A Stability Analysis and Optimization Method for Distributed Drive Electric Vehicles

Abstract

The present invention is applicable to the technical field of stability control of distributed drive electric vehicles, and provides a stability analysis and optimization method for distributed drive electric vehicles. The method combines the advantages of the phase plane method and the integral dissipated energy method, projects the integral dissipated energy method onto the phase plane and performs corrections, and determines the final stable domain, pan-stable domain, and unstable domain division forms. The method expands the range of the stable domain and makes the defined stable domain more accurate.

Description

Technical Field

[0001] The present invention belongs to the technical field of stability control of distributed drive electric vehicles, and in particular relates to a stability analysis and optimization method for distributed drive electric vehicles. Background Art

[0002] With the increasing attention paid to environmental protection and energy security around the world, electric vehicles have developed rapidly as a clean and efficient means of transportation. Among them, distributed drive electric vehicles have higher flexibility and control accuracy due to their four-wheel independent drive and independent steering characteristics. However, this new power system structure also brings new challenges in vehicle stability. Therefore, it is necessary to study the stability control of distributed drive electric vehicles to ensure good stability and handling performance when facing complex working conditions.

[0003] When studying the stability control of distributed drive electric vehicles, in order to make full use of each wire control subsystem to achieve the coordinated control of work objectives and stability, it is necessary to accurately evaluate the vehicle's motion state and divide its stable area, in order to provide a clear basis for the timing of the stability controller's intervention. The existing stability analysis methods mainly include the Lyapunov method, the phase plane method, and the energy method.

[0004] The current research on stability control of distributed drive electric vehicles mainly uses the traditional phase plane method, which has only two dimensions to describe the state and is usually unable to accurately divide the stability domain, resulting in reduced stability control effect. It is also unable to accurately divide the stability domain independently and effectively deal with interference, posing certain safety hazards. Summary of the invention

[0005] The purpose of the embodiments of the present invention is to provide a stability analysis and optimization method for a distributed drive electric vehicle, aiming to solve the problems raised in the above background technology.

[0006] The embodiment of the present invention is implemented as follows: a stability analysis and optimization method for a distributed drive electric vehicle comprises the following steps:

[0007] Step 1: Establish a vehicle dynamics model of a distributed drive electric vehicle;

[0008] Step 2: Based on Vehicle stability analysis using the phase plane method;

[0009] Step 3: Vehicle stability analysis based on the integral dissipated energy method;

[0010] Step 4: Based on Comprehensive vehicle stability analysis combining phase plane method with integrated dissipated energy method;

[0011] Step 5: Optimize weight parameters based on comprehensive stability analysis of neural population dynamic optimization algorithm.

[0012] The embodiment of the present invention provides a stability analysis and optimization method for a distributed drive electric vehicle, which combines The advantages of the phase plane method and the integral dissipated energy method are projected onto The phase plane is corrected and the final division of stable domain, pan-stable domain and unstable domain is determined. This method expands the range of stable domain and makes the delineated stable domain more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A flow chart of a stability analysis and optimization method for a distributed drive electric vehicle provided by an embodiment of the present invention;

[0014] Figure 2 is the vehicle dynamics model;

[0015] Figure 3 is a schematic diagram of the phase plane;

[0016] Figure 4 Delineation of the stability domain for the dissipated energy method;

[0017] Figure 5 Stable domain division for comprehensive method;

[0018] Figure 6 Flowchart of the optimization algorithm for neural population dynamics. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0021] like Figure 1 As shown, a stability analysis and optimization method of a distributed drive electric vehicle provided by an embodiment of the present invention includes the following steps:

[0022] Step 1: Establish a vehicle dynamics model of a distributed drive electric vehicle;

[0023] Step 2: Based on Vehicle stability analysis using the phase plane method;

[0024] Step 3: Vehicle stability analysis based on the integral dissipated energy method;

[0025] Step 4: Based on Comprehensive vehicle stability analysis combining phase plane method with integrated dissipated energy method;

[0026] Step 5: Optimize weight parameters based on comprehensive stability analysis of neural population dynamic optimization algorithm.

[0027] As a preferred embodiment of the present invention, step 1 includes the following specific steps:

[0028] like Figure 2 As shown, the body coordinate system of the distributed wire-controlled vehicle is used as a reference, and the origin is fixed at the center of mass of the vehicle. Based on the “right-hand rule”, the center axis of the vehicle body along the direction of travel is Positive axis, perpendicular to the vehicle's direction of travel The axis points to the left. Axis positive direction. is the velocity at the center of mass of the vehicle (unit: ), is the vehicle speed in the vehicle body coordinate system Component in the axial direction (unit: ), is the vehicle speed in the vehicle body coordinate system Component in the axial direction (unit: ), is the mass center side slip angle (unit: ), is the yaw angular velocity (unit: ), is the distance from the vehicle's center of mass to the front axle (unit: ), is the distance from the vehicle's center of mass to the rear axle (unit: ), is the vehicle width (unit: ), They represent the left front wheel, right front wheel, left rear wheel and right rear wheel respectively. is the longitudinal force of the tire (unit: ), is the lateral force of the tire (unit: ), is the turning angle of the front wheels (unit: ), is the tire's side slip angle (unit: ), is the speed direction of the tire (unit: ).

[0029] According to the geometric relationship and force analysis, the dynamic equations of the distributed wire-controlled vehicle in the lateral and yaw degrees of freedom are:

[0030] Lateral Movement:

[0031] (1);

[0032] Yaw motion:

[0033] (2);

[0034] in, is the lateral acceleration (unit: ), is the moment of inertia around the z-axis of the vehicle body coordinate system (unit: ), is the vehicle curb weight (unit: ), is the yaw angular acceleration (unit: ).

[0035] The tire slip angle is represented by formulas (3)-(6):

[0036] (3);

[0037] (4);

[0038] (5);

[0039] (6);

[0040] For any tire of a distributed wire-controlled vehicle, the tire expression based on the "magic" formula is as follows:

[0041] (7);

[0042] In the formula, is the lateral force (unit: ), is the slip angle (unit: ), is the stiffness factor, is the shape factor, is the peak factor, is the curvature factor, is the horizontal deformation, is the ground normal deformation. The variables in formula (7) can be represented by formula (8).

[0043] (8);

[0044] In the formula, is the tire vertical force (unit: ), a0… a 12 is a constant determined by the properties of the vehicle's tires.

[0045] As a preferred embodiment of the present invention, step 2 includes the following specific steps:

[0046] Formula (1) and Formula (2) are abstractly expressed as shown in Formula (9):

[0047] (9);

[0048] Based on different initial values ​​of lateral velocity and yaw rate, part of the phase trajectory will converge to the equilibrium point, while the other part of the phase trajectory will diverge. According to the relevant knowledge of control theory, the equilibrium point is and The saddle point is a special equilibrium point near which the system behaves in a way that attracts and repels. Specifically, the saddle point has the following characteristics: in some directions, the system tends to the saddle point (attraction); in other directions, the system moves away from the saddle point (repulsion). According to the phase plane characteristic that the derivative of the phase variable at the saddle point is 0, combined with formula (9), the following relationship can be obtained:

[0049] (10);

[0050] According to nonlinear theory, the saddle point is the point where the phase trajectory is far away, which can be defined based on the Jacobian matrix defined in formula (11): Make further judgment.

[0051] (11);

[0052] As shown in formula (11), the Jacobian matrix must satisfy The eigenvalues ​​of are one positive and one negative.

[0053] (12);

[0054] In the formula, is the Jacobian matrix The determinant value of ,in and is the coordinate value of the saddle point.

[0055] Four straight lines A1, A2, B1, and B2 are drawn through the saddle point, such as Figure 3 As shown in the figure, the closed figure is the stable area divided by the phase plane method; the area enclosed by A1 and A2, and the area enclosed by B1 and B2 are pan-stable areas; and the other areas are unstable areas.

[0056] Under the condition that the vehicle parameters, longitudinal speed, yaw rate and front and rear wheel steering angles are all known, the extreme value of the lateral speed is determined by the saturated sideslip angle. According to the "adhesion ellipse" theory, the lateral force and longitudinal force of the tire are constrained by formula (13).

[0057] (13);

[0058] in They are the lateral, longitudinal and vertical forces of the tires of the front and rear wheels respectively.

[0059] (14);

[0060] is the tire load factor, is the road adhesion coefficient.

[0061] Based on formulas (13) and (14), the absolute value of the rear wheel saturated sideslip angle can be obtained as:

[0062] (15);

[0063] In the formula, is the rear wheel saturated slip angle (unit: ), is the equivalent cornering stiffness of the rear axle. Then the saturated lateral velocity can be obtained as:

[0064] (16);

[0065] In the formula, is the saturated lateral velocity obtained based on the rear wheel saturated sideslip angle (unit: ). In most working conditions, the distributed wire-controlled vehicle can meet the control requirements with only the front wheel steering, so the saturated lateral speed can be used as a constraint:

[0066] (17);

[0067] Among them, A1 and A2 are constrained by the lateral speed. The slopes of the stability domain boundary lines A1 and A2 in the conventional steering driving mode are fixed as follows: , when only the road adhesion coefficient is considered, its height As shown in formula (18):

[0068] (18);

[0069] in, is the width of the double straight line method, It is a height auxiliary function.

[0070] Where B1 and B2 are constrained by the yaw rate, and their characteristics are as follows:

[0071] (19);

[0072] in, is the slope of B1 and B2, Width helper function.

[0073] As a preferred embodiment of the present invention, step 3 includes the following specific steps:

[0074] Since the two-degree-of-freedom model cannot fully reflect the characteristics of distributed drive electric vehicles, formula (20) is obtained based on the longitudinal motion:

[0075] Longitudinal Movement:

[0076] (20);

[0077] Combining formula (1), formula (2) and formula (20), we get the three-degree-of-freedom model of distributed drive electric vehicles:

[0078] (twenty one);

[0079] In the nonlinear system of three-degree-of-freedom vehicle planar motion, there is no external driving torque as energy input. When the vehicle is disturbed by the outside world, the system only dissipates energy through the interaction between the tires and the ground. As time goes by, until the preset time point is reached, the total amount of energy dissipated in this process is called the total dissipated energy of the system. For such generalized systems, energy can be divided into kinetic energy, potential energy and dissipated energy. The research objects cover the vehicle system itself and its interaction with the ground. In the analysis process, the elastic potential energy caused by tire deformation and the potential energy in the non-planar direction are ignored, and it is assumed that there is no energy exchange between the system and the outside world. According to the law of conservation of energy, the total energy change of the system can be expressed as:

[0080] (twenty two);

[0081] in, is the change in kinetic energy of the system, is the dissipated energy. From the perspective of work, the negative work done by the dissipative force results in positive energy dissipation, and the power of the dissipative force is always less than zero. Considering that the longitudinal energy loss can be ignored, the dissipative force is mainly caused by the lateral force of the tire. Therefore, according to the kinetic energy theorem, the dissipated energy is equal to the change in kinetic energy or the work done by the lateral force of the tire The negative value of can be expressed as:

[0082] (twenty three);

[0083] The dissipated energy integral calculation method is used to determine the boundary of the system's stable region. This method accurately defines the stable region of the system under different conditions by integrating the dissipated energy throughout the process. This method not only provides a deep understanding of the dynamic behavior of the system, but also provides an important theoretical basis for the design and optimization of vehicle control systems.

[0084] Dissipated energy integral calculation method:

[0085] According to the definition of dissipated energy, the formula can be expressed as:

[0086] (twenty four);

[0087] Since the angle between the lateral force direction and the tire speed direction is always greater than 180°, Always a positive number.

[0088] Introducing the dissipated energy integral equation into the three-degree-of-freedom dynamic model, we obtain the four-degree-of-freedom differential equation as shown in formula (25), where To dissipate energy The derivative of , which represents the rate of change of dissipated energy:

[0089] (25);

[0090] The lateral speed is axis, and the yaw angular velocity is Axis, longitudinal speed is axis, dissipating energy at any point in this three-dimensional space , characterized by the magnitude of dissipated energy, Projection on the plane to calculate dissipated energy exist The gradient of the direction is as follows:

[0091] (26);

[0092] in, Indicates direction and v y A unit vector with the same direction and length of 1; Indicates direction and A unit vector with the same direction and length of 1; according to the dissipated energy gradient Different energy clusters, such as Figure 4 As shown, The stable domain is divided into interval contour lines, and the lowest boundary is used as the boundary of the stable domain divided by the integral dissipated energy method. is the dissipated energy interval and is a constant that depends on the expected accuracy of the dissipated energy method.

[0093] As a preferred embodiment of the present invention, step 4 includes the following specific steps:

[0094] Considering the advantages and disadvantages of the double straight line method and the dissipated energy method, the specific method of final stable domain division is based on After the phase plane method uses the multi-constraint improved double straight line method to divide the stability domain, it merges the stability domain defined by the integrated dissipated energy method with it based on the same lateral velocity and yaw angular velocity variation range.

[0095] The division results are as follows Figure 5 As shown in the figure, the intersection of the stable regions divided by the phase plane method and the dissipated energy method is used as the stable region divided by the comprehensive stability analysis, the union of the pan-stable regions is used as the pan-stable region divided by the comprehensive stability analysis, and the rest is divided into the unstable region, which effectively improves the accuracy of the stable region and expands the range of the stable region-pan-stable region, providing an analysis basis for subsequent stability control.

[0096] As a preferred embodiment of the present invention, step 5 includes the following specific steps:

[0097] The area of ​​the stable region is , the area of ​​the pan-stable region is The area of ​​the unstable region is , to ensure The phase plane method is well combined with the integrated dissipated energy method, and the proportion of the stable region to the pan-stable region is required As large as possible; To ensure the effectiveness of the combination of the two methods, the total area of ​​the stable-panstable region needs to be as large as possible. To this end, the performance indicators are set as follows:

[0098] (27);

[0099] in, For the deadline, is the weight coefficient; the optimal time-varying weight coefficient is iteratively optimized based on the neural population dynamic optimization algorithm , in order to achieve the superiority of stability analysis.

[0100] The neural population dynamic optimization algorithm is as follows:

[0101] 1) Initialization:

[0102] (28);

[0103] in, for The lower limit of for The upper limit of Indicates generating a random number between 0 and 1.

[0104] 2) Attractor trend strategy:

[0105] (29);

[0106] in, is the trend of the attractor's scaling factor, Represents a random number between 0 and 1. yes The attractor, is the proportion of attractors among all neural states of the neural population, is the number of neural populations, is the number of attractors, indicating Round down, The covariance is A zero-mean Gaussian noise vector, is the peak value of the neural state, is the minimum boundary of the neural state, It has a mean of 0 and a variance of The normal distribution of .

[0107] 3) Coupling interference strategy:

[0108] The neural state of a particular neural population is also affected by the neural states of other neural populations. It is usually assumed that the interaction between neural populations occurs through additive or diffusive coupling. In this approach, both types of coupling are present in neural population dynamics. In additive coupling, the input to a neural population is a function of the sum of the neural states of other neural populations. To simplify the calculations, the neural states of other neural populations are expressed as the average of the sum of all neural states. The formula used is as follows:

[0109] (30);

[0110] in, is a random number between 0 and 0.5. is the number of neural populations.

[0111] In diffuse coupling, the input to a neural population is a function of the sum of its neural state and the neural states of other neural populations. The formula used is as follows:

[0112] (31);

[0113] in, is a random number between 0 and 0.5.

[0114] This method uses a combination of two couplings to simulate the effects of other neural groups. The formula can be expressed as follows:

[0115] (32);

[0116] in, is the proportionality factor of the coupled interference.

[0117] Equation (30) shows that one neural population is affected by another neural population through two types of coupling, and the interference of other neural populations causes the neural population to deviate from the attractor. This search strategy based on the coupling between neural populations can help enhance the exploration ability of the algorithm.

[0118] 4) Information projection strategy:

[0119] As mentioned above, information exchange and interaction between neural populations are carried out through the communication subspace. In this method, the matrix representing adjacency and communication strength is used to represent the communication subspace of information projection between neural populations.

[0120] In the attractor trend, the information transfer between the attractor and another neural state of the neural population is represented by an adjacency matrix with binary random values and the communication strength matrix with randomly generated values ​​between 0 and 1 Adopt. The formula used is as follows:

[0121] (33);

[0122] Perturbations in the neural state should have a greater effect early in the process, before the neural state stabilizes. Therefore, the effect of coupled perturbations gradually decreases, using the following formula:

[0123] (34);

[0124] in, is an adjacency matrix with binary random values, is a communication strength matrix with randomly generated values ​​between 0 and 1, is a random number between 0 and 1. is the function evaluation degree, is the maximum function evaluation degree. As the iteration process proceeds, The influence of coupling interference gradually increases, while the influence of coupling interference gradually decreases. Eventually, the neural state of each neural group will be affected by the attractors and neural states of other neural groups. The neural state update is as follows:

[0125] (35);

[0126] This method uses the above strategy to simulate information projection between neural populations and realize the transition from exploration to development. The specific implementation process is as follows Figure 6 shown.

[0127] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A stability analysis and optimization method for a distributed drive electric vehicle, characterized in that: The following steps are involved: Step 1: Establish a vehicle dynamics model of a distributed drive electric vehicle; Step 2: Based on Vehicle stability analysis using the phase plane method, where is the vehicle speed in the vehicle body coordinate system The component in the axial direction, is the yaw angular velocity; Step 3: Vehicle stability analysis based on the integral dissipated energy method; Step 4: Based on Comprehensive vehicle stability analysis combining phase plane method with integrated dissipated energy method; Step 5: Optimize weight parameters based on comprehensive stability analysis of neural population dynamic optimization algorithm.

2. The stability analysis and optimization method of distributed drive electric vehicles according to claim 1, characterized in that: The step 1 comprises the following steps: Using the body coordinate system of the distributed wire-controlled vehicle as a reference, fix the origin at the center of mass of the vehicle Based on the "right-hand rule", the center axis of the vehicle body along the direction of travel is Positive axis, perpendicular to the vehicle's direction of travel The axis points to the left. Axis positive direction; is the velocity at the center of mass of the vehicle, is the vehicle speed in the vehicle body coordinate system The component in the axial direction, is the vehicle speed in the vehicle body coordinate system The component in the axial direction, is the sideslip angle of the center of mass, is the yaw angular velocity, is the distance from the vehicle's center of mass to the front axle, is the distance from the vehicle's center of mass to the rear axle, is the vehicle width, They represent the left front wheel, right front wheel, left rear wheel and right rear wheel respectively. is the longitudinal force of the tire, is the lateral force of the tire, is the turning angle of the front wheels, is the tire slip angle, is the speed direction of the tire; According to the geometric relationship and force analysis, the dynamic equations of the distributed wire-controlled vehicle in the lateral and yaw degrees of freedom are: Lateral Movement: （1）； Yaw motion: （2）； in, is the lateral acceleration, is the moment of inertia about the z-axis of the vehicle body coordinate system, is the vehicle curb weight, is the yaw angular acceleration; The tire slip angle is represented by formulas (3)-(6): （3）； （4）； （5）； （6）； For any tire of a distributed wire-controlled vehicle, the tire expression based on the "magic" formula is as follows: （7）； In the formula, is the lateral force, is the slip angle, is the stiffness factor, is the shape factor, is the peak factor, is the curvature factor, is the horizontal deformation, is the ground normal deformation; the variable of formula (7) can be represented by formula (8): （8）； In the formula, is the tire vertical force, a 0… a 12 is a constant determined by the properties of the vehicle's tires.

3. The stability analysis and optimization method of a distributed drive electric vehicle according to claim 2, characterized in that: The step 2 comprises the following steps: Formula (1) and Formula (2) are abstractly expressed as shown in Formula (9): （9）； Based on different initial values ​​of lateral velocity and yaw rate, part of the phase trajectory will converge to the equilibrium point, while the other part of the phase trajectory will diverge. According to the relevant knowledge of control theory, the equilibrium point is and According to the phase plane characteristics that the derivative of the phase variable at the saddle point is 0, combined with formula (9), the following relationship exists: （10）； According to nonlinear theory, the saddle point is the point where the phase trajectory is far away, which can be defined based on the Jacobian matrix defined in formula (11): Make further judgment: （11）； As shown in formula (11), the Jacobian matrix must satisfy The eigenvalues ​​of are one positive and one negative; （12）； In the formula, is the Jacobian matrix The determinant value of ,in and is the coordinate value of the saddle point; Four straight lines A1, A2, B1, and B2 are drawn through the saddle point, and the closed figures enclosed by them are the stable regions divided by the phase plane method; the regions enclosed by A1 and A2, and the regions enclosed by B1 and B2 are pan-stable regions; the other regions are unstable regions; Under the condition that the vehicle parameters, longitudinal speed, yaw rate and front and rear wheel steering angles are all known, the extreme value of the lateral speed is determined by the saturated sideslip angle; according to the "adhesion ellipse" theory, the lateral force and longitudinal force of the tire are constrained by formula (13): （13）； in are the lateral, longitudinal and vertical forces of the tires of the front and rear wheels respectively; （14）； is the tire load factor, is the road adhesion coefficient; Based on formulas (13) and (14), the absolute value of the rear wheel saturated sideslip angle can be obtained as: （15）； In the formula, is the rear wheel saturated slip angle (unit: ), is the equivalent cornering stiffness of the rear axle; and the saturated lateral velocity can be obtained as: （16）； In the formula, is the saturated lateral velocity obtained based on the saturated sideslip angle of the rear wheel; in most working conditions, the distributed wire-controlled vehicle can meet the control requirements with only the front wheel steering, so the saturated lateral velocity can be used as a constraint: （17）； A1 and A2 are constrained by the lateral speed. The slopes of the stability domain boundary lines A1 and A2 in the conventional steering driving mode are fixed as follows: , when only the road adhesion coefficient is considered, its height As shown in formula (18): （18）； in, is the width of the double straight line method, is the height auxiliary function; Where B1 and B2 are constrained by the yaw rate, and their characteristics are as follows: （19）； in, is the slope of B1 and B2, Width helper function.

4. The stability analysis and optimization method of a distributed drive electric vehicle according to claim 3, characterized in that: The step 3 comprises the following specific steps: Since the two-degree-of-freedom model cannot fully reflect the characteristics of distributed drive electric vehicles, formula (20) is obtained based on the longitudinal motion: Longitudinal Movement: （20）； Combining formula (1), formula (2) and formula (20), we get the three-degree-of-freedom model of distributed drive electric vehicles: （21）； In a nonlinear system of three-degree-of-freedom vehicle planar motion, according to the law of conservation of energy, the total energy change of the system can be expressed as: （22）； in, is the change in kinetic energy of the system, is the dissipated energy; according to the kinetic energy theorem, the dissipated energy is equal to the change in kinetic energy or the work done by the lateral force of the tire The negative value of can be expressed as: （23）； The dissipated energy integral calculation method is used to determine the boundary of the system's stable domain. By integrating the dissipated energy in the entire process, the stable region of the system under different conditions can be accurately defined.

5. The stability analysis and optimization method of distributed drive electric vehicles according to claim 4, characterized in that: In step 3, the dissipated energy integral calculation method is specifically as follows: According to the definition of dissipated energy, the formula can be expressed as: （24）； Since the angle between the lateral force direction and the tire speed direction is always greater than 180°, Always a positive number; Introducing the dissipated energy integral equation into the three-degree-of-freedom dynamic model, we obtain the four-degree-of-freedom differential equation as shown in formula (25), where To dissipate energy The derivative of , which represents the rate of change of dissipated energy: （25）； The lateral speed is axis, and the yaw angular velocity is Axis, longitudinal speed is axis, dissipating energy at any point in this three-dimensional space , characterized by the magnitude of dissipated energy, Projection on the plane to calculate dissipated energy exist The gradient of the direction is as follows: （26）； in, Indicates direction and v y A unit vector with the same direction and length of 1; Indicates direction and A unit vector with the same direction and length of 1; according to the dissipated energy gradient Divide into different energy clusters to The stable domain is divided into interval contour lines, and the lowest boundary is used as the boundary of the stable domain divided by the integral dissipated energy method, where is the dissipated energy interval and is a constant that depends on the expected accuracy of the dissipated energy method.

6. The stability analysis and optimization method of distributed drive electric vehicles according to claim 5, characterized in that: The step 4 comprises the following specific steps: In support After the phase plane method uses the multi-constraint improved double straight line method to divide the stability domain, based on the same lateral velocity and yaw angular velocity variation range, the stability domain delineated by the integrated dissipated energy method is merged with it; The intersection of stable regions divided by phase plane method and dissipated energy method is taken as the stable region divided by comprehensive stability analysis, the union of pan-stable regions is taken as the pan-stable region divided by comprehensive stability analysis, and the rest is divided into unstable regions.

7. The stability analysis and optimization method of distributed drive electric vehicles according to claim 6, characterized in that: The step 5 comprises the following specific steps: The area of ​​the stable region is , the area of ​​the pan-stable region is The area of ​​the unstable region is , to ensure The phase plane method is well combined with the integrated dissipated energy method, and the proportion of the stable region to the pan-stable region is required As large as possible; To ensure the effectiveness of the combination of the two methods, the total area of ​​the stable-pan-stable region needs to be as large as possible; for this purpose, the performance indicators are set as follows: （27）； in, For the deadline, is the weight coefficient; the optimal time-varying weight coefficient is iteratively optimized based on the neural population dynamic optimization algorithm , in order to achieve the superiority of stability analysis.

8. The stability analysis and optimization method of a distributed drive electric vehicle according to claim 7, characterized in that: In step 5, the neural population dynamic optimization algorithm is as follows: 1) Initialization: （28）； in, for The lower limit of for The upper limit of Indicates generating a random number between 0 and 1; 2) Attractor trend strategy: （29）； in, is the trend of the attractor's scaling factor, Represents a random number between 0 and 1. yes The attractor, is the proportion of attractors among all neural states of the neural population, is the number of neural populations, is the number of attractors, indicating Round down, The covariance is A zero-mean Gaussian noise vector, is the peak value of the neural state, is the minimum boundary of the neural state, It has a mean of 0 and a variance of Normal distribution of 3) Coupling interference strategy: The neural state of a specific neural population is also affected by the neural states of other neural populations. It is usually assumed that the interaction between neural populations occurs through additive or diffusive coupling, and both couplings are assumed to exist in neural population dynamics; in additive coupling, the input of the neural population is a function of the sum of the neural states of other neural populations. In order to simplify the calculation, the neural state of other neural populations is expressed as the average of the sum of all neural states. The formula used is as follows: （30）； in, is a random number between 0 and 0.

5. is the number of neural populations; In diffuse coupling, the input to a neural population is a function of the sum of its neural state and the neural states of other neural populations, using the following formula: （31）； in, is a random number between 0 and 0.5; Using a combination of two couplings to simulate the effects of other neural populations, the formula can be expressed as follows: （32）； in, is the proportionality factor of the coupled interference; 4) Information projection strategy: As mentioned above, information exchange and interaction between neural populations are carried out through the communication subspace, and the communication subspace of information projection between neural populations is represented by a matrix representing adjacency and communication strength; In the attractor trend, the information transfer between the attractor and another neural state of the neural population is represented by an adjacency matrix with binary random values and the communication strength matrix with randomly generated values ​​between 0 and 1 The formula used is as follows: （33）； Perturbations in the neural state should have a greater effect early in the process, before the neural state stabilizes; therefore, the effect of coupled perturbations gradually decreases, using the following formula: （34）； in, is an adjacency matrix with binary random values, is a communication intensity matrix with randomly generated values ​​between 0 and 1, is a random number between 0 and 1. is the function evaluation degree, is the maximum function evaluation degree; as the iteration process proceeds, gradually increases, while the influence of coupling interference gradually decreases; eventually, the neural state of each neural group will be affected by the attractors and neural states of other neural groups; the current neural population The neural state update is as follows: （35）； Use the above strategies to simulate information projection between neural populations and achieve the transition from exploration to exploitation.

Images