, ρ βi is the design parameter, and ρ βi >0.
Citation Information
Cited By
Distributed driving electric vehicle double-layer game multi-agent cooperative control method
CN122525958A
