A vehicle comfort braking method and vehicle considering tire longitudinal stiffness perturbation

Abstract

The application discloses a kind of considering automobile comfort braking method of tire longitudinal stiffness perturbation, automobile, it is related to vehicle braking control technical field.The application of considering automobile comfort braking method of tire longitudinal stiffness perturbation is handled to the three-dimensional feature vector formed by the driving speed of initial braking time, brake pedal displacement, brake pedal force by random forest model, to realize the correct identification of braking intention, and only when braking intention is mild braking, and enter the late braking, start comfort braking, combined with seven degrees of freedom half-car model and LuGre tire model and introduce random model predictive control considering tire longitudinal stiffness uncertainty, construct the constraint cost function J2 of longitudinal braking system at current discrete sampling time, to driving speed, vehicle longitudinal acceleration as joint optimization index, the front and rear wheel braking torque is optimized, realizes the smoothness control under the premise of guaranteeing safety, improves the comfort in later braking.

Description

Technical Field

[0001] This invention relates to the field of vehicle braking control technology, specifically to: 1. a vehicle comfort braking method considering tire longitudinal stiffness perturbation; 2. a distributed drive vehicle. Background Technology

[0002] During vehicle braking, fluctuations in vehicle pitch angle, vertical acceleration of the center of gravity, and longitudinal deceleration are the main factors affecting braking comfort. For example, frequent starts and stops can easily cause significant vehicle pitch motion and acceleration / deceleration fluctuations, severely reducing comfort. Therefore, it is necessary to study vehicle braking control methods that can balance braking safety and ride comfort.

[0003] However, existing comfort braking technologies have two main shortcomings: First, the braking intent recognition logic is too simple, which is prone to false triggering or missed triggering, and the timing of intervention in braking control is not well designed, resulting in poor actual performance. Second, they are mostly designed based on a single optimization objective and system nominal parameters. Existing methods are mostly designed based on a single optimization objective and system nominal parameters, without fully considering the uncertainty of tire longitudinal stiffness and the strong coupling characteristics between vehicle pitch motion and vertical vibration of the center of gravity. This leads to increased model prediction error and insufficient braking torque control accuracy, which in turn affects vehicle braking control performance and ride comfort. Summary of the Invention

[0004] Therefore, it is necessary to provide a vehicle comfort braking method that takes into account tire longitudinal stiffness perturbation, addressing the problems of poor braking intention recognition and insufficient braking torque control accuracy in existing comfort braking technologies.

[0005] This invention is achieved using the following technical solution: In a first aspect, the present invention discloses a vehicle comfort braking method that takes into account tire longitudinal stiffness perturbation, for braking control of a target distributed drive vehicle traveling in a straight line.

[0006] Vehicle comfort braking methods that consider tire longitudinal stiffness perturbation include: The target distributed drive vehicle's speed, brake pedal displacement, and brake pedal force at the initial moment of braking are processed into a three-dimensional feature vector, which is then input into a trained random forest model for processing to identify the braking intention. When the target distributed drive vehicle enters the later stage of braking, if the braking intention is light braking, it enters comfort braking. Combining the LuGre tire model of the target distributed drive vehicle, stochastic model predictive control is used to divide the tire longitudinal stiffness into M types of tire longitudinal stiffness scenarios and assign corresponding probability weights. Then, a constrained cost function J2 of the longitudinal braking system at the current discrete sampling time is constructed. The driving speed and vehicle longitudinal acceleration are used as optimization indicators, and the real-time optimal solutions of the front and rear wheel braking torques are solved to control the longitudinal braking system accordingly. J2 includes upper and lower bound constraints on the tire braking torque and its increment. If the braking intention is moderate or heavy braking, the longitudinal braking system maintains conventional braking.

[0007] This method or process for implementing automotive comfort braking that takes into account tire longitudinal stiffness perturbation is based on an embodiment of the present disclosure.

[0008] Secondly, the present invention discloses a distributed drive vehicle that employs the vehicle comfort braking method considering tire longitudinal stiffness perturbation disclosed in the first aspect.

[0009] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention uses a random forest model to process the three-dimensional feature vector formed by the driving speed, brake pedal displacement, and brake pedal force at the initial moment of braking to correctly identify the braking intention. Comfort braking is initiated only when the braking intention is light braking and the braking process enters the later stage. Combining a seven-degree-of-freedom half-vehicle model and a LuGre tire model, and introducing a stochastic model predictive control that considers the uncertainty of tire longitudinal stiffness, a constrained cost function J2 of the longitudinal braking system at the current discrete sampling moment is constructed. The braking torque of the front and rear wheels is optimized using driving speed and vehicle longitudinal acceleration as joint optimization indicators. This achieves smooth control while ensuring safety and improves comfort in the later stage of braking.

[0010] 2. This invention combines a seven-degree-of-freedom half-vehicle model with model predictive control to control the active suspension system throughout the braking process, constructing a constrained cost function J1 for the active suspension system. The output force of the actuators of the front and rear active suspension systems is optimized with the vehicle pitch angle, vehicle pitch velocity, vehicle pitch acceleration, and vehicle center of gravity vertical acceleration as joint optimization objectives, thereby effectively reducing vehicle pitch and vertical vibration and improving the vertical and pitch ride comfort during braking. Attached Figure Description

[0011] To more clearly illustrate the solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0012] Figure 1 A flowchart of a vehicle comfort braking method considering tire longitudinal stiffness perturbation provided in Embodiment 1 of the present invention; Figure 2 The simulation comparison results provided for Embodiment 2 of the present invention Figure 1 ; Figure 3 The simulation comparison results provided for Embodiment 2 of the present invention Figure 2 ; Figure 4 The simulation comparison results provided for Embodiment 2 of the present invention Figure 3 . Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] It should be noted that when a component is said to be "installed on" another component, it can be directly on the other component or it may be in a component that is centered on it. When a component is said to be "set on" another component, it can be directly set on the other component or it may also be in a component that is centered on it. When a component is said to be "fixed to" another component, it can be directly fixed to the other component or it may also be in a component that is centered on it.

[0015] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "or / and" as used herein includes any and all combinations of one or more of the associated listed items.

[0016] Example 1

[0017] First, it should be noted that this embodiment 1 provides a vehicle comfort braking method that considers tire longitudinal stiffness perturbation, which is designed for vehicles traveling in a straight line (i.e., only the speed V in the forward direction needs to be considered).x The target distributed drive vehicle performs braking control.

[0018] See Figure 1 This document illustrates a flowchart of a vehicle comfort braking method considering tire longitudinal stiffness perturbation provided in Embodiment 1—primarily including the following steps: Step 1: Based on the target distributed drive vehicle's speed, brake pedal displacement, and brake pedal force at the initial moment of braking, process them into a three-dimensional feature vector, and then input it into a trained random forest model for processing to identify the braking intention.

[0019] Generally, the moment when the brake pedal displacement first exceeds a preset displacement threshold or the brake pedal force first exceeds a preset force threshold is taken as the initial braking moment. The values ​​of the preset displacement threshold and the preset force threshold can be set according to the actual situation.

[0020] Vehicle speed, brake pedal displacement, and brake pedal force can be acquired by corresponding sensors installed on the target distributed drive vehicle. Then, the vehicle speed, brake pedal displacement, and brake pedal force are continuously sampled, filtered, and outlier removed. The effective values ​​of the vehicle speed, brake pedal displacement, and brake pedal force at the initial moment are extracted, and then normalized to form a three-dimensional feature vector.

[0021] Step one uses a random forest model to process the three-dimensional feature vectors to identify braking intentions—including three cases: light braking, moderate braking, and heavy braking.

[0022] Random forest models are machine learning algorithms suitable for both classification and regression tasks, and they perform particularly well when dealing with high-dimensional and noisy data. Random forest models improve accuracy and robustness by constructing multiple decision trees and combining their predictions.

[0023] It should be noted that, in order to ensure the recognition effect, step one requires the use of a trained random forest model—which is trained based on the sample dataset.

[0024] The sample dataset contains three-dimensional feature vectors of the target distributed drive vehicle under different braking conditions and the corresponding actual braking intentions. These different braking conditions include: different driving speeds, different brake pedal displacements, and different brake pedal forces. In other words, by controlling the target distributed drive vehicle to brake under different braking conditions, the driving speed, brake pedal displacement, and brake pedal force at the initial moment of braking are processed into three-dimensional feature vectors and labeled with the corresponding actual braking intentions, thus forming the sample dataset.

[0025] When training the random forest model (a random seed of 10 is recommended), a stratified sampling method should be used to divide the sample dataset into training and test sets according to a certain ratio (70% training set and 30% test set is recommended) to ensure consistent distribution of samples across categories. Based on this, it is recommended to use decision trees as base learners, setting an appropriate minimum number of leaf node samples (5 is recommended), and ensemble a certain number of decision trees (200 is recommended) to form the random forest model. Each decision tree is trained based on a training subset obtained through bootstrapping. During node splitting, some features are randomly selected to participate in the optimal split calculation, and the final braking intent is determined by voting from all decision trees (the voting from all decision trees will form a probability distribution of mild, moderate, and severe braking; the category with the highest probability is taken as the identified braking intent). During training, the splitting features, splitting thresholds, and leaf node categories of each classification decision tree are determined based on the training samples for parameter tuning, so that the random forest model can correctly identify braking intent after training. In addition, after training, the test set is input into the random forest model to test the model's braking intention recognition performance. If the model's braking intention recognition performance meets the standard, it means that the training effect meets the requirements; otherwise, the training parameters need to be adjusted and retrained. In this embodiment 1, the random forest model achieved a recognition accuracy of 98.2% on the test set, which meets the requirements.

[0026] Step 2: Combining the seven-degree-of-freedom half-vehicle model of the target distributed drive vehicle, model predictive control (MPC) is used to construct the constrained cost function J1 of the active suspension system. The vehicle pitch angle, vehicle pitch velocity, vehicle pitch acceleration, and vehicle center of gravity vertical acceleration are used as optimization indicators, and the real-time optimal solutions of the output forces of the front and rear active suspension system actuators are solved to perform corresponding active suspension system control. J1 includes upper and lower bound constraints on the output forces of the active suspension system actuators.

[0027] It should be noted that step two is carried out throughout the entire braking process, and it focuses on the control of the active suspension system to suppress vehicle pitch and vertical motion.

[0028] First, it should be noted that, in order to achieve vehicle ride comfort during straight-line braking, a seven-degree-of-freedom (DOF) half-vehicle model of the target distributed drive vehicle is constructed. This model considers the longitudinal motion of the entire vehicle, the pitch motion of the vehicle body, the vertical motion of the sprung mass, and the vertical motion and rotation of the front and rear axle wheels, while neglecting the lateral motion of the vehicle and the load transfer between the left and right wheels. In other words, under straight-line braking conditions, the main considerations are the longitudinal and vertical changes of the vehicle, while ignoring the lateral dynamics characteristics of the vehicle and the differences between the left and right sides. Therefore, a symmetrical seven-degree-of-freedom half-vehicle model is used for analysis.

[0029] For a seven-degree-of-freedom half-car model, it is mainly constructed from two aspects: suspension system dynamics and braking system dynamics.

[0030] I. Suspension System Dynamics primarily considers the actual structure of the target distributed drive vehicle: the active suspension system connects the sprung and unsprung masses, transmits vertical forces, and dampens vibrations. The active suspension system adds force actuators to the traditional passive suspension, enabling it to output active forces in real time to adjust the vehicle's attitude.

[0031] Therefore, the dynamic equations of the suspension system include: ① Vertical motion of the sprung mass: ; In the formula, m s For the sprung mass; f is the vertical acceleration of the center of mass on the spring; f f r These are the vertical forces acting on the sprung masses by the front and rear suspensions, respectively.

[0032] ② Vehicle body pitch motion: ; In the formula, I p The moment of inertia of the vehicle body during pitch; The acceleration due to the vehicle's pitch angle; l f l r These are the horizontal distances from the vehicle's center of gravity to the front and rear axles, respectively; F xf F xr These represent the longitudinal braking forces of the front and rear wheels, respectively; h represents the height of the vehicle's center of gravity; z s z is the vertical displacement of the center of mass on the spring; rf z rr These represent the vertical excitation displacements of the road surface for the front and rear wheels, respectively.

[0033] ③ Vertical motion of the unsprung mass on the front and rear axles: ; In the formula, m uf m ur These are the unsprung masses of the front and rear axles, respectively. , These are the vertical accelerations of the unsprung masses on the front and rear axles, respectively; k tf k tr These are the vertical stiffness of the front and rear tires, respectively; c tf c tr These are the vertical damping components for the front and rear tires, respectively. , These are the vertical velocities of the unsprung masses on the front and rear axles, respectively. , These represent the vertical excitation velocities of the front and rear wheels relative to the road surface.

[0034] Furthermore, the geometric relationship between the displacements of the sprung mass at the front and rear axles and the displacement of the center of mass is as follows: ; In the formula, z sf z sr θ represents the vertical displacement of the sprung mass at the front and rear axles, respectively; L represents the pitch angle of the vehicle body; L represents the distance between the front and rear axles.

[0035] ④ Suspension force: ; In the formula, k sf k sr These are the front and rear suspension spring stiffnesses, respectively; c sf c sr These are the damping coefficients for the front and rear suspensions, respectively. , These are the vertical velocities of the sprung mass at the front and rear axles, respectively; f af f ar These are the output forces of the front and rear active suspension system actuators (positive when the direction is upward).

[0036] II. The braking system dynamics mainly considers the braking torque generated by the brake, which decelerates the wheels and generates longitudinal braking force through the interaction between the tires and the ground, thereby achieving vehicle deceleration, while ignoring the effects of air resistance and rolling resistance.

[0037] Therefore, the dynamic equations of the braking system include: ⑤ Longitudinal movement of the entire vehicle: ; In the formula, m is the total mass of the vehicle. This refers to the vehicle's longitudinal acceleration.

[0038] ⑥ The front and rear wheels rotate: ; In the formula, I wf I wr These are the moments of inertia of the front and rear wheels, respectively. These represent the angular accelerations of the front and rear wheels, respectively; R is the effective rolling radius of the wheel; T bf T br These are the braking torques of the front and rear wheels, respectively.

[0039] ⑦ The relationship between the maximum longitudinal braking force and the vertical force of a tire: ; In the formula, μ f μ r These are the coefficients of friction for the front and rear wheels, respectively; F zf F zr These are the vertical loads on the front and rear wheels, respectively.

[0040] Furthermore, the tire vertical load can be further decomposed into: ; In the formula, g is the acceleration due to gravity; ΔF zf ΔF zr These represent the vertical dynamic loads on the front and rear wheels, respectively; z uf z ur These represent the vertical displacements of the unsprung masses of the front and rear axles, respectively.

[0041] ⑧. Tire rotation: ; In the formula, v r ω represents the relative velocity of the tires; ω represents the angular velocity of the wheel.

[0042] In summary, the mathematical expression for the seven-and-a-half-degree-of-freedom vehicle model of the target distributed-drive car can be compiled as follows: .

[0043] After constructing the above seven-degree-of-freedom half-vehicle model, the process of solving the real-time optimal solution of the output force of the front and rear active suspension system actuators is as follows: S101, let the state vector of the model predictive control be: The input vector of the active suspension system is: The continuous-time state equation of the active suspension system is constructed as follows: .

[0044] In the formula, x(t) represents the state vector of the model predictive control, which varies with continuous time t; These represent, in sequence, the vertical displacement of the sprung center of mass, the vertical velocity of the sprung center of mass, the vehicle pitch angle, and the vehicle pitch angular velocity, which vary with t.

[0045] u(t) represents the input vector of the active suspension system, which varies with t; f af (t), f ar (t) represents the output force of the front and rear active suspension system actuators as t changes.

[0046] Let x(t) represent the rate of change of x(t) with respect to t; A is the continuous-time state matrix of the active suspension system, used to characterize the dynamic relationship between the internal states of the active suspension system; B is the continuous-time input matrix of the active suspension system, used to characterize the influence of the output force of the actuators of the front and rear active suspension systems on the changes in the system state.

[0047] Specifically, the expressions for A and B are: ; .

[0048] S102, let the output vector of the active suspension system be: The continuous-time output equation of the active suspension system is constructed as follows: .

[0049] In the formula, y(t) represents the output vector of the active suspension system, which varies with t; These represent the vehicle body pitch angle acceleration and the sprung center of mass vertical acceleration as they change with t, respectively.

[0050] C represents the output matrix of the active suspension system, used to establish the mapping relationship between the system state and the optimized output; D represents the direct transfer matrix of the active suspension system, used to characterize the direct influence of the actuator output force of the active suspension system on the vehicle pitch angle acceleration and the vertical acceleration of the sprung center of mass.

[0051] Specifically, the expressions for C and D are: ; ; In the formula, e2, e3, and e4 are unit column vectors; e2 = [0 1 0 0] T e3=[0 0 1 0] T e4=[0 0 0 1] T .

[0052] S103, Discretize the continuous-time state equation and continuous-time output equation of the active suspension system (forward Euler discretization method is recommended), and obtain the corresponding discrete-time state equation as follows: The discrete-time output equation is: .

[0053] In the formula, x(k) represents the actual state vector of the model predictive control at the current discrete sampling time (i.e., the kth discrete sampling time); u(k) represents the actual input vector of the active suspension system at the current discrete sampling time; y(k) represents the actual output vector of the active suspension system at the current discrete sampling time; A d B represents the discrete-time state matrix of the active suspension system. d This represents the discrete-time input matrix of the active suspension system.

[0054] Specifically, A d B d The expression is: A d =I+T s A; B d =T s B; In the formula, I is the identity matrix, and T sThe sampling time interval (usually 0.001s).

[0055] S104. Based on the discrete-time state equation and discrete-time output equation of the active suspension system, the constrained cost function J1 of the active suspension system is constructed.

[0056] The expression for the constrained cost function J1 is: ; ; In the formula, N represents the prediction time domain length (usually taken as 10); j represents the prediction step number within the prediction time domain; This represents the output error vector of the active suspension system at step j+1; y j This represents the predicted output vector of the active suspension system at step j+1. This represents the expected response of the active suspension system at step j+1; u j x represents the prediction input vector of the active suspension system at step j+1; j Q represents the predicted state vector of the active suspension system at step j+1; d This represents the output error weighting matrix of the active suspension system, where its four diagonal elements correspond to the error weights of vehicle pitch angle, vehicle pitch velocity, vehicle pitch acceleration, and vertical acceleration of the sprung center of mass, respectively; R d This represents the input weighting matrix of the active suspension system, where its two diagonal elements correspond to the weights of the output forces of the front and rear active suspension system actuators, respectively; S d The cross-term weighted matrix of the active suspension system is represented (generally a zero matrix); x0 represents the active suspension system state obtained at the start of the prediction time domain; x current This indicates the state of the active suspension system obtained at the current moment in the prediction time domain; Indicate u j The lower limit of the allowable range; Indicate u j The upper limit of the allowed range.

[0057] In other words, ;R d =diag(r af ,r ar );S d =0 4×2 .

[0058] In the formula, diag(·) represents a diagonal function; These are the error weights for vehicle pitch angle, vehicle pitch velocity, vehicle pitch acceleration, and vertical acceleration of the sprung center of mass, respectively; r af r arThese are the weights of the output forces of the actuators in the front and rear active suspension systems, respectively.

[0059] S105, Solve for the input sequence of the active suspension system that minimizes J1. and take The first input vector As the real-time optimal solution for the output force of the actuators in the front and rear active suspension systems.

[0060] in, ; express The (j+1)th input vector.

[0061] S105 will perform N control predictions in the prediction time domain with the objective of minimizing J1, thus forming Just take The first input vector contains the corresponding output forces of the front and rear active suspension system actuators, and serves as the real-time optimal solution for the output forces of the front and rear active suspension system actuators.

[0062] As time progresses, repeat steps S101 to S105 above to achieve rolling time-domain control.

[0063] Step 3: When the target distributed drive vehicle has not entered the late braking stage, apply conventional braking to the longitudinal braking system; When the target distributed drive vehicle enters the later stage of braking, if the braking intention is light braking, it enters comfort braking. Combining the LuGre tire model of the target distributed drive vehicle, stochastic model predictive control (SMPC) is used to divide the tire longitudinal stiffness into M types of tire longitudinal stiffness scenarios and assign corresponding probability weights. Then, a constrained cost function J2 of the longitudinal braking system at the current discrete sampling time is constructed. The driving speed and vehicle longitudinal acceleration are used as optimization indicators, and the real-time optimal solutions of the front and rear wheel braking torques are solved to control the longitudinal braking system accordingly. J2 includes upper and lower bound constraints on the tire braking torque and its increment. If the braking intention is moderate or heavy braking, the longitudinal braking system maintains conventional braking.

[0064] It should be noted that step three differs from step two; it is designed for the later stages of braking and focuses on controlling the longitudinal braking system. Comfort braking aims to optimize longitudinal deceleration, while conventional braking prioritizes braking safety.

[0065] The criteria for determining the later stage of braking are as follows: if the target distributed drive vehicle's speed is lower than a preset speed threshold during braking, then the target distributed drive vehicle enters the later stage of braking; otherwise, the target distributed drive vehicle has not entered the later stage of braking. The preset speed threshold can be set according to actual conditions.

[0066] Comfort braking incorporates the LuGre tire model for target distributed drive vehicles—which takes into account longitudinal forces and friction.

[0067] Therefore, the dynamic equations of the LuGre tire model include: ① Longitudinal force: ; In the formula, F x For the longitudinal force of the tire; z t For the internal friction state variables of the tire rubber; Indicate z t First-order differential; v r σ0 is the tire relative velocity; σ1 is the tire rubber longitudinal lumped stiffness (recommended value of 1); σ2 is the tire rubber longitudinal lumped damping (recommended value of 1); σ2 is the tire viscous relative damping (recommended value of 0); F z This refers to the vertical load on the tire.

[0068] ② Internal friction state variables: ; In the formula, θ t For pavement characterization parameters (recommended value is 1); g(v) r ) is the friction characteristic function; κ is the parameter representing the vertical force distribution on the tire contact surface (6.0 is recommended); ω is the wheel angular velocity; R is the effective rolling radius of the wheel.

[0069] In addition, g(v r This describes the relationship between the static friction coefficient and the kinetic friction coefficient, and its expression is: ; In the formula, μ c μ is the nominal coulomb coefficient of friction (0.8 is recommended); s v is the nominal static friction coefficient (recommended value is 1.5); s α is the relative velocity of Stribeck (5.5 is recommended); α is the shape parameter (0.5 is recommended).

[0070] In summary, the mathematical expression for the LuGre tire model of the target distributed drive vehicle can be compiled as follows: .

[0071] After constructing the LuGre tire model described above, and combining it with the previous seven-degree-of-freedom half-vehicle model, the process of solving for the real-time optimal solution of the braking torque of the front and rear wheels is as follows: S201, let the state vector of the stochastic model predictive control be: The input vector of the longitudinal braking system is: The nonlinear state equation of the longitudinal braking system is constructed as follows: .

[0072] In the formula, X(t) represents the state vector of the stochastic model predictive control, which varies with t; The values ​​represent, in order, the vehicle's longitudinal speed, the internal friction state variables of the front tire rubber, the internal friction state variables of the rear tire rubber, the relative speed of the front tire, and the relative speed of the rear tire, all varying with time (t).

[0073] U(t) represents the input vector of the longitudinal braking system, which varies with t; T bf (t), T br (t) represents the front and rear wheel braking torques as they change with t.

[0074] denoted by X(t), f(·) represents the rate of change of X(t) with respect to t; σ0 represents the tire longitudinal stiffness.

[0075] S202, let the output vector of the longitudinal braking system be: The nonlinear output equation of the longitudinal braking system is constructed as follows: .

[0076] In the formula, represents the vehicle's longitudinal acceleration as a function of time t; h(·) represents the nonlinear output function.

[0077] S203, the nonlinear state equation and nonlinear output equation of the longitudinal braking system are locally linearized for each of the M tire longitudinal stiffness scenarios, resulting in the continuous-time increment linearized state equation for the i-th tire longitudinal stiffness scenario: The linear output equation is: ;i=1,2,..,M.

[0078] Specifically, dividing σ0 into M types of tire longitudinal stiffness scenarios, we obtain: σ 0,1 σ 0,2 、…、σ 0,M ; where σ 0,i This represents the longitudinal stiffness of the tire in the i-th scenario.

[0079] σ0 follows a normal distribution: ; This represents the average longitudinal stiffness of the tire under M different tire longitudinal stiffness scenarios. This represents the variance of tire longitudinal stiffness for M different tire longitudinal stiffness scenarios.

[0080] Then, the probability weight W for the i-th tire longitudinal stiffness scenario is calculated based on the Gaussian distribution. i ,but: .

[0081] For the i-th type of tire longitudinal stiffness scenario, there exists a linear operating point related to t. : express The linearized operating point state; express Linearized operating point input; This represents the linearized operating point output for the i-th type of tire longitudinal stiffness scenario.

[0082] exist The nonlinear state equation and nonlinear output equation of the longitudinal braking system are linearized to order one, and the linearization constant term is incorporated into the reference trajectory, thereby obtaining the continuous-time incremental linearized state equation and linear output equation for the i-th tire longitudinal stiffness scenario: , .

[0083] In the formula, This represents the state deviation of the longitudinal braking system under the i-th tire longitudinal stiffness scenario, i.e. ; for The derivative with respect to t; This indicates the input deviation of the longitudinal braking system, i.e. ; This represents the output deviation of the longitudinal braking system under the i-th tire longitudinal stiffness scenario, i.e. A i Let B be the continuous-time state Jacobian matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario; i The continuous-time input Jacobian matrix for the longitudinal braking system in the i-th tire longitudinal stiffness scenario; C i Output the Jacobian matrix for the longitudinal braking system in the i-th tire longitudinal stiffness scenario; D i Let be the input-output Jacobian matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario.

[0084] S204, Discretize the continuous-time increment linearized state equation and linear output equation for the i-th tire longitudinal stiffness scenario (forward Euler discretization method is recommended) to obtain the corresponding discrete-time increment state equation: The discrete-time increment output equation is: .

[0085] In the formula, This represents the state deviation of the longitudinal braking system at the current discrete sampling moment under the i-th tire longitudinal stiffness scenario; This represents the input deviation of the longitudinal braking system at the current discrete sampling time; A represents the output deviation of the longitudinal braking system at the current discrete sampling moment under the i-th tire longitudinal stiffness scenario; d,i B represents the discrete-time state matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario; d,i This represents the discrete-time input matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario.

[0086] in addition, , , satisfy: , , ; In the formula, express The state in; express Input in; express The output; X represents the linear operating point corresponding to the current discrete sampling time under the i-th tire longitudinal stiffness scenario; i (k) represents the state vector of the longitudinal braking system at the current discrete sampling moment under the i-th tire longitudinal stiffness scenario; U(k) represents the actual input vector of the longitudinal braking system at the current discrete sampling moment; Y i (k) represents the output vector of the longitudinal braking system at the current discrete sampling time under the i-th tire longitudinal stiffness scenario.

[0087] A d,i B d,i The expression is: A d,i =I+T s A i ; B d,i =T s B i .

[0088] It should be noted that different tire longitudinal stiffness scenarios share the same set of actual inputs, so the subscript i of U(k) can be ignored.

[0089] S205, based on the discrete-time incremental state equation and discrete-time incremental output equation under various tire longitudinal stiffness scenarios, constructs the constrained cost function J2 of the longitudinal braking system at the current discrete sampling time.

[0090] The constrained cost function J2 of the longitudinal braking system at the current discrete sampling time is expressed as follows: ; ; In the formula, W i Let be the probability weight for the i-th type of tire longitudinal stiffness scenario; Let be the output error vector of the longitudinal braking system at the (j+1)th step under the i-th tire longitudinal stiffness scenario; ;Y i,j This represents the predicted output vector of the longitudinal braking system at the (j+1)th step under the i-th tire longitudinal stiffness scenario. U represents the expected response of the longitudinal braking system at step j+1; j Let U(k) represent the predicted input vector of the longitudinal braking system at step j+1 (similar to U(k), all tire longitudinal stiffness scenarios share the same predicted input vector at step j+1, therefore U...). j The subscript i is also ignored); U(k-1) represents the input vector at the previous discrete sampling time (i.e., the (k-1)th discrete sampling time); U0 represents the initial input vector in the prediction time domain, which corresponds to the current discrete sampling time; ΔU0 represents the increment of U0 compared to U(k-1); U h ΔU represents the input vector at step h+1 in the prediction time domain; h U h Compared to U h-1 The increment; Q d,i R represents the output error weighting matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario, where its two diagonal elements correspond to the error weights of driving speed and vehicle longitudinal acceleration, respectively; d,i This represents the input weighting matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario, with its two diagonal elements corresponding to the weights of the front and rear wheel braking torque increments, respectively. X represents the state of the longitudinal braking system at the initial moment in the time domain under the i-th tire longitudinal stiffness scenario; current This indicates the state of the longitudinal braking system obtained at the current moment in the prediction time domain; X represents the state deviation of the longitudinal braking system at step j+1 under the i-th tire longitudinal stiffness scenario; i,j This represents the predicted state vector of the longitudinal braking system at step j+1 under the i-th tire longitudinal stiffness scenario. This represents the output deviation of the longitudinal braking system at step j+1 under the i-th tire longitudinal stiffness scenario. This represents the input deviation of the longitudinal braking system at step j+1. , They represent U respectively j The lower and upper limits of the allowed range; , They represent ΔU respectively j The lower and upper limits of the allowed range; express The state in; express Input in; express The output in; Let be the linear operating point of the longitudinal braking system at step j+1 under the i-th tire longitudinal stiffness scenario.

[0091] Another point to note is ΔU j The input increment at step j compared to step j-1 in the prediction time domain can be written as: ;ΔT bf,j Indicates ΔU j Increment of braking torque of the inner front wheel; ΔT br,j Indicates ΔU j Increased braking torque of the inner rear wheel.

[0092] In other words, ;R d,i =diag(r Δbf,i ,r Δbr,i ).

[0093] In the formula, These are the error weights for the driving speed and vehicle longitudinal acceleration under the i-th tire longitudinal stiffness scenario, respectively; r Δbf,i r Δbr,i These are the weights of the increase in braking torque of the front and rear wheels under the i-th tire longitudinal stiffness scenario.

[0094] S206, Solve for the longitudinal braking system input increment sequence that minimizes J2. and take Increment of the first input vector The input vector U(k-1) at the previous discrete sampling time is added to obtain the real-time optimal solution of the braking torque of the front and rear wheels.

[0095] in, ; express The increment of the (j+1)th input vector.

[0096] As defined in S205, N control predictions will be performed within the prediction time domain with the objective of minimizing J2, thus forming Just take within Adding it to U(k-1) yields the optimal control input vector at the current discrete sampling time. ,Right now .

[0097] It includes the corresponding front and rear wheel braking torques and serves as the real-time optimal solution for the front and rear wheel braking torques. That is... , This is the real-time optimal solution for the front wheel braking torque; This is the real-time optimal solution for the rear wheel braking torque.

[0098] As time progresses, repeat steps S201 to S206 above to achieve rolling time-domain control.

[0099] In summary, the vehicle comfort braking method considering tire longitudinal stiffness perturbation provided in Embodiment 1 can effectively improve pitch, vertical and longitudinal ride comfort during braking while ensuring braking safety.

[0100] This embodiment 1 also discloses a distributed drive vehicle, which adopts the above-mentioned vehicle comfort braking method that considers tire longitudinal stiffness perturbation, and can effectively optimize the braking experience of the target distributed drive vehicle traveling in a straight line.

[0101] Example 2

[0102] To illustrate the effectiveness of the vehicle comfort braking method considering tire longitudinal stiffness perturbation provided in Example 1, Example 2 simulates it under a straight-line light braking condition and compares it with full-range conventional braking under the same conditions. The results are as follows: Figures 2-4 .

[0103] See Figure 2 It shows the variation curves of the vertical acceleration of the center of mass under the two schemes; see [link / reference]. Figure 3 The diagram illustrates the changes in vehicle pitch angle acceleration under two different scenarios. It shows that during braking, the active suspension system using model predictive control in Example 1 is significantly superior to conventional braking in suppressing vertical acceleration and vehicle pitch angle acceleration, and almost eliminates the sudden change in vehicle pitch angle acceleration at the moment of stopping.

[0104] See Figure 4 The diagram illustrates the braking acceleration variation curves under two different schemes. It shows that the method in Example 1, employing stochastic model predictive control, results in a smoother braking acceleration variation, and the acceleration amplitude at the critical point is significantly lower than that under conventional braking conditions throughout the entire braking process, effectively optimizing longitudinal ride comfort.

[0105] Furthermore, when the initial driving speed is set the same, the vehicle comfort braking method that takes into account the longitudinal stiffness perturbation of the tires results in only a slight increase in braking distance compared to conventional braking throughout the entire braking process: for example, when the initial driving speed is 50 km / h, the braking distance is only 3 cm longer—its impact on braking safety is almost negligible.

[0106] Example 3

[0107] This embodiment 3 discloses a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the automobile comfort braking method considering tire longitudinal stiffness perturbation disclosed in embodiment 1.

[0108] The computer equipment can be either a mobile terminal or a fixed terminal. Examples of the former include mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (such as in-vehicle navigation terminals); examples of the latter include digital TVs and desktop computers.

[0109] This embodiment 3 also discloses a readable storage medium storing computer program instructions. When the computer program instructions are read and executed by a processor, the steps of the vehicle comfort braking method considering tire longitudinal stiffness perturbation disclosed in embodiment 1 are performed.

[0110] The readable storage medium may include, but is not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above.

[0111] This embodiment 3 also discloses a computer program product, including a computer program. When executed by a processor, the computer program implements the steps of the vehicle comfort braking method considering tire longitudinal stiffness perturbation disclosed in embodiment 1.

[0112] It should be noted that the computer program used to execute the above can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—as well as conventional procedural programming languages—such as C or similar languages. The computer program can be executed entirely on the user's computer, partially on the user's computer, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer through any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN).

[0113] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0114] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for comfortable braking of a vehicle considering tire longitudinal stiffness perturbation, characterized in that, It is used for braking control of a target distributed drive vehicle traveling in a straight line; it includes: The target distributed drive vehicle's speed, brake pedal displacement, and brake pedal force at the initial moment of braking are processed into a three-dimensional feature vector, which is then input into a trained random forest model for processing to identify the braking intention. When the target distributed drive vehicle enters the later stage of braking, if the braking intention is light braking, it enters comfort braking. Combining the seven-degree-of-freedom half-vehicle model and LuGre tire model of the target distributed drive vehicle, stochastic model predictive control is used to divide the tire longitudinal stiffness into M types of tire longitudinal stiffness scenarios and assign corresponding probability weights. Then, a constrained cost function J2 of the longitudinal braking system at the current discrete sampling time is constructed. The driving speed and vehicle longitudinal acceleration are used as optimization indicators, and the real-time optimal solutions of the front and rear wheel braking torques are solved to control the longitudinal braking system accordingly. J2 includes upper and lower bound constraints on the tire braking torque and its increment. If the braking intention is moderate or heavy braking, the longitudinal braking system maintains conventional braking.

2. The vehicle comfort braking method considering tire longitudinal stiffness perturbation according to claim 1, characterized in that, The initial braking moment is the moment when the brake pedal displacement first exceeds the preset displacement threshold or the brake pedal force first exceeds the preset force threshold. The method for obtaining the three-dimensional feature vector includes: continuously sampling the driving speed, brake pedal displacement, and brake pedal force of the target distributed drive vehicle, filtering and outlier removal, then extracting the effective values ​​of driving speed, brake pedal displacement, and brake pedal force corresponding to the initial moment, and then performing normalization processing to form the three-dimensional feature vector.

3. The vehicle comfort braking method considering tire longitudinal stiffness perturbation according to claim 1, characterized in that, Random forest models are trained based on sample datasets; The sample dataset contains the three-dimensional feature vectors of the target distributed drive vehicle under different braking conditions and the corresponding true braking intentions; among which, different braking conditions include: different driving speeds, different brake pedal displacements, and different brake pedal forces.

4. The vehicle comfort braking method considering tire longitudinal stiffness perturbation according to claim 1, characterized in that, If the target distributed drive vehicle's speed is lower than a preset speed threshold during braking, the target distributed drive vehicle enters the later stage of braking; otherwise, the target distributed drive vehicle does not enter the later stage of braking. When the target distributed drive vehicle has not entered the late braking stage, the longitudinal braking system is subjected to conventional braking.

5. The vehicle comfort braking method considering tire longitudinal stiffness perturbation according to claim 1, characterized in that, The mathematical expression for the seven-and-a-half-degree-of-freedom vehicle model of the target distributed drive vehicle is: ； In the formula, m s For the sprung mass; f is the vertical acceleration of the center of mass on the spring; f f r These are the vertical forces acting on the sprung masses by the front and rear suspensions, respectively; I p The moment of inertia of the vehicle body during pitch; The acceleration due to the vehicle's pitch angle; l f l r These are the horizontal distances from the vehicle's center of gravity to the front and rear axles, respectively; F xf F xr These represent the longitudinal braking forces of the front and rear wheels, respectively; h represents the height of the vehicle's center of gravity; z rf z rr These represent the vertical excitation displacements of the front and rear wheels on the road surface, respectively; m uf m ur These are the unsprung masses of the front and rear axles, respectively. These are the vertical accelerations of the unsprung masses on the front and rear axles, respectively; k tf k tr These are the vertical stiffness of the front and rear tires, respectively; c tf c tr These are the vertical damping components for the front and rear tires, respectively. These are the vertical velocities of the unsprung masses on the front and rear axles, respectively. These represent the vertical excitation velocities of the front and rear wheels on the road surface, respectively; z sf z sr These represent the vertical displacements at the front and rear axles of the sprung mass, respectively; z s θ is the vertical displacement of the sprung center of mass; θ is the vehicle pitch angle; k sf k sr These are the front and rear suspension spring stiffnesses, respectively; c sf c sr These are the damping coefficients for the front and rear suspensions, respectively. These are the vertical velocities of the sprung mass at the front and rear axles, respectively; f af f ar These represent the output forces of the front and rear active suspension system actuators, respectively; m is the total mass of the vehicle. I is the longitudinal acceleration of the vehicle. wf I wr These are the moments of inertia of the front and rear wheels, respectively. These represent the angular accelerations of the front and rear wheels, respectively; R is the effective rolling radius of the wheel; T bf T br These are the braking torques of the front and rear wheels, respectively; μ f μ r These are the coefficients of friction for the front and rear wheels, respectively; F zf F zr The vertical loads on the front and rear wheels are respectively; g is the acceleration due to gravity; ΔF zf ΔF zr These represent the vertical dynamic loads on the front and rear wheels, respectively; z uf z ur These represent the vertical displacements of the unsprung masses of the front and rear axles, respectively; v r ω is the relative velocity of the tires; ω is the angular velocity of the wheel. The mathematical expression for the tire model of the target distributed drive vehicle is: ； In the formula, F x For the longitudinal force of the tire; z t For the internal friction state variables of the tire rubber; Indicate z t The first derivative; σ0 is the longitudinal lumped stiffness of the tire rubber; σ1 is the longitudinal lumped damping of the tire rubber; σ2 is the viscous relative damping of the tire; F z For the vertical load on the tire; θ t For road surface characterization parameters; g(v) r ) is the friction characteristic function; κ is the parameter characterizing the vertical force distribution on the tire contact surface.

6. The vehicle comfort braking method considering tire longitudinal stiffness perturbation according to claim 5, characterized in that, The methods for finding the real-time optimal solution for the braking torque of the front and rear wheels include: S201, let the state vector of the stochastic model predictive control be: The input vector of the longitudinal braking system is: The nonlinear state equation of the longitudinal braking system is constructed as follows: ； In the formula, X(t) represents the state vector of the stochastic model predictive control, which varies with t; The following variables represent the vehicle's longitudinal speed, the internal friction state variables of the front tire rubber, the internal friction state variables of the rear tire rubber, the relative speed of the front tire, and the relative speed of the rear tire, respectively, varying with time t; U(t) represents the input vector of the longitudinal braking system, which varies with time t; T bf (t), T br (t) represents the front and rear wheel braking torques as they change with t; X(t) represents the rate of change of X(t) with respect to t; f(·) represents the nonlinear state function; σ0 is the longitudinal stiffness of the tire; S202, let the output vector of the longitudinal braking system be: The nonlinear output equation of the longitudinal braking system is constructed as follows: ; In the formula, The longitudinal acceleration of the vehicle varies with t; h(·) represents the nonlinear output function; S203, the nonlinear state equation and nonlinear output equation of the longitudinal braking system are locally linearized for each of the M tire longitudinal stiffness scenarios, resulting in the continuous-time increment linearized state equation for the i-th tire longitudinal stiffness scenario: The linear output equation is: i = 1, 2, ..., M; In the formula, This represents the state deviation of the longitudinal braking system under the i-th tire longitudinal stiffness scenario; for The derivative with respect to t; This indicates the input deviation of the longitudinal braking system; A represents the output deviation of the longitudinal braking system under the i-th tire longitudinal stiffness scenario; i Let B be the continuous-time state Jacobian matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario; i The continuous-time input Jacobian matrix for the longitudinal braking system in the i-th tire longitudinal stiffness scenario; C i Output the Jacobian matrix for the longitudinal braking system in the i-th tire longitudinal stiffness scenario; D i Let be the input-output Jacobian matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario; S204, Discretize the continuous-time increment linearized state equation and linear output equation for the i-th tire longitudinal stiffness scenario to obtain the corresponding discrete-time increment state equation: The discrete-time increment output equation is: ; In the formula, This represents the state deviation of the longitudinal braking system at the current discrete sampling moment under the i-th tire longitudinal stiffness scenario; This represents the input deviation of the longitudinal braking system at the current discrete sampling time; A represents the output deviation of the longitudinal braking system at the current discrete sampling moment under the i-th tire longitudinal stiffness scenario; d,i B represents the discrete-time state matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario; d,i This represents the discrete-time input matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario; S205, Based on the discrete-time incremental state equation and discrete-time incremental output equation under various tire longitudinal stiffness scenarios, the constrained cost function J2 of the longitudinal braking system at the current discrete sampling time is constructed. The constrained cost function J2 of the longitudinal braking system at the current discrete sampling time is expressed as follows: ； ； In the formula, W i Let be the probability weight for the i-th type of tire longitudinal stiffness scenario; Y is the output error vector of the longitudinal braking system at the (j+1)th step under the i-th tire longitudinal stiffness scenario; i,j U represents the predicted output vector of the longitudinal braking system at the (j+1)th step under the i-th tire longitudinal stiffness scenario; j U(k-1) represents the predicted input vector of the longitudinal braking system at step j+1; U(k-1) represents the previous discrete sampling time; U0 represents the initial input vector in the prediction time domain; ΔU0 represents the increment of U0 compared to U(k-1); U h ΔU represents the input vector at step h+1 in the prediction time domain; h U h Compared to U h-1 The increment; Q d,i R represents the output error weighting matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario, where its two diagonal elements correspond to the error weights of driving speed and vehicle longitudinal acceleration, respectively; d,i This represents the input weighting matrix of the longitudinal braking system in the i-th tire longitudinal stiffness scenario, with its two diagonal elements corresponding to the weights of the front and rear wheel braking torque increments, respectively. X represents the state of the longitudinal braking system at the initial moment in the time domain under the i-th tire longitudinal stiffness scenario; current This indicates the state of the longitudinal braking system obtained at the current moment in the prediction time domain; X represents the state deviation of the longitudinal braking system at step j+1 under the i-th tire longitudinal stiffness scenario; i,j This represents the predicted state vector of the longitudinal braking system at step j+1 under the i-th tire longitudinal stiffness scenario. This represents the output deviation of the longitudinal braking system at step j+1 under the i-th tire longitudinal stiffness scenario. This represents the input deviation of the longitudinal braking system at step j+1. , They represent U j The lower and upper limits of the allowed range; , They represent ΔU respectively j The lower and upper bounds of the allowable range; ΔU j This represents the input increment at step j compared to step j-1 within the prediction time domain; express The state in; express Input in; express The output in; Let J be the linear operating point of the longitudinal braking system at step j+1 under the i-th tire longitudinal stiffness scenario. S206, Solve for the longitudinal braking system input increment sequence that minimizes J2. and take Increment of the first input vector Add the input vector U(k-1) from the previous discrete sampling time to obtain the real-time optimal solution for the braking torque of the front and rear wheels; in, ; express The increment of the (j+1)th input vector.

7. The vehicle comfort braking method considering tire longitudinal stiffness perturbation according to claim 6, characterized in that, σ0 follows a normal distribution: ; This represents the average longitudinal stiffness of the tire under M different tire longitudinal stiffness scenarios. This represents the variance of tire longitudinal stiffness for M different tire longitudinal stiffness scenarios. W i Based on calculations using the Gaussian distribution: ; In the formula, σ 0,i This represents the longitudinal stiffness of the tire in the i-th scenario.

8. The vehicle comfort braking method considering tire longitudinal stiffness perturbation according to claim 5, characterized in that, It also includes: Based on the seven-degree-of-freedom half-vehicle model of the target distributed drive vehicle, a constrained cost function J1 for the active suspension system is constructed using model predictive control. The vehicle pitch angle, vehicle pitch velocity, vehicle pitch acceleration, and vehicle center of gravity vertical acceleration are used as optimization indicators, and the real-time optimal solutions of the output forces of the front and rear active suspension system actuators are solved to perform corresponding active suspension system control. J1 includes upper and lower bound constraints on the output forces of the active suspension system actuators.

9. The vehicle comfort braking method considering tire longitudinal stiffness perturbation according to claim 8, characterized in that, The methods for finding the real-time optimal solution of the actuator output force of the front and rear active suspension systems include: S101, let the state vector of the model predictive control be: The input vector of the active suspension system is: The continuous-time state equation of the active suspension system is constructed as follows: ; In the formula, x(t) represents the state vector of the model predictive control, which varies with continuous time t; The vertical displacement of the sprung center of mass, the vertical velocity of the sprung center of mass, the vehicle pitch angle, and the vehicle pitch angular velocity, respectively, vary with t; u(t) represents the input vector of the active suspension system, which varies with t; f af (t), f ar (t) represents the output force of the front and rear active suspension system actuators as t changes; Let x(t) represent the rate of change of x(t) with respect to t; A is the continuous-time state matrix of the active suspension system; B is the continuous-time input matrix of the active suspension system. S102, let the output vector of the active suspension system be: The continuous-time output equation of the active suspension system is constructed as follows: ; In the formula, y(t) represents the output vector of the active suspension system, which varies with t; The values ​​represent the vehicle pitch angle acceleration and the vertical acceleration of the sprung center of mass as a function of t, respectively; C represents the output matrix of the active suspension system; and D represents the direct transmission matrix of the active suspension system. S103, Discretize the continuous-time state equation and continuous-time output equation of the active suspension system to obtain the corresponding discrete-time state equation: The discrete-time output equation is: ; In the formula, x(k) represents the actual state vector of the model predictive control at the current discrete sampling time; u(k) represents the actual input vector of the active suspension system at the current discrete sampling time; y(k) represents the actual output vector of the active suspension system at the current discrete sampling time; A d B represents the discrete-time state matrix of the active suspension system. d This represents the discrete-time input matrix of the active suspension system; S104. Based on the discrete-time state equation and discrete-time output equation of the active suspension system, the constrained cost function J1 of the active suspension system is constructed. The expression for the constrained cost function J1 is: ； ； In the formula, N represents the prediction time domain length; j represents the prediction step number within the prediction time domain; y represents the output error vector of the active suspension system at step j+1; j This represents the predicted output vector of the active suspension system at step j+1; u j x represents the prediction input vector of the active suspension system at step j+1; j Q represents the predicted state vector of the active suspension system at step j+1; d R represents the weighted matrix of the output error of the active suspension system; d S represents the input weighting matrix of the active suspension system; d The cross-term weighted matrix of the active suspension system is represented; x0 represents the active suspension system state obtained at the start of the prediction time domain; x current This indicates the state of the active suspension system obtained at the current moment in the prediction time domain; Indicate u j The lower limit of the allowable range; Indicate u j The upper limit of the allowable range; S105, Solve for the input sequence of the active suspension system that minimizes J1. and take The first input vector As the real-time optimal solution for the output force of the actuators in the front and rear active suspension systems; in, ; express The (j+1)th input vector.

10. A distributed drive vehicle, characterized in that, It employs a vehicle comfort braking method that considers tire longitudinal stiffness perturbation as described in any one of claims 1-9.

Images