Steer-by-wire vehicle variable transmission ratio design method and system based on fuzzy neural network

Abstract

The invention provides a steering-by-wire vehicle variable transmission ratio design method and system based on a fuzzy neural network, and belongs to the field of vehicle steering-by-wire. According to a fixed gain method, the steering angle transmission ratio is too small at a low speed, and the sensitivity is low at a high speed; the existing intelligent algorithm excessively depends on the experience of a designer and the quality of a data sample. The method comprises the following steps: establishing a closed-loop driver-vehicle system, designing a multi-target evaluation method by using a quadratic cost function of a vehicle dynamic state, and obtaining a data relationship between a vehicle ideal variable transmission ratio characteristic and a vehicle longitudinal speed and a steering wheel angle; a nonlinear control model is established through a fuzzy RBF network, and a globally optimal solution is obtained based on nonlinear model learning. The robustness of the control system is improved, the stable and reliable transmission ratio can be provided under various vehicle conditions, and the problem that a steering system is light and flexible is solved, so that the experience feeling of a driver under various vehicle conditions is improved, and meanwhile the operation stability and safety of the vehicle are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle steer-by-wire technology, and in particular to a fuzzy neural network-based steer-by-wire vehicle transmission ratio design method and system. Background Art

[0002] The transmission ratio of the mechanical steering used in traditional steering systems is fixed, which cannot solve the contradiction between "light" and "flexible" steering. The wire-controlled steering system eliminates the mechanical connection between the steering wheel and the steering wheel, realizes structural decoupling, broadens the vehicle's torque transmission and angle transmission characteristics, and can flexibly change the angular transmission ratio according to vehicle conditions to reduce the driver's steering load, thereby improving the vehicle's handling stability.

[0003] Currently, common steer-by-wire variable transmission ratio design methods are mainly based on the "fixed-gain yaw rate gain method", "fixed-gain lateral acceleration gain method", or "hybrid gain method based on intelligent algorithms such as fuzzy algorithm, genetic algorithm, particle swarm optimization algorithm, etc." Common steer-by-wire variable transmission ratio design methods have the following problems:

[0004] The variable transmission ratio design method of steer-by-wire based on the "fixed gain method" has difficulty in establishing a mathematical model for complex or nonlinear systems:

[0005] The "fixed-gain yaw rate gain method" only considers setting the steady-state yaw rate gain to a constant value. Although this ensures consistent steering response at different vehicle speeds, it will result in an excessively small steering angle transmission ratio at low speeds and does not consider the impact of the lateral acceleration gain on the driver.

[0006] The "constant lateral acceleration gain design method" will cause the steering angle transmission ratio at high speed to be too large, resulting in too low high-speed steering sensitivity, which is not conducive to driving safety.

[0007] The wire-controlled steer-by-wire variable transmission ratio design method based on the "hybrid gain method of intelligent algorithms such as fuzzy algorithm, genetic algorithm, particle swarm optimization algorithm, etc." is too dependent on the designer's experience and data sample quality, and the physical meaning of the steering variable angle transmission ratio is unclear.

[0008] Both the "fixed gain method" and the "intelligent algorithm" based steer-by-wire variable transmission ratio design methods are difficult to provide a stable and reliable transmission ratio under special working conditions.

[0009] To address the limitations of the variable transmission ratio design solutions based on the "fixed gain method" and "intelligent algorithms", the present invention designs a multi-objective evaluation method by establishing a closed-loop driver-vehicle system and using the quadratic cost function of the vehicle's dynamic state, obtaining the data relationship between the ideal variable transmission ratio characteristics of the vehicle and the vehicle's longitudinal speed and steering wheel angle; then, a non-linear control model is established through a fuzzy RBF network, and the global optimal solution is obtained based on the learning of the non-linear model. Summary of the Invention

[0010] The technical problem to be solved by the present invention is:

[0011] To solve the problem that the existing fixed gain method cannot balance the yaw rate and lateral acceleration gains, resulting in too small a steering angle transmission ratio at low speeds and too large a steering angle transmission ratio at high speeds; and the existing intelligent algorithms based on fuzzy algorithms, genetic algorithms, and particle swarm optimization algorithms overly rely on the designer's experience and the quality of data samples, and the physical meaning of the steering variable angle transmission ratio is not clear.

[0012] The technical solution adopted by the present invention to solve the above technical problems:

[0013] The present invention provides a variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network, including the following steps:

[0014] S100. Establish a linear two-degree-of-freedom vehicle dynamics model based on Newton's second law;

[0015] S200. Establish a dynamics model based on a closed-loop driver-vehicle system;

[0016] S300. Design a multi-objective evaluation method based on the quadratic cost function of the vehicle's dynamic state to evaluate the relationship between the ideal variable steering ratio characteristics of the vehicle and the vehicle's longitudinal speed and steering wheel angle;

[0017] S400. Construct a non-linear control model based on a fuzzy neural network of the RBF model to obtain the global optimal solution of the variable transmission ratio.

[0018] Further, in step S100, the linear two-degree-of-freedom vehicle dynamics model established based on Newton's second law is:

[0019]

[0020] In the formula, a is the distance from the vehicle's center of mass to the front axle; b is the distance from the vehicle's center of mass to the rear axle; C f 、C r are the cornering stiffnesses of the front and rear tires of the vehicle respectively; m is the vehicle mass; I Z is the moment of inertia; v xis the longitudinal vehicle speed; δ is the front wheel steering angle; β is the vehicle center of mass sideslip angle; ω r is the vehicle yaw rate.

[0021] Furthermore, in step S200, additional state variables are defined: Y is the lateral displacement, and ψ is the yaw angle of the vehicle relative to the road;

[0022] Meanwhile, where, v y is the vehicle lateral motion speed, is the vehicle lateral acceleration; thus, the closed-loop driver-vehicle system dynamics model is obtained:

[0023]

[0024] Furthermore, in step S300, the steering ratio is defined as: In the formula, is the steering ratio of the steer-by-wire system; δ s is the steering wheel angle; δ p is the steering gear angle; δ f is the front wheel steering angle; is the steering actuator steering ratio; τ is the steering ratio compensation coefficient calculated by the control algorithm.

[0025] Furthermore, in S310, for the ideal transmission ratio characteristic varying with the vehicle speed, the steering ratio τ min is considered with respect to the lower limit value of the vehicle speed; in addition, an upper limit τ max is also adopted to avoid slow response during high-speed driving;

[0026] The relationship between the compensation coefficient τ and the longitudinal vehicle speed v x is described by the following non-linear function:

[0027]

[0028] where, v0 is the threshold value for the vehicle to drive in the low-speed section; v1 is the threshold value for the vehicle to drive in the high-speed section;

[0029] To determine the relationship between the compensation coefficient τ and the longitudinal vehicle speed V x a multi-objective evaluation index system for vehicle handling performance is designed; based on the vehicle dynamic state, the evaluation model is constructed through a quadratic cost function:

[0030]

[0031] In the formula, J e is the index for evaluating the vehicle trajectory tracking performance, describing the lateral displacement deviation during vehicle driving. Among them, e is the lateral deviation of the vehicle trajectory tracking during driving, e tis the threshold value of the lateral deviation technical index; J b is an index for evaluating the ease of a driver's operation of a vehicle, describing the driving comfort and steering load of the driver. The larger this index, the greater the difficulty for the driver to operate the vehicle; J c and J r are indexes for evaluating the rollover risk during vehicle driving, describing the influence of the lateral acceleration a y of the vehicle and the roll angle on the handling stability, where t represents the duration of the driving experiment;

[0032] In order to conduct a multi-objective evaluation of the steering transmission ratio, the above quadratic cost function is used to calculate the normalized performance index J, as shown in the following formula:

[0033]

[0034] In the formula: ω1, ω2, ω3, and ω4 are all weighting coefficients.

[0035] Furthermore, S320. For the ideal transmission ratio characteristic varying with the steering wheel angle, the yaw rate gain during the steering control process is an important parameter reflecting the vehicle's handling performance. The relationship between the yaw rate gain and the steering wheel angle is defined as:

[0036]

[0037] The vehicle yaw rate gain decreases with the increase of the steering wheel angle. Therefore, the response between the vehicle's steering wheel angle and the yaw rate is non-linear; in order to make the vehicle yaw rate gain relatively linear, a variable steering ratio compensation τ' is added to keep the yaw rate gain unchanged when the steering wheel angle changes; τ' is shown in the following formula:

[0038]

[0039] where, is the yaw rate gain relative to the front wheel angle, derived from the vehicle dynamics model:

[0040]

[0041] Furthermore, in step S400, the four-layer RBF type fuzzy neural network includes an input layer, a membership function layer, a rule layer, and an output layer. Among them,

[0042] In the input layer, each node of this layer is directly connected to each component of the input quantity, used to transmit the input variables to the next layer; the input variables are the vehicle longitudinal speed v x and the steering wheel angle δ s; The output of each input node i of this layer is expressed as: f1(i) = x i i = 1, 2;

[0043] In the membership function layer, each node of this layer has the function of a membership function. The Gaussian function is used as the membership function, and five fuzzy sets of Gaussian membership functions are designed: extremely small, small, medium, large, and extremely large; for the j-th node: where c ij and b j are respectively the mean and standard deviation of the Gaussian function of the j-th fuzzy set of the i-th input variable; represents the square of the Euclidean distance between the input vector i and the center of the hidden layer neuron, characterizing the matching degree of the input i to this fuzzy rule;

[0044] In the rule layer, this layer completes fuzzy rule matching through connection with the fuzzification layer, and fuzzy operations are implemented between nodes; the output of each node k is the product of all input signals of this node:

[0045]

[0046] where N i is the number of the i-th input membership functions, that is, the number of nodes in the corresponding fuzzification layer;

[0047] In the output layer, the output of each node of this layer is the weighted sum of all input signals of this node, that is:

[0048]

[0049] where l is the serial number of the output layer node, and w is the connection weight matrix between the output node and each node of the third layer.

[0050] Furthermore, in the RBF type fuzzy neural network, the adjustable parameters include: the weight coefficient of the rule; the mean c ij of the Gaussian function, the parameter of the input membership function; the standard deviation b j , the parameter of the input membership function.

[0051] A steer-by-wire vehicle variable transmission ratio design system based on a fuzzy neural network, this system has program modules corresponding to the above steps, and executes the steps in the above-mentioned steer-by-wire vehicle variable transmission ratio design method based on a fuzzy neural network when running.

[0052] A computer-readable storage medium, the computer-readable storage medium stores a computer program, and the computer program is configured to implement the steps of the steer-by-wire vehicle variable transmission ratio design method based on a fuzzy neural network when called by a processor.

[0053] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0054] The present invention provides a variable transmission ratio design method and system for a steer-by-wire vehicle based on a fuzzy neural network. By using the variable transmission ratio design method for steer-by-wire based on a fuzzy RBF neural network, the robustness of the control system is improved, and a stable and reliable transmission ratio is provided under various vehicle conditions. Experiments have shown that a smaller transmission ratio is provided under low-speed conditions of 10 km / h, meeting the requirements of steering sensitivity and operation ease at low speeds; a larger transmission ratio is provided under high-speed conditions of 120 km / h, making the vehicle steering "sluggish", solving the problem of "lightness" and "sensitivity" of the steering system, thereby improving the driving experience of the driver under various vehicle conditions and at the same time improving the handling stability and safety of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 is a flowchart of a variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network according to an embodiment of the present invention;

[0056] Figure 2 is a structural diagram of a fuzzy RBF neural network controller according to an embodiment of the present invention;

[0057] Figure 3 is a whole vehicle model diagram with a steer-by-wire system according to an embodiment of the present invention;

[0058] Figure 4 is a relationship diagram of vehicle speed, steering wheel angle and ideal transmission ratio according to an embodiment of the present invention;

[0059] Figure 5 is a comparison diagram of variable transmission ratio and fixed transmission ratio steering wheel angle curves under double lane change conditions at different vehicle speeds according to an embodiment of the present invention. Among them, (a) is a comparison diagram of variable transmission ratio and fixed transmission ratio steering wheel angle curves under double lane change conditions at a vehicle speed of 10 km / h, (b) is a comparison diagram of variable transmission ratio and fixed transmission ratio steering wheel angle curves under double lane change conditions at a vehicle speed of 60 km / h, and (c) is a comparison diagram of variable transmission ratio and fixed transmission ratio steering wheel angle curves under double lane change conditions at a vehicle speed of 120 km / h. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] In order to make the above objects, features and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be given with reference to the accompanying drawings.

[0061] Specific Embodiment 1: As shown in Figures 1 to 2 , the present invention provides a variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network, including the following steps:

[0062] S100. Establish a linear two-degree-of-freedom vehicle dynamics model based on Newton's second law:

[0063]

[0064] Wherein, a is the distance from the center of mass to the front axle; b is the distance from the center of mass to the rear axle; C f , C r are the cornering stiffnesses of the front and rear tires of the vehicle respectively; m is the mass of the vehicle; I Z is the moment of inertia; v x is the longitudinal vehicle speed; δ is the front wheel steering angle; β is the sideslip angle of the vehicle center of mass; ω r is the yaw rate of the vehicle;

[0065] S200. When considering the motion of the vehicle relative to the road by the driver, two additional state variables need to be defined: Y is the lateral displacement, and ψ is the yaw angle of the vehicle relative to the road;

[0066] Meanwhile, v y is the lateral vehicle speed, is the lateral vehicle acceleration, thus obtaining the closed-loop driver-vehicle system dynamics model:

[0067]

[0068] S300. Design a multi-objective evaluation method based on the quadratic cost function of the vehicle dynamic state, specifically including,

[0069] To evaluate the dynamic state of the vehicle, the relationship between the ideal variable steering ratio characteristic of the vehicle and the vehicle longitudinal speed and the steering wheel angle needs to be obtained;

[0070] Among them, the steering ratio of the vehicle steering system is defined as: Wherein, is the steering ratio of the steer-by-wire system; δ s is the steering wheel angle; δ f is the front wheel steering angle; δ p is the steering gear angle; is the steering ratio of the steering actuator; τ is the steering ratio compensation coefficient calculated by the control algorithm;

[0071] S310. The ideal transmission ratio characteristic changes with the vehicle speed: when the vehicle is driving at a low speed below 25 km / h, the expected steering ratio is relatively small, so as to improve the maneuverability and steering flexibility; on the other hand, to improve the driving stability and safety at a high vehicle speed above 100 km / h, it is necessary to increase the steering ratio to reduce the steering sensitivity; due to the mechanical position limitation on the steering rack, the lower limit value of the steering ratio τ min relative to the vehicle speed is considered; in addition, an upper limit τ max is also adopted to avoid the slow response during high-speed driving;

[0072] Based on the above analysis, the compensation coefficient τ and the longitudinal vehicle speed v x The relationship between them can be described by the non-linear function shown in the following formula:

[0073]

[0074] v0 is the threshold value when the vehicle is driving in the low-speed section; v1 is the threshold value when the vehicle is driving in the high-speed section;

[0075] In order to determine the relationship between the compensation coefficient τ and the longitudinal vehicle speed V x A multi-objective evaluation index system for vehicle handling performance is designed; based on the vehicle dynamic state, the evaluation model is represented by the following quadratic cost function:

[0076]

[0077] In the formula, J e is the key index to evaluate the vehicle trajectory tracking performance. e is the lateral deviation of vehicle trajectory tracking during driving, and e t is the threshold value of the lateral deviation technical index, which describes the lateral displacement deviation during vehicle driving; J b is the key index to evaluate the difficulty of the driver to operate the vehicle, which describes the driving comfort and steering load of the driver. The larger this index is, the greater the difficulty for the driver to operate the vehicle; J c and J r are the evaluation indexes for rollover risk during vehicle driving, which describe the influence of the lateral acceleration a y and the roll angle on the handling stability. t represents the duration of the driving experiment;

[0078] In order to conduct a multi-objective evaluation of the steering gear ratio, the above cost function is used to calculate the normalized performance index J, as shown in the following formula; generally speaking, the smaller the performance index is, the better the handling stability and comfort of the vehicle under this gear ratio:

[0079]

[0080] In the formula: ω1, ω2, ω3, ω4 are all weighting coefficients;

[0081] S320, the ideal gear ratio characteristics vary with the steering wheel angle:

[0082] The yaw rate gain during the steering control process is an important parameter reflecting the vehicle handling performance. The relationship between the yaw rate gain and the steering wheel angle is defined as:

[0083]

[0084] Yaw rate gain of vehicle It decreases with the increase of the steering wheel angle. Therefore, the response between the vehicle steering wheel angle and the vehicle yaw rate is non - linear. Especially when a large steering wheel angle is required, it is difficult for the driver to operate. In order to make the vehicle yaw rate gain relatively linear, a variable steering ratio compensation τ' is added to keep the vehicle yaw rate gain unchanged when the steering wheel angle changes. τ' is shown as follows: Unchanged; τ' is as follows:

[0085]

[0086] Wherein, is the yaw rate gain relative to the front wheel angle, which can be derived from the vehicle dynamics model:

[0087]

[0088] S400. The four - layer RBF - type fuzzy neural network is used as the non - linear control model, specifically including,[

[0089] Since the vehicle by - wire steering system is a complex non - linear system with multiple inputs and a single output, to solve this non - linear control problem, the present invention adopts a fuzzy neural network based on the RBF model. Based on the data relationship between the ideal variable transmission ratio characteristics of the vehicle, the vehicle longitudinal speed and the steering wheel angle obtained above, it is used to train the fuzzy neural network controller. The structure of the proposed fuzzy neural network is as Figure 2 shown, and the design of the four - layer is described as follows:

[0090] The first layer: input layer

[0091] Each node of this layer is directly connected to each component of the input quantity, and is used to transmit the input variable to the next layer. The input variables are the vehicle longitudinal speed v x and the steering wheel angle δ s ; The output of each input node i of this layer is expressed as: f1(i)=x i i = 1, 2.;

[0092] The second layer: membership function layer, that is, the fuzzification layer

[0093] Each node of this layer has the function of a membership function. The Gaussian function is used as the membership function, and 5 fuzzy sets of Gaussian membership functions are designed, namely very small (VS), small (S), medium (M), large (B) and very large (VB); For the j - th node:

[0094]

[0095] Wherein, c ij and b jThey are the mean and standard deviation of the Gaussian function of the j-th fuzzy set of the i-th input variable respectively; It represents the square of the Euclidean distance between the input vector i and the center of the hidden layer neuron, and characterizes the matching degree of the input i to this fuzzy rule;

[0096] The third layer: the rule layer, that is, the fuzzy inference layer

[0097] This layer completes the fuzzy rule matching through the connection with the fuzzification layer, and realizes the fuzzy operation between nodes; the output of each node k is the product of all input signals of this node:

[0098]

[0099] where, N i is the number of the i-th input membership functions, that is, the number of corresponding nodes in the fuzzification layer;

[0100] The fourth layer: the output layer

[0101] The output of each node in this layer is the weighted sum of all input signals of this node, that is:

[0102]

[0103] where, l is the serial number of the nodes in the output layer, w is the connection weight matrix between the output nodes and each node in the third layer, is the transmission ratio of the steer-by-wire system;

[0104] In the RBF type fuzzy neural network, there are three types of adjustable parameters:

[0105] The first type is the weight coefficient of the rule;

[0106] The second type is the mean c of the Gaussian function ij ; the parameters of the input membership function;

[0107] The third type is the standard deviation b j , the parameters of the input membership function.

[0108] Specific implementation plan two: The present invention provides a variable transmission ratio design system for a steer-by-wire vehicle based on a fuzzy neural network. This system has program modules corresponding to the above steps, and executes the steps in the above-mentioned variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network when running.

[0109] The other combinations and connection relationships in this implementation plan are the same as those in the specific implementation plan one.

[0110] Specific Embodiment 3: The present invention provides a computer-readable storage medium storing a computer program configured to implement the steps of a variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network when called by a processor.

[0111] Other combinations and connection relationships in this embodiment are the same as those in Specific Embodiment 1.

[0112] Simulation experiment: The feasibility of the invention is verified through a joint simulation of Carsim / Simulink. The default B-Class Hatchback model in Carsim is selected for simulation modeling. The parameters of this car model are shown in the following table.

[0113]

[0114] Since the original vehicle model in Carsim is a traditional mechanical steering system, when establishing a steer-by-wire vehicle model, the original steering system needs to be replaced. In the Steering System subsystem in the vehicle characteristics setting interface, set OPT_STEER_EXT = 0, and replace the entire steering part in the original vehicle model with the built steer-by-wire system.

[0115] In the CarSim / Simulink joint simulation, the parameter input to the CarSim vehicle model is the corrected ideal front wheel angle. The parameters output to the Simulink dynamics model include the actual front wheel angle, vehicle speed, yaw rate, lateral acceleration, steering wheel angle, and simulation time. The built vehicle model with a steer-by-wire system is as Figure 3 shown.

[0116] To verify the correctness of the built variable transmission ratio steer-by-wire model, a simulation comparison is made between the steer-by-wire model and the existing mechanical steering model in Carsim. When simulating, the influence of road surface changes and external disturbances is ignored. On the premise of ensuring the same working conditions and other conditions, the response of relevant indicators of the two models is compared. Through Figure 4 and Figure 5 it can be seen that through the variable transmission ratio design method for steer-by-wire based on a fuzzy RBF neural network, the present invention improves the robustness of the control system, provides a stable and reliable transmission ratio under various vehicle conditions, provides a smaller transmission ratio under the low-speed condition of 10 km / h, meeting the requirements of steering sensitivity and operation ease at low speeds; provides a larger transmission ratio under the high-speed condition of 120 km / h, making the vehicle steering "sluggish", solving the problem of "light" and "sensitive" of the steering system, thereby improving the driving experience of drivers under various vehicle conditions, and at the same time improving the handling stability and safety of the vehicle.

[0117] Although the present invention is disclosed as above, the scope of protection of the present invention is not limited thereto. Those skilled in the art of the present invention can make various changes and modifications without departing from the spirit and scope of the present disclosure, and these changes and modifications will all fall within the scope of protection of the present invention.

Claims

1. A design method for variable transmission ratio of steer-by-wire vehicles based on a fuzzy neural network, characterized in that It includes the following steps: S100. Establish a linear two-degree-of-freedom vehicle dynamics model based on Newton's second law; S200. Establish a dynamics model based on a closed-loop driver-vehicle system; S300. Design a multi-objective evaluation method based on a quadratic cost function of vehicle dynamic states to evaluate the relationship between the ideal variable steering ratio characteristics of the vehicle, the vehicle longitudinal speed, and the steering wheel angle; S400. Construct a nonlinear control model based on a fuzzy neural network of the RBF model to obtain the global optimal solution of the variable transmission ratio.

2. The variable transmission ratio design method for steer-by-wire vehicles based on a fuzzy neural network according to claim 1, characterized in that: In step S100, the linear two-degree-of-freedom vehicle dynamics model established based on Newton's second law is: Where a is the distance from the vehicle's center of mass to the front axle; b is the distance from the vehicle's center of mass to the rear axle; C f , C r are the cornering stiffnesses of the front and rear tires of the vehicle respectively; m is the vehicle mass; I Z is the moment of inertia; v x is the longitudinal vehicle speed; δ is the front wheel steering angle; β is the vehicle center of mass sideslip angle; ω r is the vehicle yaw rate.

3. The variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network according to claim 1, characterized in that: In step S200, additional state variables are defined: Y is the lateral displacement, and ψ is the yaw angle of the vehicle relative to the road; Meanwhile, where v y is the lateral movement speed of the vehicle, is the lateral acceleration of the vehicle; thus, a closed-loop driver-vehicle system dynamics model is obtained:

4. The variable transmission ratio design method for steer-by-wire vehicles based on a fuzzy neural network according to claim 1, characterized in that: In step S300, the steering ratio is defined as: In the formula, is the steering ratio of the steer-by-wire system; δ s is the steering wheel angle; δ p is the steering gear angle; δ f is the front wheel angle; is the steering ratio of the steering actuator; τ is the steering ratio compensation coefficient calculated by the control algorithm.

5. The variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network according to claim 4, characterized in that: S310. For the ideal transmission ratio characteristic varying with vehicle speed, consider the steering ratio τ min with respect to the lower limit value of the vehicle speed; in addition, an upper limit τ is also adopted max to avoid slow response during high-speed driving; Compensation factor τ and longitudinal vehicle speed v x The relationship between them is described by the following non-linear function: where, v0 is the threshold for the vehicle to travel in the low-speed section; v1 is the threshold for the vehicle to travel in the high-speed section; To determine the relationship between the compensation coefficient τ and the longitudinal vehicle speed V x A multi-objective evaluation index system for vehicle handling performance was designed; based on the vehicle's dynamic state, the evaluation model was constructed through a quadratic cost function: where J e is an index for evaluating the vehicle trajectory tracking performance, describing the lateral displacement deviation during vehicle driving. Among them, e is the lateral deviation of trajectory tracking during vehicle driving, and e t is the threshold value of the lateral deviation technical index; J b is an index for evaluating the ease of driving a vehicle by a driver, describing the driving comfort and steering load of the driver. The larger this index is, the greater the difficulty for the driver to drive the vehicle; J c and J r are the evaluation indexes of rollover risk during vehicle driving, describing the influence of the lateral acceleration a y and the roll angle on the handling stability, and t represents the duration of the driving experiment; To conduct a multi-objective evaluation of the steering transmission ratio, the above quadratic cost function is used to calculate the normalized performance index J, as shown in the following formula: In the formula: ω1, ω2, ω3, and ω4 are all weighting coefficients.

6. The design method of variable transmission ratio for steer-by-wire vehicles based on fuzzy neural network according to claim 5, characterized in that: S320. For the ideal transmission ratio characteristic varying with the steering wheel angle, the yaw rate gain during the steering control process is an important parameter reflecting the vehicle's handling performance. The relationship between the yaw rate gain and the steering wheel angle is defined as: Vehicle yaw rate gain It decreases with the increase of the steering wheel angle. Therefore, the response between the vehicle's steering wheel angle and yaw rate is non-linear. To make the vehicle yaw rate gain relatively linear, a variable steering ratio compensation τ' is added to keep the vehicle yaw rate gain unchanged when the steering wheel angle changes. τ' is shown as follows: wherein, is the yaw rate gain relative to the front wheel angle, which is derived from the vehicle dynamics model:

7. The design method of variable transmission ratio for steer-by-wire vehicle based on fuzzy neural network according to claim 6, characterized in that: In step S400, the four-layer RBF-type fuzzy neural network includes an input layer, a membership function layer, a rule layer, and an output layer, where At the input layer, each node of this layer is directly connected to each component of the input quantity, and is used to transfer the input variable to the next layer; the input variable is the longitudinal vehicle speed v x and the steering wheel angle δ s ; the output of each input node i of this layer is expressed as: f1(i) = x i i = 1, 2.; In the membership function layer, each node in this layer has the function of a membership function. The Gaussian function is used as the membership function, and five fuzzy sets of Gaussian membership functions are designed: extremely small, small, medium, large, and extremely large. For the j-th node: where c ij and b j are the mean and standard deviation of the Gaussian function of the j-th fuzzy set of the i-th input variable, respectively; represents the square of the Euclidean distance between the input vector i and the center of the hidden layer neuron, characterizing the matching degree of the input i to this fuzzy rule; In the rule layer, this layer completes fuzzy rule matching through connections with the fuzzification layer, and fuzzy operations are implemented between nodes; the output of each node k is the product of all input signals of this node: Among them, N i is the number of membership functions of the i-th input, that is, the number of nodes in the corresponding fuzzification layer; In the output layer, the output of each node in this layer is the weighted sum of all input signals of this node, that is: where, l is the serial number of the output layer node, and w is the connection weight matrix between the output node and each node in the third layer.

8. The design method of variable transmission ratio for steer-by-wire vehicles based on fuzzy neural network according to claim 7, characterized in that: In the RBF type fuzzy neural network, the adjustable parameters include: the weight coefficients of the rules; the mean value c of the Gaussian function ij , the parameters of the input membership function; the standard deviation b j , the parameters of the input membership function.

9. A variable transmission ratio design system for a steer-by-wire vehicle based on a fuzzy neural network, characterized in that: This system has program modules corresponding to the steps of any one of the above claims 1-8, and when running, it executes the steps in the above-mentioned variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is configured to implement the steps of the variable transmission ratio design method for a steer-by-wire vehicle based on a fuzzy neural network according to any one of claims 1-8 when called by a processor.

Images