A lateral and longitudinal coordinated control method for emergency braking of vehicles on opposite sides of the road

Abstract

A method for coordinated lateral and longitudinal control of emergency braking of a vehicle on an open road belongs to the field of vehicle safety control technology. The purpose of the present invention is to provide a method for coordinated lateral and longitudinal control of emergency braking of a vehicle on an open road, which can prevent vehicle instability while reducing the braking distance by controlling the braking torque of the vehicle. The present invention builds a linear two-degree-of-freedom reference model, replaces the different adhesion coefficients on both sides of the open road with fitted adhesion coefficients, and takes into account the influence of the direct yaw torque obtained by the controller on the reference signal, and obtains lateral speed and yaw angular velocity reference signals that meet actual road conditions, and determines the optimal slip rate and expected acceleration based on the road adhesion coefficient, thereby obtaining the expected values ​​of wheel speed and longitudinal speed. The present invention integrates tracking of the expected acceleration into the objective function to effectively reduce the braking distance.

Description

Technical Field

[0001] The invention belongs to the technical field of vehicle safety control. Background Art

[0002] When a vehicle makes an emergency brake on an oncoming road, the adhesion coefficients on both sides of the road are different, and the wheel braking forces are different, which in turn generates a larger yaw moment, and the vehicle body will lean toward the track with greater braking force; when a vehicle makes an emergency brake on an icy or snowy road, the road adhesion coefficient is smaller, locking often occurs, and the braking distance increases. In severe cases, the vehicle becomes unstable and causes a traffic accident; when a vehicle makes an emergency brake, its longitudinal and lateral dynamics exhibit strongly coupled nonlinear characteristics, and the separate longitudinal or lateral motion control does not take into account the mutual influence between the two.

[0003] At present, the lateral and longitudinal coordinated control of vehicles during emergency braking on the oncoming road has the following problems:

[0004] 1. The indicators for evaluating vehicle lateral stability are mainly the tracking effect of the vehicle's lateral speed and yaw rate on their expected values. Most traditional control algorithms have a relatively simple setting of the expected values ​​and do not take into account the actual road conditions, which makes the design of the reference model unreasonable and thus affects the control performance of the controller.

[0005] 2. Although most of the current longitudinal control methods can control the slip rate, they do not take into account the safety constraints of the vehicle, and most traditional control algorithms only consider the longitudinal slip rate control or lateral stability of the vehicle. However, when the vehicle performs emergency braking on an oncoming road, the longitudinal force and lateral force of the tire are coupled in a nonlinear relationship, which will affect each other and cause a decrease in the control effect.

[0006] 3. The current slip rate control method mainly reduces the wheel braking pressure on high-adhesion roads to maintain a balance in the braking force of the wheels on both sides and prevent the vehicle from running off the track. However, this will reduce the utilization of the ground adhesion coefficient, reduce the braking force and thus increase the braking distance. Summary of the invention

[0007] The purpose of the present invention is to provide a lateral and longitudinal coordinated control method for emergency braking of an automobile on an oncoming road, which can prevent the vehicle from losing stability and reduce the braking distance by controlling the braking torque of the automobile.

[0008] The steps of the present invention are:

[0009] S1. Control on the open road:

[0010] ① Lateral and horizontal swing expected values:

[0011]

[0012]

[0013] Among them, β is the sideslip angle of the vehicle center of mass, γ is the yaw rate, δ is the steering wheel angle given by the driver, and v x represents the longitudinal velocity of the vehicle, μ1 and μ2 are the road adhesion coefficients on the left and right sides of the vehicle, M z is the direct yaw moment obtained based on the controller;

[0014] ΔT Br =T Brl -T Brr , ΔT Bf =T Bfl -T Bfr ;

[0015] At the tire adhesion limit, the lateral force must satisfy the following constraints: y |≤μg; when the sideslip angle is very small, the vehicle lateral acceleration a y It can be expressed as: y ≈γv x ; Use a fitting adhesion coefficient μ′ to fit the two adhesion coefficients so that the output of their transfer functions is the same, where μ′ = (μ1 + μ2) / 2; Combining the two equations, it can be seen that the ideal yaw rate should satisfy the following constraints: At the same time, the upper limit of the ideal center of mass sideslip angle is: Finally, the reference model with the fitted adhesion coefficient μ′ is obtained:

[0016]

[0017]

[0018] Derivation of γ ref and β ref The expression is as follows:

[0019]

[0020]

[0021] Where: n is the undamped frequency, ξ is the damping coefficient, τ β1 From δ to β ref The differential coefficient of the transfer function, τ γ1 From δ to γ ref The differential coefficient of the transfer function, τ γ2 M Z to γ ref The differential coefficient of the transfer function, G β1 From δ to βref The transfer function of the center of mass sideslip angle stabilization gain, G β2 M Z To β ref The transfer function of the center of mass sideslip angle stabilization gain, G γ1 From δ to γ ref The yaw rate stabilization gain of the transfer function, G γ2 M Z To β ref The yaw rate stabilization gain of the transfer function, K is the gradient coefficient, L = a + b represents the distance from the front axle to the rear axle. y / v x It can be seen that: yref =v x β ref ;

[0022] ②Vertical Expected Value

[0023] Longitudinal acceleration a x ≤μg; and It can be seen that xref (k)=(a xref +v y γ)T+v x (k-1), where v x (k) is the longitudinal velocity at the current moment, v x (k-1) is the longitudinal velocity at the previous moment, T is the sampling time; the optimal slip rate λ for different road surfaces ref =λ top , then the optimal angular velocity S2. NMPC-based lateral and longitudinal collaborative controller: Build a seven-degree-of-freedom vehicle dynamics model based on the Dugoff tire model: ① Seven-degree-of-freedom vehicle dynamics model

[0024] Considering the longitudinal motion, lateral motion, yaw motion and rotation of the four wheels of the vehicle, a seven-degree-of-freedom vehicle dynamics model is obtained:

[0025]

[0026]

[0027]

[0028]

[0029] Among them, v y is the lateral velocity of the vehicle, ωi is the wheel speed, T Ei and T Bi are the driving force and braking force applied to the wheels, respectively, x and F y They represent the longitudinal force and lateral force of the tire respectively, and the subscripts fl, fr, rl, and rr represent the left front, right front, left rear, and right rear wheels respectively;

[0030] ②Tire model

[0031] For longitudinal and lateral forces under combined transverse and longitudinal conditions

[0032]

[0033]

[0034]

[0035]

[0036] Among them, λ i is the longitudinal slip rate of the wheel, F z,i is the tire vertical load, α i is the wheel slip angle;

[0037] The vertical wheel load is:

[0038]

[0039]

[0040]

[0041]

[0042] The wheel slip rate is Among them, the longitudinal speeds of the four wheels are:

[0043] v xfl =(v x -Lγ)cosδ+(v y +aγ)sinδ (7a)

[0044] v xfr =(v x +Lγ)cosδ+(v y +aγ)sinδ (7b)

[0045]

[0046]

[0047] The tire's side slip angle α is:

[0048]

[0049]

[0050] ③ Prediction model

[0051] Normalize the state quantity x, that is,

[0052] x=[v x / v xmax v y / v ymax γ / γ max ω1 / ω max ω2 / ω max ω3 / ω max ω4 / ω max ] T ,in

[0053] v xmax ,v ymax ,γ max ,ω max are the upper limits of the vehicle longitudinal velocity, lateral velocity, yaw rate and wheel speed respectively; the control amount u=[ΔT bfr / ΔT bmax ΔT bfl / ΔT bmax ΔT brr / ΔT bmax ΔT brl / ΔT bmax ] T , where ΔT bmax Indicates the maximum value of additional braking torque;

[0054] ④Objective function and constraints

[0055] Define cost functions as J2 and J3; define cost function as J1; define cost functions J4 and J5:

[0056]

[0057]

[0058]

[0059]

[0060]

[0061] The control variable braking torque is constrained as follows:

[0062] -1≤u(k+i-1)≤1,i=1,2,…,N (10a)

[0063] The state quantity constraints are:

[0064] -1≤x2(k+i)≤1,i=1,2,…,N (10b)

[0065] The optimization problem is described as:

[0066] min J = min(Γ ax J1+Γ β J2+Γ γ J3+Γ λ1 J4+Γ λ2 J5)

[0067] st(4),(10)

[0068] Among them, Γ ax is the weight coefficient of acceleration, Γ γ is the weight coefficient of the yaw rate, Γ β is the weight coefficient of the sideslip angle of the center of mass, Γ λ1 is the weight coefficient of the left wheel slip rate of the vehicle, Γ λ2 is the weight coefficient of the right wheel slip rate of the vehicle.

[0069] The beneficial effects of the present invention are:

[0070] 1. The present invention adopts a two-degree-of-freedom reference model designed for split road surfaces to derive the vehicle lateral velocity and yaw rate reference signals as expected values. When the vehicle is differentially braked, the direct yaw moment M obtained by the controller is Z will affect the reference model; at the same time, when the road adhesion coefficient is different, the lateral force that the road can provide to the tire is also different. Different from the traditional reference model, the reference model designed by the present invention takes into account M Z The influence of different adhesion coefficients on both sides of the vehicle on the reference model makes the tracking signal more reasonable;

[0071] 2. Most traditional control algorithms do not consider the mutual influence between the longitudinal force and lateral force of the tire during emergency braking of the vehicle, and ignore the nonlinear coupling relationship between the lateral force and the longitudinal slip rate. The present invention takes into account multiple control objectives in the lateral and longitudinal directions, and takes into account control constraints and safety constraints, which can ultimately improve the stability and maneuverability of the vehicle during emergency braking on the open road.

[0072] 3. The current slip rate control reduces the wheel braking pressure on the high-adhesion road surface to maintain the balance of the braking force on both sides of the wheel, which will increase the braking distance. The present invention integrates the tracking of the expected acceleration into the objective function to effectively reduce the braking distance. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a flowchart of the present invention;

[0074] Figure 2a is a comparison diagram of the yaw rate reference signal of the present invention;

[0075] Figure 2b is a comparison diagram of the reference signal of the center of mass sideslip angle of the present invention;

[0076] Figure 2a and Figure 2b The solid line in the middle is the linear two-degree-of-freedom reference signal, and the double-strike line is the direct yaw moment M obtained based on the controller when only vehicle differential braking is considered. Z The reference signal obtained, the dotted line is considered while M Z and reference signals of different adhesion coefficients on both sides of the split road;

[0077] Figure 3a It is a reference signal comparison diagram when the yaw angular velocity is fitted to the adhesion coefficient of the present invention;

[0078] Figure 3b It is a comparison diagram of reference signals when the center of mass sideslip angle is fitted to the adhesion coefficient of the present invention;

[0079] Figure 3a and Figure 3b The solid line is the one that takes M into consideration at the same time. Z and the reference signals of different adhesion coefficients on both sides of the split road. The double-line is the reference signal obtained by fitting the adhesion coefficient.

[0080] Figure 4 is a schematic diagram of a vehicle dynamics model of the present invention;

[0081] Figure 5a It is a speed comparison diagram of the left front wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied in a straight line working condition;

[0082] Figure 5b It is a speed comparison diagram of the right front wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied in a straight line working condition;

[0083] Figure 5c It is a speed comparison diagram of the left rear wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied in a straight line working condition;

[0084] Figure 5d It is a speed comparison diagram of the right rear wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied in a straight line working condition;

[0085] Figure 5a , Figure 5b , Figure 5c and Figure 5dThe solid line in the middle is the wheel speed when the vehicle does not apply the control strategy, and the double-strike line is the wheel speed when the control strategy is applied;

[0086] Figure 6a is a comparison diagram of the slip rate of the left front wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under straight-line working conditions;

[0087] Figure 6b is a comparison diagram of the slip rate of the right front wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under straight-line working conditions;

[0088] Figure 6c is a comparison diagram of the slip rate of the left rear wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under straight-line working conditions;

[0089] Figure 6d is a comparison diagram of the slip rate of the right rear wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under straight-line working conditions;

[0090] Figure 6a , Figure 6b , Figure 6c and Figure 6d The solid line in the middle is the vehicle slip rate when no control strategy is applied to the vehicle, and the double-strike line is the vehicle slip rate when the control strategy is applied;

[0091] Figure 7a is a comparison diagram of the yaw angular velocity of the vehicle of the present invention under straight-line working conditions;

[0092] Figure 7b is a comparison diagram of the center of mass side slip angle of the vehicle of the present invention under straight line working conditions;

[0093] Figure 7a and Figure 7b The solid line is the reference value, the double dashed line is the yaw rate and sideslip angle of the vehicle when it is not controlled, and the dashed line is the yaw rate and sideslip angle of the vehicle when the control strategy is applied;

[0094] Figure 8 is the control quantity of the vehicle of the present invention under straight-line working conditions;

[0095] Fig. 9 is the acceleration of the vehicle of the present invention under straight-line conditions, wherein the solid line is the acceleration of the vehicle when no control strategy is applied, and the double-strike line is the acceleration of the vehicle when the control strategy is applied;

[0096] Fig.10 is the trajectory of the vehicle of the present invention under straight line conditions, wherein the solid line is the reference trajectory, the double-dash line is the trajectory of the vehicle when no control strategy is applied, and the dot-dash line is the trajectory of the vehicle when the control strategy is applied;

[0097] Fig.11ais the speed of the left front wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under the J-turn condition;

[0098] Fig.11b is the speed of the right front wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under the J-turn condition;

[0099] Fig.11c is the speed of the left rear wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under the J-turn condition;

[0100] Fig.11d is the speed of the right rear wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under the J-turn condition;

[0101] Fig.11a , Fig.11b , Fig.11c and Fig.11d The solid line in the middle is the wheel speed when the vehicle does not apply the control strategy, and the double-strike line is the wheel speed when the control strategy is applied;

[0102] Fig.12a is the slip rate of the left front wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under the J-turn condition;

[0103] Figure 12b is the slip rate of the right front wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under the J-turn condition;

[0104] Fig.12c is the slip rate of the left rear wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under the J-turn condition;

[0105] Fig.12d is the slip rate of the right rear wheel of the vehicle of the present invention when the control strategy is applied and when the control strategy is not applied under the J-turn condition;

[0106] Fig.12a , Figure 12b , Fig.12c and Fig.12d The solid line in the middle is the vehicle slip rate when no control strategy is applied to the vehicle, and the double-strike line is the vehicle slip rate when the control strategy is applied;

[0107] Fig.13a is the yaw angular velocity of the vehicle in the J-turn condition according to the present invention;

[0108] Fig.13b is the center of mass sideslip angle of the vehicle in the J-turn condition of the present invention;

[0109] Fig.13a and Fig.13bThe solid line is the reference value, the double dashed line is the yaw rate and sideslip angle of the vehicle when it is not controlled, and the dashed line is the yaw rate and sideslip angle of the vehicle when the control strategy is applied;

[0110] Fig.14 is the control quantity of the vehicle of the present invention under the J-turn condition;

[0111] Fig.15 is the acceleration of the vehicle of the present invention under the J-turn condition, wherein the solid line is the acceleration of the vehicle when no control strategy is applied, and the double-strike line is the acceleration of the vehicle when the control strategy is applied;

[0112] Fig.16 It is the trajectory of the vehicle of the present invention under the J-turn condition, wherein the solid line is the reference trajectory, the double-dash line is the trajectory of the vehicle when no control strategy is applied, and the dot-dash line is the trajectory of the vehicle when the control strategy is applied. DETAILED DESCRIPTION

[0113] In order to solve the problem of poor stability of vehicles during emergency braking on the open road, the present invention proposes a lateral and longitudinal coordinated control method based on model predictive control (MPC), which enables the vehicle to track the reference signal well when braking on the open road, and effectively reduces the braking distance. A linear two-degree-of-freedom reference model is built, and the different adhesion coefficients on both sides of the open road are replaced by the fitted adhesion coefficient. At the same time, the direct yaw moment M obtained by the controller is converted into Z Taking the influence of the reference signal into consideration, the lateral velocity and yaw rate reference signals that meet the actual road conditions are obtained, and the optimal slip rate and expected acceleration are determined by the road adhesion coefficient, thereby obtaining the expected values ​​of wheel speed and longitudinal speed. A seven-degree-of-freedom vehicle dynamics model based on the Dugoff tire model is built to describe the vehicle kinematic characteristics and predict the future state of the system. Under the MPC framework, a vehicle lateral and longitudinal collaborative controller is designed to integrate requirements such as suppressing the sideslip angle of the center of mass, tracking the yaw rate, reducing the slip rate, and tracking the expected longitudinal speed into the objective function. At the same time, the safety constraints of the vehicle are considered, and the braking torque of the vehicle is controlled to prevent vehicle instability while reducing the braking distance.

[0114] In order to solve the problems of the prior art, the present invention is implemented by adopting the following technical solutions:

[0115] The method for coordinated lateral and longitudinal control of emergency braking of a vehicle on an on-road surface comprises the following steps:

[0116] Step 1: Use the simulation software CarSim to obtain an emergency braking vehicle model and provide vehicle status information in real time;

[0117] Step 2: Control demand analysis. Analysis of lateral and yaw expected values: When the vehicle is differentially braked on the open road, the direct yaw moment M obtained by the controller isZ Different adhesion coefficients on both sides of the road will affect the reference signal. The present invention designs a reference model that conforms to the actual road conditions for emergency braking of the vehicle on the open road to determine the ideal state of the vehicle's lateral and swaying. Longitudinal expected value design: According to the classification of braking intensity levels and the constraint of the vehicle's braking force on the ground, a reasonable acceleration expected value is designed, and the expected value of the longitudinal speed is obtained from the relationship between acceleration and speed; according to the adhesion coefficient of the open road, the optimal slip rate can be obtained, thereby obtaining the expected wheel speed;

[0118] Step 3: NMPC controller design. Based on the Dugoff tire model, a seven-degree-of-freedom vehicle dynamics model is built. Under the MPC framework, a vehicle lateral and longitudinal coupling collaborative optimization controller is designed to coordinate multiple lateral and longitudinal control objectives, while considering control constraints and state constraints, solve the optimization problem, and obtain the optimal control law to improve the safety and maneuverability of the vehicle in emergency braking on the open road and reduce the braking distance.

[0119] The present invention is described in detail below in conjunction with the accompanying drawings:

[0120] The structure diagram of the collaborative control method of the present invention is as follows: Figure 1 As shown in the figure, the purpose of the control demand analysis is to obtain the expected values ​​of the longitudinal motion, lateral motion, and yaw motion of the vehicle; the input of the NMPC controller is the expected yaw angular velocity, expected lateral velocity, expected wheel speed, expected longitudinal velocity, and the output measurement value of the controlled object, and the output is the additional braking torque of the four tires; the actual braking torque of the vehicle can be obtained by adding the additional braking torque solved by the controller and the pedal braking torque given by the driver. The NMPC controller is built in MATLAB / Simulink; the controlled object is a vehicle model constructed using CarSim.

[0121] The control target of the present invention is to obtain a control target that is more suitable for open roads based on actual road conditions. Based on the dynamic model of the vehicle with seven degrees of freedom, namely longitudinal motion, lateral motion, yaw motion and rotation of four wheels, a vehicle lateral and longitudinal coordinated controller is designed. Requirements such as suppressing the sideslip angle of the center of mass, tracking the yaw angular velocity, reducing the slip rate, and tracking the expected longitudinal speed are integrated into the objective function. At the same time, the control constraints and state constraints of the vehicle are taken into consideration. By controlling the braking torque of the vehicle, the vehicle's maneuverability is improved, and the vehicle is prevented from becoming unstable while reducing the braking distance.

[0122] The present invention provides a set of joint simulation models based on the above operating principles and operating processes, and its construction and operating process are as follows:

[0123] 1. Software selection

[0124] The simulation models of the controller and the controlled object of the control system are built by software MATLAB / Simulink and CarSim, respectively. The software versions are MATLAB R2020a and CarSim 2019.0, respectively, and the simulation step is 0.001s. The present invention completes the construction of the simulation model of the controller through MATLAB / Simulink, and completes the operation of the controller in the control system through Simulink programming; CarSim software is a commercial simulation software, the main function of which is to provide a high-fidelity vehicle dynamics model, replace the real car as the implementation object of the control method in the simulation experiment, and provide a simulation environment for the open road.

[0125] 2. Joint simulation settings

[0126] To achieve joint simulation between MATLAB / Simulink and CarSim, you must first set the working path in CarSim to Simulink Model, then add the output interface module, add the set vehicle model to Simulink in the form of S-function, and run Simulink to achieve joint simulation and communication between the two. If the model structure or parameter settings in CarSim are modified, you need to resend it.

[0127] 3. Construction of electric vehicle model in joint simulation software

[0128] The CarSim electric vehicle model is mainly composed of the body, transmission system, steering system, brake system, tire, suspension, aerodynamics, working condition configuration and other systems. Its braking torque input uses IMP_MYUSM_L1, IMP_MYUSM_L2, IMP_MYUSM_R1, IMP_MYUSM_R2, and the electric vehicle parameters are shown in Table 1.

[0129] 4. The control principle of the vehicle's lateral and longitudinal stability under extreme working conditions of the present invention

[0130] The control goal of the present invention is to improve the lateral and longitudinal stability of the vehicle during emergency braking on the split road. The main design process of the control method is described as follows: First, the high-fidelity vehicle dynamics model of the vehicle is obtained using the simulation software CarSim; second, a linear two-degree-of-freedom reference model is built, and the different adhesion coefficients on both sides of the split road are replaced by the fitted adhesion coefficients, while considering the direct yaw moment M obtained based on the controller Z, a two-degree-of-freedom reference model with a more reasonable control target is obtained, and the expected values ​​of yaw rate and lateral speed are obtained through the reference model. The optimal slip rate and expected acceleration are obtained from the road adhesion coefficient, thereby obtaining the expected values ​​of wheel speed and longitudinal speed; then, a seven-degree-of-freedom vehicle dynamics model based on the Dugoff tire model is built to describe the vehicle kinematic characteristics and predict the future state of the system; finally, under the MPC framework, a vehicle lateral and longitudinal collaborative controller is designed, which integrates the requirements of suppressing the sideslip angle of the center of mass, tracking the yaw rate, reducing the slip rate, and tracking the expected longitudinal speed into the objective function, and controls the braking torque of the vehicle to ensure the lateral and longitudinal stability of the vehicle while reducing the braking distance.

[0131] Table 1 Electric vehicle parameters

[0132] symbol Physical Description Value / Unit m Vehicle quality 1286 / kg <![CDATA[R ω ]]> Wheel radius 0.308 / m a Distance from vehicle center of mass to front axle 1 / m b Distance from vehicle center of mass to rear axle 1.6 / m d Left and right wheel track 1.55 / m <![CDATA[k f0 ]]> Front tire cornering stiffness constant <![CDATA[53353 / N·rad -1 ]]> <![CDATA[k r0 ]]> Rear tire cornering stiffness constant <![CDATA[64118 / N·rad -1 ]]> <![CDATA[I z ]]> The vehicle's moment of inertia about the z-axis <![CDATA[1970 / kg·m -2 ]]> B Left and right wheel track 1.1 / m h Center of mass height 0.52 / m w Half car width 0.773 / m <![CDATA[I ω ]]> Wheel moment of inertia <![CDATA[10 / kg·m 2 ]]> <![CDATA[C σ ]]> Tire longitudinal stiffness <![CDATA[68000 / N·rad -1 ]]> <![CDATA[C α ]]> Tire lateral stiffness <![CDATA[48000 / N·rad -1 ]]> g Gravity coefficient <![CDATA[9.8 / m·s -2 ]]>

[0133] The specific steps of the control method of the present invention are introduced below:

[0134] The method for coordinated lateral and longitudinal control of emergency braking of a vehicle on an on-road surface comprises the following steps:

[0135] Step 1: Use the simulation software CarSim to obtain a car model. In CarSim, the car model is set to simulate the real controlled object. Its main function is to provide real-time information about the vehicle's status, read the vehicle's motion state parameters into Simulink, build a simulation condition for driving on a split road, and simulate the kinematic characteristics of the actual vehicle.

[0136] Step 2: Analyze the control requirements for the split road. Use the fitted adhesion coefficient to replace the different adhesion coefficients on both sides of the split road, and consider the direct yaw moment M obtained based on the controller. Z , a two-degree-of-freedom reference model with a more reasonable control target is obtained, and the expected values ​​of lateral velocity and yaw angular velocity are obtained through the reference model, the ideal lateral and yaw motion states of the vehicle are determined, and the optimal slip rate is determined according to the adhesion coefficient of the open road surface. The expected acceleration is obtained from the division of braking intensity levels and the constraint of braking force by the ground, thereby obtaining the expected wheel speed and expected longitudinal speed.

[0137] ① Analysis of lateral and horizontal swing expected values

[0138] When studying the stability of a car, a linear two-degree-of-freedom model is often used to provide a reference signal. This model ignores the effects of air resistance and the steering system, assumes that the transmission system is linear, and assumes that the vehicle has no vertical movement. At this time, the road adhesion coefficient is represented by μ. However, when the vehicle is differentially braked, the direct yaw moment M obtained based on the controller zIt will affect the above two-degree-of-freedom reference model. At the same time, when the road adhesion coefficient is different, the lateral force that the road can provide to the tire is also different. At this time, it is necessary to consider M z The influence of different adhesion coefficients on both sides of the road on the reference model is shown in Figure 2. At this time, the different adhesion coefficients on both sides of the road are represented by μ1 and μ2. The equation is as follows:

[0139]

[0140]

[0141] Among them, β is the sideslip angle of the vehicle center of mass, γ is the yaw rate, δ is the steering wheel angle given by the driver, and v x represents the longitudinal velocity of the vehicle, μ1 and μ2 are the road adhesion coefficients on the left and right sides of the vehicle, M z is the direct yaw moment obtained based on the controller, ΔT Br =T Brl -T Brr , ΔT Bf =T Bfl -T Bfr .

[0142] As shown in Figure 2, in the linear two-degree-of-freedom reference model and considering only M z In the reference model, the road adhesion coefficient is set to μ = 0.85, while considering M z In the reference model of the split road, μ1 = 0.85 and μ2 = 0.25. From the simulation results, it can be seen that the direct yaw moment M obtained based on the controller z Different adhesion coefficients do have an impact on the reference signal, and the impact is greater when the vehicle is turning. Therefore, in order to make the control target more reasonable, both of the above situations need to be taken into account.

[0143] In actual situations, the tire force will reach saturation, and the lateral force must satisfy the following constraints under the tire adhesion limit: y |≤μg. When the sideslip angle is very small, the vehicle lateral acceleration a y It can be expressed as: y ≈γv x When considering the situation where the two adhesion coefficients on both sides of the split road are different, consider using a fitting adhesion coefficient μ′ to fit the two adhesion coefficients so that the output of their transfer functions is the same, where μ′=(μ1+μ2) / 2. Combining the two equations, it can be seen that the ideal yaw rate should satisfy the following constraints: At the same time, the upper limit of the ideal center of mass sideslip angle is: Finally, the reference model with the fitted adhesion coefficient μ′ can be obtained:

[0144]

[0145]

[0146] Considering the frequency characteristics, the response of the vehicle reference motion state to the front wheel steering angle δ can be obtained from the formula, and γ can be derived ref and β ref The expression is as follows:

[0147]

[0148]

[0149] Where: n is the undamped frequency, ξ is the damping coefficient,

[0150] τ β1 From δ to β ref The differential coefficient of the transfer function, τ γ1 From δ to γ ref The differential coefficient of the transfer function, τ γ2 M Z to γ ref The differential coefficient of the transfer function, G β1 From δ to β ref Transfer function of the center of mass sideslip angle stabilization gain, G β2 M Z To β ref Transfer function of the center of mass sideslip angle stabilization gain, G γ1 From δ to γ ref The yaw rate stabilization gain of the transfer function, G γ2 M Z To β ref The yaw rate stabilization gain of the transfer function, K is the gradient coefficient, L = a + b represents the distance from the front axle to the rear axle. y / v x It can be seen that: yref =v x β ref .

[0151] As shown in Figure 3, considering M ZIn the reference model of the split road, μ1=0.85, μ2=0.4, and the adhesion coefficient μ'=0.625 in the reference model of the fitted adhesion coefficient. The images of the sideslip angle and yaw rate obtained by using the fitted adhesion coefficient are consistent with those of the M Z The image obtained from the reference model of the opposing road surface completely overlaps.

[0152] ② Longitudinal expected value analysis Determine the optimal slip rate and expected deceleration based on the road adhesion coefficient: a x As a more important variable in the control system, it not only affects the stability and ride comfort of the vehicle, but also affects the braking distance. The longitudinal acceleration is divided into four levels according to the braking intensity, as shown in Table 2; at the same time, the vehicle braking force is constrained by the ground adhesion coefficient, and the longitudinal acceleration a x ≤μg, to avoid vehicle skidding during braking, the expected acceleration value is set to different values ​​according to the road conditions, as shown in Table 3.

[0153] Depend on and It can be seen that xref (k)=(a xref +v y γ)T+v x (k-1), where v x (k) is the longitudinal velocity at the current moment, v x (k-1) is the longitudinal velocity at the previous moment, and T is the sampling time.

[0154] Table 2 Classification of longitudinal acceleration levels

[0155]

[0156] Table 3 Maximum longitudinal acceleration under different road conditions

[0157]

[0158] Different adhesion coefficients correspond to different optimal slip rates. The μ-λ curve model proposed by Burckhardt is as follows:

[0159]

[0160] Given three parameters c1, c2, c3 under typical road conditions, mathematical analysis of the above formula can yield the optimal slip rate λ of the road surface. top and the peak road adhesion coefficient μ max The values ​​of the parameters of the exponential function model under different road conditions are shown in Table 4. The optimal slip rate λ of different road surfaces ref =λ top , then the optimal angular velocity

[0161] Table 4 Parameter values ​​of the exponential function model under different road conditions

[0162]

[0163] Step 3: Design of lateral and longitudinal coordinated controller based on NMPC. Build a seven-degree-of-freedom vehicle dynamics model based on the Dugoff tire model to describe the vehicle kinematic characteristics and predict the future state of the system. Under the MPC framework, design the vehicle lateral and longitudinal coordinated controller, integrate the requirements of suppressing the sideslip angle of the center of mass, tracking the yaw rate, reducing the slip rate, and tracking the expected longitudinal speed into the objective function, and optimize the solution to obtain the optimal control quantity.

[0164] ① Seven-degree-of-freedom vehicle dynamics model

[0165] The schematic diagram of the vehicle dynamics model of the present invention is as follows Figure 4 As shown, ignoring the influence of air resistance and considering the longitudinal motion, lateral motion, yaw motion and rotation of the four wheels of the vehicle, a seven-degree-of-freedom vehicle dynamics model is obtained:

[0166]

[0167]

[0168]

[0169]

[0170] Among them, v y is the lateral velocity of the vehicle, ω i is the wheel speed, T Ei and T Bi are the driving force and braking force applied to the wheels, respectively, x and F y They represent the longitudinal force and lateral force of the tire respectively, and the subscripts fl, fr, rl, and rr represent the left front, right front, left rear, and right rear wheels respectively.

[0171] ②Tire model

[0172] The longitudinal stiffness and lateral stiffness of the Dugoff tire model are very different, which allows the model to have independent tire stiffness values ​​in the longitudinal and lateral directions. It can be used to solve the longitudinal force and lateral force under the combined transverse and longitudinal working conditions. The expression is:

[0173]

[0174]

[0175]

[0176]

[0177] Among them, λ i is the longitudinal slip rate of the wheel, F z,i is the tire vertical load, α i is the wheel slip angle.

[0178] The wheel vertical load expression is:

[0179]

[0180]

[0181]

[0182]

[0183] The wheel slip ratio expression is:

[0184] Among them, the longitudinal speeds of the four wheels are:

[0185] v xfl =(v x -Lγ)cosδ+(v y +aγ)sinδ (7a)

[0186] v xfr =(v x +Lγ)cosδ+(v y +aγ)sinδ (7b)

[0187]

[0188]

[0189] The calculation of the tire's side slip angle α is as follows:

[0190]

[0191]

[0192] ③ The prediction model can be obtained by combining the seven-degree-of-freedom vehicle dynamics model and the tire model to obtain a prediction model for controller design. Its state quantity x is composed of the vehicle longitudinal speed, lateral speed, yaw angular velocity and four wheel speeds. Since the different orders of magnitude of the system state are not conducive to comparison and weighting, in order to better adjust the weighting factor and obtain better control performance, the state quantity is normalized, that is, x = [v x / v xmax vy / v ymax γ / γ max ω1 / ω max ω2 / ω max ω3 / ω max ω4 / ω max ] T , where v xmax ,v ymax ,γ max ,ω max are the upper limits of the vehicle longitudinal velocity, lateral velocity, yaw rate and wheel speed respectively; the control amount u=[ΔT bfr / ΔT bmax ΔT bfl / ΔT bmax ΔT brr / ΔT bmax ΔT brl / ΔT bmax ] T , where ΔT bmax Indicates the maximum value of the additional braking torque.

[0193] ④Objective function and constraints

[0194] According to the control objectives mentioned above, the vehicle must first ensure stability, so it should track β ref and γ ref , define the cost functions as J2 and J3; in order to increase the utilization of the ground adhesion coefficient and shorten the braking distance, define the cost function as J1; in order to prevent wheel locking and reduce the wheel slip rate, define cost functions J4 and J5.

[0195]

[0196]

[0197]

[0198]

[0199]

[0200] Since the actuator will be saturated, the control variable braking torque is constrained as follows:

[0201] -1≤u(k+i-1)≤1,i=1,2,…,N (10a)

[0202] Too large a lateral speed will reduce maneuverability, so the state constraint is:

[0203] -1≤x2(k+i)≤1,i=1,2,…,N (10b)

[0204] Finally, the optimization problem can be described as:

[0205]

[0206] st(4),(10)

[0207] in, is the weight coefficient of acceleration, Γ γ is the weight coefficient of the yaw rate, Γ β is the weight coefficient of the sideslip angle of the center of mass, Γ λ1 is the weight coefficient of the left wheel slip rate of the vehicle, Γ λ2 is the weight coefficient of the right wheel slip rate of the vehicle.

[0208] The effectiveness of the control method of the present invention is verified by the following simulation experiments:

[0209] In order to verify the effectiveness of the control method described in the present invention, a simulation experiment was designed in the joint simulation environment of CarSim and MATLAB / Simulink. This paper selects the straight line working condition and J-turn working condition with different adhesion conditions on the left and right sides. The adhesion coefficient on the left side is μ1=0.4, and the adhesion coefficient on the right side is μ2=0.85. The vehicle speed is set to 60km / h and the pedal brake pressure is 1MPa. The model of the controlled vehicle is E-Class SUV, the controller prediction time domain N=30, and the sampling time T s =0.01s, the parameters and weight coefficients used in the simulation experiment are shown in Table 4. The Casadi toolbox is used to solve the above optimization problem, and finally the control sequence can be obtained.

[0210] As shown in FIG5 , when the control strategy is not applied, the vehicle decelerates to 0 in about 10.2 s, while after the control strategy is applied, the vehicle decelerates to 0 in about 8.8 s, and the vehicle can decelerate to 0 faster.

[0211] As shown in Figure 6, the vehicle's slip rate can be well controlled within 0.2, and the vehicle is in a rolling state to prevent the vehicle from sliding sideways.

[0212] As shown in Figure 7, the vehicle's yaw rate and center of mass sideslip angle can track the reference value very well.

[0213] like Fig. 9 As shown, the vehicle acceleration with the control strategy applied is greater than the vehicle acceleration without the control strategy applied.

[0214] Table 4 Simulation experiment parameters

[0215]

[0216]

[0217] like Fig.10 As shown, the longitudinal displacement of the vehicle to which the control strategy is applied is reduced by 8m relative to the longitudinal displacement of the vehicle to which the control strategy is not applied. It can be seen that the control strategy of the present invention can effectively reduce the braking distance of the vehicle while ensuring the lateral and longitudinal stability of the vehicle.

[0218] As shown in Figure 11, when the control strategy is not applied, the vehicle's speed decelerates to 0 in about 9 seconds. After the control strategy is applied, the vehicle decelerates to 0 in about 8.6 seconds. The vehicle can decelerate to 0 faster.

[0219] As shown in Figure 12, when the control strategy is not applied, the wheel slip rate sometimes exceeds 0.2. After the control strategy is applied, the vehicle slip rate can be controlled within 0.2, and the controller can quickly reduce the slip rate to ensure that the tire is in a rolling state, prevent the vehicle from slipping longitudinally, and improve the safety of the vehicle.

[0220] As shown in Figure 13, the vehicle's yaw rate and center of mass sideslip angle can track the reference value very well. When no control strategy is applied, the vehicle's yaw rate and center of mass sideslip angle reference values ​​are quite different, the maximum yaw rate can reach 1 rad / s, and the maximum absolute value of the center of mass sideslip angle can reach 0.75 rad; under the action of the controller, the maximum absolute value of the controlled vehicle's yaw rate is 0.3 rad / s, and the center of mass sideslip angle is close to 0, indicating that the controller can track the yaw rate very well, suppress the center of mass sideslip angle, and improve vehicle stability.

[0221] like Fig.15 As shown, the vehicle acceleration with the control strategy applied is greater than the vehicle acceleration without the control strategy applied.

[0222] like Fig.16 As shown, the longitudinal displacement of the vehicle to which the control strategy is applied is reduced by 1 m relative to the displacement of the vehicle to which the control strategy is not applied. It can be seen that the control strategy of the present invention can reduce the braking distance of the vehicle while ensuring the lateral and longitudinal stability of the vehicle.

[0223] Through verification of straight-line working conditions and J-turn working conditions, the lateral and longitudinal stability coordinated control method of the present invention can effectively improve the lateral and longitudinal stability of the vehicle during emergency braking on an oncoming road, improve the handling performance, reduce the braking distance, and ensure driving safety.

[0224] The symbols involved in the present invention are as follows:

[0225] MPC: Model Predictive Control; Z : direct yaw moment; m: vehicle mass; R ω: wheel radius; a: distance from the center of mass of the vehicle to the front axle; b: distance from the center of mass of the vehicle to the rear axle; L: distance from the front axle to the rear axle of the vehicle; k f0 : Front wheel tire cornering stiffness constant; k r0 : rear wheel tire cornering stiffness constant; I z : The moment of inertia of the vehicle around the z axis; B: The wheelbase of the left and right wheels; h: The height of the center of mass; w: Half the width of the vehicle; I ω : Wheel moment of inertia; C σ : Tire longitudinal stiffness; C α : tire lateral stiffness; g: gravity coefficient; μ: road adhesion coefficient; μ1: road adhesion coefficient on the left side; μ2: road adhesion coefficient on the right side; β: vehicle center of mass sideslip angle; γ: yaw rate; δ: steering wheel angle given by the driver; a y : vehicle lateral acceleration; μ': fitting adhesion coefficient; v x : vehicle longitudinal velocity; v xfl : longitudinal speed of the left front wheel of the vehicle; v xfr : longitudinal speed of the right front wheel of the vehicle; v xrl : longitudinal speed of the left rear wheel of the vehicle; v xrr : longitudinal speed of the right rear wheel of the vehicle; T Bfl : braking torque of the left front wheel; T Bfr : braking torque of the right front wheel; T Brl : braking torque of the left rear wheel; T Brr : braking torque of the right rear wheel; γ ref : yaw rate reference value; β ref : Reference value of the sideslip angle at the center of mass; ω n : undamped frequency; ξ: damping coefficient; τ β1 : δ to β ref The differential coefficient of the transfer function; τ γ1 : δ to γ ref The differential coefficient of the transfer function; τ γ2 :M Z to γ ref Differential coefficient of transfer function; G β1 : δ to β ref Transfer function of the center of mass sideslip angle stability gain; G β2 :M Z To β ref Transfer function of the center of mass sideslip angle stability gain; G γ1 : δ to γ ref The yaw rate stabilization gain of the transfer function; G γ2 :M Z To β ref The yaw rate stabilization gain of the transfer function; K: gradient coefficient; v x (k): longitudinal velocity at the current moment; v x(k-1): the longitudinal velocity at the previous moment; v yref : lateral speed reference value; T: sampling time; c1, c2, c3: three parameters under typical road conditions; λ ref : optimal slip ratio; λ top : Slip rate corresponding to the peak adhesion coefficient of the road surface; μ max : peak road adhesion coefficient; ω iref : optimal angular velocity; v y : lateral speed of vehicle; ω fl : left front wheel speed; ω fr : right front wheel speed; ω rl : left rear wheel speed; ω rr : right rear wheel speed; T Efl : driving force applied to the left front wheel of the vehicle; T Efr : driving force applied to the right front wheel of the vehicle; T Erl : driving force applied to the left rear wheel of the vehicle; T Err : driving force applied to the right rear wheel of the vehicle; T Bfl : braking force applied to the left front wheel of the vehicle; T Bfr : braking force applied to the right front wheel of the vehicle; T Brl : braking force applied to the left rear wheel of the vehicle; T Brr : braking force applied to the right rear wheel of the vehicle; F xfl : longitudinal force of the left front tire of the vehicle; F xfr : longitudinal force of the right front tire of the vehicle; F xrl : longitudinal force of the left rear tire of the vehicle; F xrr : longitudinal force of the right rear tire of the vehicle; F yfl : lateral force of the left front tire of the vehicle; F yfr : lateral force of the right front tire of the vehicle; F yrl : lateral force of the left rear tire of the vehicle; F yrr : lateral force of the right rear tire of the vehicle; λ fl : longitudinal slip rate of the left front wheel of the vehicle; λ fr : longitudinal slip rate of the right front wheel of the vehicle; λ rl : longitudinal slip rate of the left rear wheel of the vehicle; λ rr : longitudinal slip rate of the right rear wheel of the vehicle; F zfl : Vertical load on the left front tire of the vehicle; F zfr : Vertical load on the right front tire of the vehicle; F zrl : Vertical load on the left rear tire of the vehicle; F zrr : Vertical load on the right rear tire of the vehicle; α fl : The left front wheel slip angle of the vehicle; α fr : The right front wheel slip angle of the vehicle; α rl: The left rear wheel slip angle of the vehicle; α rr : Vehicle right rear wheel slip angle; Γ ax : weight coefficient of acceleration; Γ γ : weight coefficient of yaw rate; Γ β : weight coefficient of the sideslip angle of the center of mass; Γ λ1 : Weight coefficient of the vehicle's left wheel slip rate; Γ λ2 : Weight coefficient of the vehicle's right wheel slip rate; v xmax : upper limit of vehicle longitudinal speed; v ymax : upper limit of tire lateral speed; γ max : upper limit of yaw rate; ω max : Upper limit of wheel speed; ΔT bmax : upper limit of additional braking torque; Γ ax : acceleration weight in NMPC; Γ γ : yaw rate weight in NMPC; Γ β : weight of the sideslip angle of the center of mass in NMPC; Γ λ1 : The slip rate weight of the left side of the vehicle in NMPC; Γ λ2 : The slip rate weight of the right side of the vehicle in NMPC.

Claims

1. A method for coordinated lateral and longitudinal control of emergency braking of a vehicle on an open road, characterized in that: The steps are: S1. Control on the open road: ① Lateral and horizontal swing expected values: Among them, β is the sideslip angle of the vehicle center of mass, γ is the yaw rate, δ is the steering wheel angle given by the driver, and v x represents the longitudinal velocity of the vehicle, μ1 and μ2 are the road adhesion coefficients on the left and right sides of the vehicle, M z is the direct yaw moment obtained based on the controller; ΔT Br =T Brl -T Brr ,ΔT Bf =T Bfl -T Bfr ; At the tire adhesion limit, the lateral force must satisfy the following constraints: y |≤μg; when the sideslip angle is very small, the vehicle lateral acceleration a y It can be expressed as: y ≈γv x ; Use a fitting adhesion coefficient μ' to fit the two adhesion coefficients so that the output of their transfer functions is the same, where μ' = (μ1 + μ2) / 2; Combining the two equations, it can be seen that the ideal yaw rate should satisfy the following constraints: At the same time, the upper limit of the ideal center of mass sideslip angle is: Finally, the reference model with the fitted adhesion coefficient μ' is obtained: Derivation of γ ref and β ref The expression is as follows: Where: n is the undamped frequency, ξ is the damping coefficient, τ β1 From δ to β ref The differential coefficient of the transfer function, τ γ1 From δ to γ ref The differential coefficient of the transfer function, τ γ2 M Z to γ ref The differential coefficient of the transfer function, G β1 From δ to β ref The transfer function of the center of mass sideslip angle stabilization gain, G β2 M Z To β ref The transfer function of the center of mass sideslip angle stabilization gain, G γ1 From δ to γ ref The yaw rate stabilization gain of the transfer function, G γ2 M Z To β ref The yaw rate stabilization gain of the transfer function, K is the gradient coefficient, L = a + b represents the distance from the front axle to the rear axle; β = v y / v x It can be seen that: yref =v x β ref ; ②Vertical Expected Value Longitudinal acceleration a x ≤μg; and It can be seen that xref (k)=(a xref +v y γ)T+v x (k-1), where v x (k) is the longitudinal velocity at the current moment, v x (k-1) is the longitudinal velocity at the previous moment, T is the sampling time; the optimal slip rate λ for different road surfaces ref =λ top , then the optimal angular velocity S2. NMPC-based lateral and longitudinal collaborative controller: Build a seven-degree-of-freedom vehicle dynamics model based on the Dugoff tire model: ① Seven-degree-of-freedom vehicle dynamics model Considering the longitudinal motion, lateral motion, yaw motion and rotation of the four wheels of the vehicle, a seven-degree-of-freedom vehicle dynamics model is obtained: Among them, v y is the lateral velocity of the vehicle, ω i is the wheel speed, T Ei and T Bi are the driving force and braking force applied to the wheels, respectively, x and F y They represent the longitudinal force and lateral force of the tire respectively, and the subscripts fl, fr, rl, and rr represent the left front, right front, left rear, and right rear wheels respectively; ②Tire model For longitudinal and lateral forces under combined transverse and longitudinal conditions Among them, λ i is the longitudinal slip rate of the wheel, F z,i is the tire vertical load, α i is the wheel slip angle; The vertical wheel load is: The wheel slip rate is Among them, the longitudinal speeds of the four wheels are: v xfl =(v x -Lγ)cosδ+(v y +aγ)sinδ (7a) v xfr =(v x +Lγ)cosδ+(v y +aγ)sinδ (7b) The tire's side slip angle α is: ③ Prediction model Normalize the state quantity x, that is, x=[v x / v xmax v y / v ymax c / c max ω1 / ω max ω2 / ω max 3 / hour max 4 / hour max ] T , among them v xmax ,v ymax ,γ max ,ω max are the upper limits of the vehicle longitudinal velocity, lateral velocity, yaw rate and wheel speed respectively; the control amount u=[ΔT bfr / ΔT bmax ΔT bfl / ΔT bmax ΔT brr / ΔT bmax ΔT brl / ΔT bmax ] T , where ΔT bmax Indicates the maximum value of additional braking torque; ④Objective function and constraints Define cost functions as J2 and J3; define cost function as J1; define cost functions J4 and J5: The control variable braking torque is constrained as follows: -1≤u(k+i-1)≤1,i=1,2,…,N (10a) The state quantity constraints are: -1≤x2(k+i)≤1,i=1,2,…,N (10b) The optimization problem is described as: st(4),(10) in, is the weight coefficient of acceleration, Γ γ is the weight coefficient of the yaw rate, Γ β is the weight coefficient of the sideslip angle of the center of mass, Γ λ1 is the weight coefficient of the left wheel slip rate of the vehicle, Γ λ2 is the weight coefficient of the right wheel slip rate of the vehicle.

Images