A corner module vehicle electronic mechanical braking system turning braking coordination control strategy

Abstract

The present application belongs to the technical field of intelligent electric vehicle chassis control and drive-by-wire technology, and particularly relates to a turning braking coordination control strategy of an angular module vehicle electronic mechanical braking system. The present application takes a 4WID-4WIS angular module vehicle as a research object, innovatively proposes a layered collaborative control architecture facing the overdrive characteristics of the angular module, realizes deep fusion optimization of longitudinal-horizontal-yaw dynamics through real-time information interaction and decision coordination between four-wheel independent control units, and breaks through the limitations of traditional centralized control.

Description

Technical Field

[0001] This invention belongs to the field of intelligent electric vehicle chassis control and drive-by-wire technology, and specifically relates to a corner module vehicle electromechanical braking system cornering braking coordination control strategy. Background Technology

[0002] As the automotive industry evolves towards "software-defined chassis," the integration and intelligence of chassis systems have become core trends in the development of intelligent electric vehicles. Distributed drive technology, as a key path to achieving full-domain chassis control, has evolved from traditional in-wheel motor drive to a "corner module" architecture. This physical form, which highly integrates the power unit, steer-by-wire actuator (SBW), and electromechanical braking system (EMB) into independent units, endows the four "corners" of the vehicle with independent drive, steering, braking, and suspension capabilities, constituting a typical MIMO over-actuated system. However, the increased physical integration also brings unprecedented control challenges, especially in the typical complex condition of brake-in-turn. The complexity of this condition stems from its significant longitudinal-lateral dynamic coupling characteristics; the longitudinal braking force and the lateral force are fundamentally competing under the constraint of a "friction ellipse." The compact design of the corner modules often leads to unexpected dynamic interference from the braking system to the steering system, exacerbating the control difficulty.

[0003] Research on the coordinated control of corner module vehicle turning and braking can be divided into three aspects: longitudinal and lateral decoupling control, dynamic coordinated distribution of braking force, and precise control of EMB system.

[0004] Lateral and longitudinal decoupling control is used to solve the problems of longitudinal and lateral force coupling interference and multi-steering mode adaptation under turning and braking conditions. It separates the longitudinal and lateral control objectives through decoupling algorithms, improving control accuracy and response speed. Yan et al. proposed a corner module vehicle hierarchical motion control framework, allocating front and rear axle steering angles through feedforward-feedback control and optimizing four-wheel steering angles using the Ackerman principle, achieving internal decoupling of the steering system. However, it did not deeply integrate the strong coupling constraints of longitudinal and lateral forces under braking conditions, nor did it achieve universal decoupling adaptation between 2WS and 4WS modes. Hang et al. designed a four-wheel steering vehicle path tracking controller based on a linear parameter variation system, combining the LQR algorithm and feedforward control. It achieved partial decoupling through parameter adaptation, adapting to different longitudinal speeds and friction coefficients, but did not address longitudinal and lateral dynamic decoupling under braking conditions. Li et al. adopted sliding mode control design and hierarchical four-wheel steering path tracking control, transforming path tracking into yaw rate tracking and combining it with lateral stability control to achieve lateral internal decoupling, but did not consider the impact of corner module integration characteristics on the decoupling effect. Under extreme conditions, the nonlinear and multidimensional coupling effects of vehicles are significantly enhanced. Existing decoupling control strategies mostly focus on a single steering mode or pure steering conditions, and do not adequately consider the coupling and decoupling of braking and steering. The overdrive advantage of corner module vehicles cannot be fully utilized in decoupling control, and external disturbance compensation and adaptability to variable adhesion road surfaces still need to be improved.

[0005] The key to dynamic coordination and distribution of braking force is adapting to the overdrive characteristics of the corner module, balancing braking efficiency and steering stability based on longitudinal and lateral decoupling. Wu et al. used a 2-DOF vehicle model as a basis and combined the attached ellipse constraint to solve the maximum longitudinal force online to achieve dynamic distribution of braking force. However, they did not optimize the distribution coefficient based on the difference in front and rear wheel steering angles in the 4WS mode, and did not fully adapt to the dynamic load changes after decoupling. ZHANG Z et al. proposed a coordinated control strategy for lateral stability and braking performance to optimize the cornering braking effect, but they did not introduce a lateral stability correction mechanism, resulting in insufficient coordination between the decoupled braking force distribution and the steering system. Traditional braking force distribution methods are mostly based on ideal I-curves or static constraints, which are difficult to deal with the nonlinear coupling problem of cornering braking after decoupling, and the overdrive advantage of the corner module is not fully released.

[0006] The EMB (Electronic Braking Module) system aims to improve actuator response speed and control accuracy, providing reliable execution guarantees for longitudinal and lateral decoupling control and dynamic distribution of braking force. Mou et al. designed a phased three-loop control strategy for EMB, combining sliding mode control to enhance anti-interference capabilities, but did not fully consider the real-time impact of dynamic load changes on clamping force requirements under cornering and braking conditions, and failed to adapt to the coordinated control requirements of steering and braking after decoupling. Linevai-Soos et al. designed an EMB controller based on sliding mode variable structure control to enhance system robustness, but did not design the coordinated control logic after decoupling in conjunction with the dynamic requirements of cornering and braking. Existing EMB control research mostly focuses on single-objective optimization, with few designs for multi-loop control architectures that incorporate the dynamic requirements after longitudinal and lateral decoupling. The robustness of sensorless clamping force estimation and its compatibility with the steering system still need improvement.

[0007] In summary, existing technologies, when addressing the complex problem of corner module vehicle cornering braking, typically treat longitudinal and lateral decoupling, braking force distribution, and actuator control as relatively isolated components, lacking a deeply integrated and coordinated system solution. This design disconnect between components prevents the full utilization of the corner module's overdrive potential. Under conditions of drastic dynamic load changes, strong coupling, and varying adhesion road surfaces, it becomes difficult to simultaneously guarantee excellent braking performance, precise steering follow-through, and robust lateral stability. Therefore, there is an urgent need in this field for an innovative and systematic coordinated control strategy to overcome these technical bottlenecks. Summary of the Invention

[0008] To address the problems existing in the prior art, this invention provides a coordinated control strategy for cornering braking in an electromechanical braking system for corner modular vehicles. Taking the 4WID-4WIS corner modular vehicle as the research object, this invention innovatively proposes a hierarchical cooperative control architecture oriented towards the overdrive characteristics of corner modules. Through real-time information interaction and decision-making coordination among the four independent control units, it achieves deep fusion and optimization of longitudinal, lateral, and yaw dynamics, breaking through the limitations of traditional centralized control.

[0009] A corner module vehicle electromechanical braking system cornering braking coordination control strategy includes the following steps:

[0010] S1. Establish a three-degree-of-freedom vehicle dynamics model that includes longitudinal, lateral, and yaw motions; a tire model that describes the coupling of longitudinal and lateral forces in the tire; and an EMB braking model.

[0011] S2. Construct an upper-level controller. The upper-level controller is based on the Ackerman steering principle and adopts a "feedforward + feedback" composite control strategy to allocate the steering angle of the four wheels. It also combines the model predictive control (MPC) method and the tire friction ellipse constraint to solve the target lateral force.

[0012] S3. Construct a lower-level controller. The lower-level controller receives the target lateral force from the upper-level controller. Based on braking strength, steering strength, and real-time lateral stability, it dynamically solves the target clamping force of each wheel through an adaptive braking force distribution algorithm. Based on the EMB system model, it drives the EMB actuator to output the corresponding braking torque through closed-loop control.

[0013] Furthermore, in S1, the EMB system model employs a two-stage identification method, including:

[0014] Offline identification stage: Based on EMB test bench data, the least squares method is used to identify key motor parameters and fit the nonlinear relationship between brake disc clamping force and ball screw axial displacement.

[0015] Online adaptive correction stage: The recursive least squares (RLS) algorithm with forgetting factor is introduced to update the friction torque model parameters and transmission efficiency in real time based on armature current, motor speed and wheel vertical load.

[0016] Furthermore, the specific operations of S2 are as follows:

[0017] S2.1 Receives vehicle steering wheel angle input, real-time yaw rate and vehicle speed signals;

[0018] S2.2 Based on the Ackermann steering principle, a "feedforward + feedback" composite control strategy is adopted to calculate and allocate the steering angle of the four wheels;

[0019] S2.3 Using a three-degree-of-freedom vehicle dynamics model as the prediction model, and combining the allocated four-wheel steering angles, construct a multi-objective optimization function with tire force utilization rate as the optimization objective and tire friction ellipse as the hard constraint.

[0020] S2.4. The optimal target lateral force that meets steering requirements and does not exceed the tire adhesion limit is obtained by using the model predictive control method for rolling solution.

[0021] Furthermore, the specific steps in S2.2 include:

[0022] S2.2.1 Convert the steering wheel angle input into a front axle angle reference;

[0023] S2.2.2. Based on the steady-state steering condition where the vehicle's center of gravity sideslip angle is zero, determine the feedforward coefficient of the rear axle steering angle relative to the front axle steering angle;

[0024] S2.2.3. Obtain the real-time yaw rate of the vehicle and determine the feedback coefficient in combination with the vehicle speed;

[0025] S2.2.4 Summing the feedforward and feedback terms yields the rear axle rotation angle;

[0026] S2.2.5. Based on the Ackermann steering geometry, the independent steering angles of the four wheels are calculated from the front axle angle and the rear axle angle.

[0027] Furthermore, the specific operations of S3 are as follows:

[0028] S3.1 Receive the target lateral force sent by the upper-level controller and obtain the real-time motion state of the vehicle;

[0029] S3.2 Based on braking strength, steering strength, and the sideslip angle and yaw rate that characterize lateral stability, the target clamping force of each wheel is dynamically solved through an adaptive braking force distribution algorithm.

[0030] S3.3 The adaptive braking force distribution algorithm includes a hierarchical adaptation function ψ(z,σ), which assigns different braking force distribution weights to the three working conditions of weak, medium and strong according to different combinations of braking intensity and steering intensity.

[0031] S3.4 Based on the EMB system model, the EMB actuator is driven by closed-loop control to output a braking torque corresponding to the target clamping force.

[0032] Furthermore, the specific operation of S3.3 is as follows: the hierarchical adaptation function ψ(z,σ) According to braking intensity z and steering strength σ The system is divided into three operating conditions: weak, medium, and strong. Different weighting coefficients are set for different operating conditions. In the weak operating condition, the weights are evenly distributed. In the medium operating condition, the braking performance and tracking requirements are balanced. In the strong operating condition, lateral stability is given priority.

[0033] Furthermore, in step S3.3, the adaptive braking force distribution algorithm also includes a lateral stability correction function ξ(β, ω), which is applied when the sideslip angle β or the yaw rate ω... z When the deviation from its expected value exceeds a preset threshold, the function reduces the braking force distribution weight of the corresponding wheel to prioritize lateral stability.

[0034] Furthermore, in S3.4, the lower-level controller drives the EMB actuator using a position-speed-force three-closed-loop control, wherein:

[0035] The force closed loop is the outer loop, which takes the deviation between the target clamping force and the clamping force estimated based on the EMB system model as input, and outputs the target motor speed through the PID controller;

[0036] The speed closed loop is the intermediate loop, which takes the deviation between the target speed and the actual speed of the motor as input, and outputs the target current through the PI controller;

[0037] The position closed loop is the inner loop. The target current and the motor rotor position feedback signals are used to generate the drive voltage through space vector pulse width modulation (SVPWM) to control the motor output torque.

[0038] Compared with the prior art, the beneficial technical effects of the present invention are as follows:

[0039] This invention focuses on a four-wheel steering-four-wheel independent braking intelligent vehicle (4WID-4WIS CAV). Addressing the strong coupling of longitudinal and lateral dynamics and the competition for tire adhesion resources under steering and braking conditions, a novel hierarchical electromechanical braking (EMB) cooperative control method is proposed. The effectiveness of this method is verified through joint simulations using MATLAB / Simulink and CarSim under various operating conditions, as well as hardware-in-the-loop (HIL) testing.

[0040] This invention solves the lateral force by combining the upper-level controller with the MPC and the tire friction ellipse hard constraint, and the lower-level adaptive braking force distribution ensures that the longitudinal and lateral forces are always coordinated within the tire adhesion limit, effectively suppressing the yaw rate fluctuation and preventing steering instability caused by excessive braking force intervention.

[0041] For four-wheel steering vehicles, the proposed hierarchical coordinated control strategy combined with the three-intensity adaptive allocation algorithm can make full use of the independent braking force control degrees of freedom of each wheel. In 4WS mode, it can significantly reduce the difference in braking torque fluctuation between the front and rear axles, optimize load transfer, reduce braking pitch, and improve ride comfort.

[0042] A two-stage identification method is used to accurately model the dynamic nonlinear characteristics of EMB. Combined with position-velocity-force three-loop control and sliding mode control, the system response delay is significantly reduced, and high-precision and fast tracking of the target clamping force is achieved, providing a reliable execution guarantee for upper-level coordination and decision-making.

[0043] The control method proposed in this invention performs well under various typical working conditions, including high, low, and variable adhesion road surfaces, as well as constant steering angle braking and lane-changing braking, verifying its strong adaptability to different road conditions and driving operations. Attached Figure Description

[0044] Figure 1 Configuration for corner module vehicles (CAVs).

[0045] Figure 2 This is an EMB coordination and control strategy based on a hierarchical architecture.

[0046] Figure 3 This is a schematic diagram of the integration of three-degree-of-freedom vehicle dynamics and four-wheel steering distribution for a angular module vehicle.

[0047] Figure 4This is a schematic diagram of the ellipse formed by tire friction adhesion.

[0048] Figure 5 This is for solving the four-wheel steering angle distribution and lateral force based on ASP.

[0049] Figure 6 The logic diagram of the lower-level controller for brake distribution coordination control.

[0050] Figure 7 The steering wheel and brake pedal are designed for emergency braking scenarios with a fixed steering angle.

[0051] Figure 8 The wheel lateral force curve is shown for emergency braking at a constant turning angle on a high-adhesion road surface at 90 km / h.

[0052] Figure 9 The wheel braking torque curve is for emergency braking at a constant turning angle on a high-adhesion road surface at 90km / h.

[0053] Figure 10 The yaw rate curve is for emergency braking at a constant turning angle on a high-adhesion road surface at 90 km / h.

[0054] Figure 11 The longitudinal acceleration curve is for emergency braking at a constant turning angle on a high-adhesion road surface at 90 km / h.

[0055] Figure 12 The yaw rate curve is for emergency braking under constant steering angle conditions at 60km / h on a variable adhesion road surface.

[0056] Figure 13 The steering wheel and brake pedal are designed for lane-changing braking scenarios on roads with varying adhesion at 60km / h.

[0057] Figure 14 The yaw rate curve is for lane changing braking under the condition of 60km / h with variable adhesion road surface.

[0058] Figure 15 This is a hardware-in-the-loop test platform for EMB braking systems.

[0059] Figure 16 Comparison of vehicle speed and wheel speed test results for emergency braking at 60 km / h on variable adhesion road surface under constant steering angle conditions.

[0060] Figure 17 A comparison of the trajectories of the 2WS and 4WS steering modes on the HIL test platform for 60km / h variable adhesion road surface.

[0061] Figure 18 Comparison of vehicle speed and wheel speed test results for lane change braking scenario at 60km / h on variable adhesion road surface.

[0062] Figure 19A comparison of the trajectories of the 2WS and 4WS steering modes on the HIL test platform at 60km / h with variable adhesion road surface for lane change braking scenario. Detailed Implementation

[0063] 1. 4WID-4WIS Corner Module Vehicle Structure

[0064] Typical corner module vehicles are equipped with four-wheel independent drive (4WID), four-wheel independent steering (4WIS), and an electromechanical braking system (EMB), and their structure is as follows: Figure 1 As shown. The vehicle control unit (VCU) of the corner module receives steering, drive, braking signals, battery signals, and other sensor signals from the vehicle. The VCU and the MCU of the corner module are connected via a wiring harness to complete the interaction of vehicle information and the processing of vehicle algorithms; the MCU acts as a domain controller to control the drive and steering actions of the corner modules; the brake control unit (BCU) is part of the MCU and is integrated into the EMB braking system, distributed in each corner module, responsible for the closed-loop control of the braking system, realizing the control of clamping force output. Table 1 shows the relevant parameters and output characteristics of the permanent magnet synchronous motor (PMSM, model: B20) used in the EMB system of this embodiment.

[0065] Table 1. Key Parameters and Output Characteristics of EMB Motors

[0066]

[0067] 2. Cornering Braking EMB Layered Coordination Control Framework

[0068] A framework for a layered coordinated control system for cornering and braking of a four-wheel independent steering vehicle with a cornering module configuration, such as Figure 2 As shown, this includes settings for cornering braking conditions, four-wheel steering angles, layered and coordinated distribution of braking force, and vehicle control.

[0069] In the cornering and braking condition setting module, the driver operation peripherals such as the simulator steering wheel and pedals are used in conjunction with MATLAB / Simulink simulation to inject signals such as steering wheel angle and brake pedal travel into the control algorithm and vehicle dynamics model, so as to achieve effective integration of real driving intention and simulation environment.

[0070] In the four-wheel steering angle and braking force hierarchical coordination module, the upper-level controller considers the differences in longitudinal and lateral velocities, simplifying the vehicle model into a three-degree-of-freedom single-track nonlinear model to reflect the vehicle's longitudinal-lateral-yaw coupling dynamics. A multi-objective optimization function is constructed based on tire adhesion elliptical constraints, combining the MPC method with a nonlinear vehicle dynamics model to solve for tire lateral forces, ensuring the adhesion of the steering and braking mechanisms and improving safety. Combining the Ackermann steering principle, a "feedforward + feedback" composite control strategy is adopted, adjusting the rear axle steering angle with yaw rate feedback to achieve four-wheel steering angle distribution for the corner module vehicle. The lower-level controller embeds a dynamic braking force distribution and sensorless clamping force estimation algorithm, integrating real-time vehicle information such as wheel speed and vehicle posture to solve for the four-wheel braking clamping force requirements. Leveraging the advantages of electromechanical braking (EMB)—no hydraulic circuit and rapid response—a position-velocity-force three-closed-loop control algorithm is used to accurately output braking torque, completing the closed-loop distribution of braking force in cornering braking scenarios. This hierarchical coordinated control strategy fully leverages the integrated advantages of the corner module's "drive-brake-turn" system and the independent and precise control characteristics of the EMB, effectively shortening the vehicle's dynamic stability response time, reducing yaw rate fluctuations, and improving vehicle comfort and steering following accuracy.

[0071] 3. System Modeling

[0072] 3.1, 4WIS Corner Module Vehicle Model

[0073] Based on the dynamic characteristics of the CAV, and neglecting the effects of yaw, vertical, and pitch motions, a three-degree-of-freedom nonlinear vehicle dynamics model is established, incorporating longitudinal, lateral, and yaw motions. The model assumes linear tire characteristics and satisfies small-angle conditions. Force analysis is as follows: Figure 3 As shown, this simplified model can reflect the longitudinal deceleration caused by vehicle braking, the distribution of vertical load on the tires, and the coupling effect of longitudinal and lateral forces, providing a dynamic basis for coordinated control of turning and braking. The specific expression is formula (1):

[0074]

[0075] Where: m is the vehicle's curb weight; , These are longitudinal and lateral accelerations, respectively. ω z , α and β are the yaw rate and yaw acceleration, respectively; a and b are the distances from the front and rear axles to the center of mass, respectively. C lf , C lr These are the longitudinal stiffnesses of the front and rear wheels, respectively. C cf , C crThese are the lateral stiffness of the front and rear wheels, respectively. λ f , λ r These are the slip ratios of the front and rear wheels, respectively. δ f Front axle steering angle; δ r This refers to the rear axle rotation angle.

[0076] Kinematic models are fundamental to trajectory planning and tracking control. To reduce the complexity of controller design, the 4WS vehicle kinematic model is simplified to a single-track model under pure rolling and small steering angle assumptions, such as... Figure 3 As shown. Figure 3 The black dots on the front and rear axles represent the centers of the vehicle's front and rear axles, respectively. The blue dot (CG) is the vehicle's geometric center, and point C is the center of rotation. R is the vehicle's radius of rotation. Wheelbase L refers to the distance between the front and rear axles, and wheel diameter B refers to the distance between the left and right wheels. Yaw angle. θ It is the angle between the direction of the vehicle's front and the X-axis. (Center of gravity sideslip angle) β It is the angle between velocity V at point CG and the vehicle's longitudinal axis. Ackerman steering ensures that the vertical lines of all wheels point towards the center of the circle when the vehicle is turning, resulting in a more stable turning posture, reduced tire wear, and improved driving safety. Based on the Ackerman steering principle, the ideal relationship between the vehicle's wheel angles under the steering module configuration is as follows: Figure 3 As shown. Front axle steering angle δ f and rear steering angle δ r Satisfying the Ackermann steering geometry, the steering angle of each wheel δ ij Satisfies formula (2), where, i=f / r , j=l / r , L Wheelbase B This refers to the wheel track.

[0077]

[0078] 3.2 Tire Model

[0079] When studying steering and braking coordination control strategies, the influence of air resistance is ignored, and the only external force changing the vehicle's speed is the force exerted by the ground on the tires. Therefore, the mechanical properties of the tires are of great significance to the vehicle's braking performance. Currently, tire mechanical model construction methods are generally divided into two categories: theoretical models and empirical models. This embodiment adopts the "Magic Tire Formula" empirical model. (Tire longitudinal force) F x The calculation formula is:

[0080]

[0081] In the formula: X 1 For the longitudinal force combination independent variable: X 1=( s r +S h) , s r This refers to the longitudinal slip ratio of the wheel; B is the stiffness factor, representing the slope at the origin of the characteristic curve; C The shape factor of the curve; D The peak factor represents the maximum value of the curve; E The curvature factor represents the shape near the maximum value of the curve; S h This represents the horizontal offset of the characteristic curve. S V This represents the vertical drift of the curve.

[0082] The longitudinal and lateral slip velocities of the wheel center are the cause of the longitudinal and lateral forces of the tire. The longitudinal slip stiffness of the tire is not affected by the lateral force and tire carcass deformation, but the lateral slip stiffness is affected by the longitudinal force and tire carcass deformation. Considering the coupling effect of the longitudinal and lateral forces of the tire, their simplified relationship is shown in formula (4):

[0083]

[0084] in, F xijmax and F yijmax These are the current tire slip direction angles. β ij The corresponding maximum longitudinal force and lateral force. C xij and C yij These are the friction elliptic coefficients for the longitudinal and lateral forces of the tire, respectively, which can be fitted based on experimental data. μ max The maximum adhesion coefficient of the road surface. F zij This refers to the vertical load on the tire. For example... Figure 4 As shown, in the limiting case, the friction ellipse... F xijmax With the major semi-axis, F yijmax For the short half-shaft, the actual longitudinal force of the tire at any given time... F xij and actual lateral force F yij The determined tire force vector F ij=(F xij , F yij) T It must be located inside the friction ellipse.

[0085] 3.3 EMB Braking Model

[0086] The EMB actuator uses a permanent magnet brushless DC torque motor. Motor friction directly affects the generated clamping force. Therefore, establishing an accurate and suitable motor friction model is crucial for improving the overall braking performance of the EMB system. Thus, the friction torque... T D A model consisting of static friction, Coulomb friction, and viscous friction is adopted. The dynamic equation of the motor is shown in formula (5):

[0087]

[0088] In the formula, U a Armature voltage; I a R is the armature current; R is the armature resistance; L a For armature inductance; K t The torque constant; θ m For rotor angle; J m This refers to the moment of inertia of the motor. T D This is the frictional torque; T L This is the load torque; T w External torque; ω Δ Indicates relative sliding speed; T s This represents the maximum static friction torque. T k The frictional torque is the Coulomb friction torque. G It is the coefficient of viscous friction.

[0089] To capture the EMB nonlinear characteristics during CAV steering and braking, a two-stage identification method is proposed:

[0090] Offline identification: Based on EMB test bench data (clamping force range 300-10000N), the key motor parameters in equation (5) are identified using the least squares method. K t , J m In addition, the clamping force on the brake disc. Fcl Axial displacement of the ball screw x emb There is a nonlinear relationship between them, which can be fitted using a cubic polynomial. The EMB transmission mechanism and braking force model are shown in Equation (6).

[0091] Online adaptive correction: Introducing a recursive least squares (RLS) algorithm with a forgetting factor to update friction torque in real time. T D And transmission efficiency. Input parameters include I a , and wheel vertical load F zij .

[0092] The output torque of the drive motor is transmitted through the transmission mechanism, generating a clamping force on the brake disc. This also results in a load torque. T L The clamping force acting on the drive motor and the brake disc. F cl Axial displacement of the ball screw x emb There is a nonlinear relationship between them, which can be fitted by a cubic polynomial. The EMB transmission mechanism and braking force model are shown in formula (6):

[0093]

[0094] In the formula, x emb This refers to the axial displacement of the ball screw; P h For the ball screw lead; i p Indicates the reduction ratio of the planetary gear; R b The radius of the brake disc; μ b Indicates the coefficient of friction of the brake disc; η s To improve the efficiency of ball screw transmission; η p This indicates the efficiency of the planetary gear transmission. EMB The braking torque of the brake disc is T EMB .

[0095] 4. Design of 4WD-4WIS Corner Module Vehicle Layered Coordination Control Strategy

[0096] To address the overdrive characteristics of 4WD-4WIS corner module vehicles and the issues of longitudinal and lateral force coupling and load transfer under cornering and braking conditions, this embodiment adopts a hierarchical control architecture to optimize and coordinate the control strategy. The upper layer uses a "feedforward + feedback" composite control to achieve four-wheel steering angle distribution, and combines MPC to solve for the lower-level input of lateral force. The lower layer proposes a graded adaptive braking force distribution algorithm, embedding a sliding mode control ABS module, and combining it with the EMB system's three-closed-loop control to achieve actuator load balancing and reasonable allocation of tire adhesion resources while satisfying steering stability.

[0097] 4.1 Upper-level controller based on ASP for four-wheel steering angle distribution and lateral force calculation

[0098] like Figure 5 As shown, the upper-level controller adopts a "feedforward + feedback" composite control strategy to achieve rapid and stable distribution of the front and rear axle rotation angles. The feedforward control based on the Ackerman principle completes the rotation angle distribution, while the yaw rate feedback adjustment achieves dynamic coordination. When the lateral force demand increases, the rear axle rotation angle is corrected in real time to reduce the lateral load on the inner wheel and reserve attachment space for longitudinal braking force. At the same time, the lateral force target solved by MPC is used as a hard constraint and transmitted to the lower-level braking force distribution module to ensure that the longitudinal and lateral forces do not exceed the friction ellipse limit.

[0099] The controller uses steering wheel angle input. δ SW Based on this, the rear axle rotation angle is adjusted by combining yaw rate feedback. δ r To ensure vehicle steering stability, the specific allocation relationship is shown in the formula for feedforward coefficients. k 1 With feedback coefficient k 2 The expression is derived from the constraint that the sideslip angle of the center of gravity is zero during steady-state steering of the vehicle, and is given by formula (7):

[0100]

[0101] in, k 1 A constant less than zero, used to improve vehicle maneuverability; k 2 A function of vehicle speed, which can provide real-time feedback of yaw rate. ω z The rear axle angle is dynamically adjusted to enhance stability during high-speed braking and steering.

[0102] Based on the lateral force balance relationship of the 3DOF vehicle dynamics model, the expression for calculating the lateral force of the front and rear wheels is formula (8). F yf , F yrThese represent the lateral forces of the front and rear wheels, and the total lateral force required for steering. F ycmd The sum of the lateral forces of the front and rear wheels represents the real-time state of the vehicle (v). x v y , ω z ) and steering wheel angle input δ sw Substituting the above formulas and combining them with the tire model in Section 3.2, a multi-objective optimization function is constructed, and the MPC method is used to solve for the tire lateral force.

[0103]

[0104] 4.2 Lower-level controller of EMB braking system coordinated control strategy

[0105] Under cornering braking conditions, the vehicle exhibits significant longitudinal and lateral force coupling and load transfer. To address this, the lower-level controller, combining the ideal braking force distribution curve and ECE regulatory constraints, proposes an adaptive braking force distribution algorithm. Based on tire adhesion ellipse constraints, the algorithm introduces a lateral stability correction function and constructs a hierarchical adaptation function to solve the insufficient adaptability of traditional distribution methods under cornering braking conditions. The algorithm also includes a dynamic distribution coefficient. C brij The solution is shown in formula (9):

[0106]

[0107] in, ψ(z,σ) For the graded adaptation function, combined with ECE regulatory constraints, based on braking intensity z and steering strength σ Three levels of dynamic adaptation are set: weak, medium, and strong. In the weak condition, the weights are evenly distributed to avoid excessive interference of braking force with steering stability. In the medium condition, braking performance and tracking requirements are balanced. In the strong condition, lateral stability is prioritized. (Equation follows) k z1 , k z2 , kσ 1 , kσ 2 These are the braking strength correction coefficient and steering strength correction coefficient, respectively; lateral stability correction function. ξ ( β,ω z To address the nonlinear dynamics of cornering braking, when the sideslip angle approaches its limit or the yaw rate deviates significantly from the desired value, the braking force weight of the corresponding wheel is reduced to prevent tire instability. When the vehicle's lateral state is stable, the braking force weight is maximized to ensure braking efficiency. In the formula... k β and kω These are the correction coefficients for the center of gravity sideslip angle and the yaw rate, respectively; the steering strength σ quantifies the intensity of the steering condition through changes in steering angle and steering stability deviation, where δ fmax This is the maximum steering angle of the front wheels. ω zmax For the maximum yaw rate, ω zd Let be the desired yaw rate.

[0108] like Figure 6 As shown, the lower-level control strategy uses the vehicle's lateral acceleration. a y Front axle steering angle δ f Rear axle steering angle δ r Longitudinal speed v x , vertical force F z Slip ratio λ ij Target lateral force F ycmd and yaw rate ω z As input, a hierarchical adaptive braking force allocation algorithm is used to output braking force allocation coefficients. C brij The target clamping force calculation submodule, in conjunction with EMB parameters such as brake disc radius and friction coefficient, solves for the target clamping force. F Clcmd Considering changes in road surface adhesion during braking, the lower-level controller incorporates a sliding mode control ABS module, through... k abs1 , k abs2 , k abs3 The module calibrates the sliding surface and the approach law to achieve tracking control of the wheel slip ratio. It receives feedback on the target clamping force and wheel speed, and corrects the clamping force in real time. This ensures braking performance while preventing excessive longitudinal braking force from consuming adhesion resources, and also controls lateral force. F ycmd To achieve dynamic adaptation.

[0109] Table 2. Key parameters of EMB coordination control strategy

[0110]

[0111] The revised clamping force requirement is transmitted to the EMB system, where the BCU receives the signal and initiates the position-velocity-force three-loop control algorithm. Key parameters of the EMB coordinated control strategy are shown in Table 2. J embTo include the equivalent moment of inertia of the reduction gear, the three closed-loop control parameters are used. k And the corresponding subscript representation.

[0112] The force closed loop is the outer loop, taking the deviation between the target clamping force and the sensorless estimated clamping force as input. A PID controller outputs the target motor speed, and the sensorless clamping force is based on the motor current. i a / i b / i c Rotation speed ω r The system uses signals such as the EMB transmission mechanism model (Formula 6) to estimate in real time and simultaneously compensate for nonlinear interference caused by static friction, Coulomb friction, and viscous friction. The speed closed loop is the intermediate loop, receiving the target speed output from the force closed loop and comparing it with the actual motor speed acquired by the encoder. After adjustment by the PI controller, it outputs the target current to achieve rapid speed tracking. The position closed loop is the inner loop, based on the target current output from the speed closed loop and combined with the motor rotor position signal fed back from the encoder, to generate the drive voltage through space vector pulse width modulation (SVPWM). U k (subscript) k=a,b,c The permanent magnet synchronous motor is driven to output torque, which is combined with the load torque. T load Feedback compensation, through the ball screw, converts the motor torque into braking clamping force, completing the closed-loop control of "current-speed-clamping force", and ultimately generating four-wheel braking torque. T bij It applies to corner module vehicles.

[0113] 5. Simulation Analysis and Experimental Verification

[0114] 5.1 Co-simulation platform setup

[0115] Based on the proposed mathematical model, this embodiment builds a MATLAB / Simulink and CarSim co-simulation platform for algorithm verification and debugging, such as... Figure 2 As shown in Table 3, the co-simulation platform follows the overall framework logic of the control strategy and is divided into four parts: operating condition settings, upper-level controller, lower-level controller, and CarSim vehicle status. A distributed drive B-class vehicle is selected from the CarSim vehicle library for configuration, with the steering and braking systems set as external inputs. The key parameters of the vehicle dynamics model are shown in Table 3.

[0116] Table 3. Parameters of the Vehicle Simulation Model

[0117]

[0118] in, ms For the sprung mass, h CG For the height of the center of mass, r b The tire's rolling radius, I x and I z These are the moments of inertia along the x and z axes, respectively. I w This represents the longitudinal rotational inertia of the tire; the meanings of the other symbols are the same as those mentioned above.

[0119] 5.2 Co-simulation setup and result analysis

[0120] 5.2.1 Setting up and analyzing the results of an emergency braking scenario with a fixed steering angle

[0121] The constant steering angle emergency braking simulation describes the emergency braking state of a vehicle while driving on a curve. Figure 7 As shown, the steering wheel angle increases from 0 to 60° within 500ms, the brake pedal is pressed to 100% in 300ms, and the variable adhesion road surface is set so that the road surface adhesion coefficient changes abruptly from 0.85 to 0.3 starting from a longitudinal distance of 40m, simulating the elevated ice surface scenario of a real road.

[0122] Table 4. Simulation Parameters for Emergency Braking Scenario with Fixed Steering Angle

[0123]

[0124] Three typical operating conditions were set for emergency braking scenarios with a fixed steering angle, and the parameters are shown in Table 4. The three operating conditions tested two-wheel steering (2WS) and four-wheel steering (4WS) conditions, respectively, and simultaneously covered cornering braking conditions on high-adhesion, low-adhesion, and variable-adhesion road surfaces. This setup allows for a comprehensive assessment of the robustness of the coordinated control strategy. In the simulation results, the lateral force and torque curves show that the black and red solid lines represent the left and right front axle wheels, respectively, and the blue and green dashed lines represent the left and right rear axle wheels, respectively. In the yaw rate and acceleration curves, the red and blue solid lines represent the 2WS and 4WS steering conditions, respectively.

[0125] Simulation results for high-adhesion pavement at 90km / h are as follows: Figure 8-11 As shown, Figure 8 The curves show the variation of lateral force on the vehicle. Under both steering conditions, the lateral force on the front axle tires is greater than that on the rear axle, and the lateral force on the right side is generally greater than that on the left. This reflects the load transfer characteristics under left-hand steering braking. Observing the 2-6 second interval, the mean change of the vehicle's lateral force under 4WS conditions is more stable compared to 2WS conditions, indicating that the steering strength σ term in the braking force distribution algorithm plays a role in adjusting steering stability. Figure 10The yaw rate variation curve provides a more intuitive observation that the yaw rate in the 4WS case is controlled below the desired yaw rate curve. Figure 9 The curve shows the vehicle's braking torque. The right wheel oscillates due to the ABS function being triggered by load transfer. In both 2WS and 4WS modes, the dynamic braking force distribution strategy comes into play. In the 4WS mode, the average difference between the front and rear axle braking torques is reduced by about 19.7% compared to the 2WS mode, which can effectively reduce braking pitch. Figure 11 It shows the acceleration curve; the average longitudinal deceleration of a 4WS vehicle is approximately 7.5 m / s². 2 This indicates that the vehicle has good braking performance.

[0126] Analysis of the simulation results for the 60km / h low-adhesion road surface condition shows that the performance of various physical quantities on the low-adhesion road surface is consistent with the overall pattern of the high-adhesion road surface. The coordinated control strategy has a stable control effect on the low-adhesion road surface. The oscillation amplitude is smaller in the 4WS case compared to the 2WS case, and the difference in the mean values ​​of the front and rear axle lateral forces and braking torques is smaller. The 4WS case has better lateral stability and longitudinal deceleration effect, further verifying the robustness of the coordinated control algorithm.

[0127] Simulation results for a 60 km / h variable adhesion road surface condition show that when a vehicle transitions from a high adhesion surface (0.85) to a low adhesion surface (0.3), the lateral force and braking torque of the wheels respond rapidly. Through the adaptation and adjustment of the inner and outer wheels, the vehicle deceleration decreases rapidly after 2.4 seconds, with an average value of 2.4 m / s². 2 The left and right positions meet the emergency braking requirements for low-adhesion road surfaces. Figure 12 As shown, the yaw rate in both 2WS and 4WS cases meets the expectations, and the 4WS case has better cornering stability, which verifies the effectiveness of the proposed adaptive braking force distribution algorithm.

[0128] 5.2.2 Lane Change Braking Scenario Setup and Result Analysis

[0129] Table 5. Simulation Parameters for Lane Changing Braking Scenarios

[0130]

[0131] Three typical operating conditions were set for the lane change braking scenario, and the parameters for these three conditions are shown in Table 5. Conditions No.1 and No.2 have a uniform road surface adhesion coefficient, while condition No.3 has a variable adhesion contact road surface. Taking No.3 as an example, the steering wheel angle and brake pedal percentage are as follows: Figure 13 As shown, this is used to simulate the driver's braking operation during lane changing. After the lane change is completed, the brake pedal returns to 0%, and the vehicle continues to travel at the original speed. The variable adhesion road surface is set to 70m-100m for the lane changing section, and the road surface adhesion coefficient changes abruptly from 0.85 to 0.3 starting from 85m.

[0132] The braking condition during cornering on a uniform road surface can be comprehensively reflected by the braking condition on a variable adhesion road surface. Therefore, the No. 3 braking condition for lane changing on a 60km / h variable adhesion road surface is discussed and analyzed, and its simulation results are as follows. Figure 14 As shown, the steering and braking intensity are relatively small under variable adhesion lane change braking conditions. During the period when steering and braking are applied simultaneously, the lateral force of the outer wheel is larger during the single lane change braking process. The lateral force changes are gradual under 2WS and 4WS steering conditions, verifying the effect of the dynamic distribution strategy. The right rear wheel generates braking torque when the left steering amplitude is large, indicating that the hierarchical adaptation function plays a role in balancing braking performance and tracking requirements under intermediate conditions. The average braking torque under 4WS conditions is significantly smaller than that under 2WS conditions. Combined with the yaw rate change curve, this shows that the hierarchical coordinated control strategy can achieve lateral and longitudinal coordinated control in lane change braking scenarios under variable adhesion surfaces.

[0133] 5.3 HIL Test Verification

[0134] 5.3.1 EMB Hardware-in-the-Loop Experimental Platform

[0135] A hardware-in-the-loop (HIL) test platform combining a driving simulator and an EMB bench was built to evaluate the real-time control effect of a hierarchical coordinated control strategy, such as... Figure 15 As shown, the steering wheel and brake pedal serve as driver input devices, injecting steering angle and braking demand signals under combined turning and braking conditions into the test platform. The SIMULINK hierarchical coordinated control algorithm is deployed on an NI real-time industrial computer connected to the host computer via a TCP communication network, ensuring the real-time nature of control decisions and dynamic simulation. The host computer, NI real-time industrial computer, and dedicated EMB programmable power supply are all integrated into the braking system HIL test control cabinet. The control cabinet provides a stable 24V DC power supply to the EMB bench and establishes a communication link via a CAN bus to ensure the issuance of control commands and feedback on execution status. The host computer uses TSMaster with integrated CarSim-Controller function modules in conjunction with NIVeriStand to complete algorithm deployment, while simultaneously adjusting and visualizing parameters such as control algorithm output, vehicle dynamic response, EMB clamping force, wheel speed, and slip ratio in real time. It can also synchronously record test process data, providing reliable support for verifying the effectiveness of the control strategy.

[0136] 5.3.2 Analysis of Test Results for Emergency Braking Scenario with Fixed Steering Angle

[0137] First, three operating conditions of the constant steering angle emergency braking scenario were experimentally verified. The steering wheel and pedal signals were consistent with the simulation, and the test results for the variable adhesion road surface were as follows: Figure 16-17 As shown, where Figure 16 The colors of the four curves correspond to those in the simulation test, with the purple curve representing vehicle speed. Figure 17 The red vehicles correspond to the 2WS scenario, and the blue vehicles correspond to the 4WS scenario.

[0138] Figure 16 The curves show the changes in vehicle speed and wheel speed under two steering conditions. The oscillating changes in wheel speed indicate that the dynamic distribution of the control strategy and the slippage control ABS function are effective. The overall vehicle speed changes smoothly, which is basically consistent with the trend of the simulated braking torque and lateral force changes. Figure 16 trajectory curve and Figure 17 Screenshots from the CarSim scenario on the HIL test platform show that in the 4WS steering configuration, the vehicle utilizes more lateral force, resulting in better tracking performance. Compared to the expected trajectory, the lateral error is reduced by 10.9% in the 2WS configuration. At the same time, the braking distance is longer than in the 2WS configuration, with a braking time of approximately 5.1 seconds. The vehicle achieves rapid braking force response on variable adhesion surfaces, with smooth speed changes and good cornering braking trajectory tracking.

[0139] 5.3.3 Analysis of Lane Changing Braking Scenario Test Results

[0140] When conducting lane change braking scenario tests, the steering braking parameter settings were kept consistent with the lane change braking scenario simulation. The variable adhesion road surface test results are as follows: Figure 18-19 As shown, the legend definition is consistent with the previous text.

[0141] Depend on Figure 18 A comparison of the speed curves shows that the speed variation range is approximately 3.4 km / h with 4WS and approximately 8.4 km / h with 2WS. Both steering, turning, braking, and tracking effects are good, and speed changes are smooth. Figure 19 As shown, at the end of the test, the trajectory of the 2WS vehicle was slightly behind that of the 4WS vehicle, and the 4WS tracking performance was better than that of the 2WS vehicle; the vehicle achieved smooth lane change braking under the coordinated control algorithm.

[0142] Conclusion: This invention focuses on a four-wheel steering-four-wheel independent braking intelligent vehicle (4WID-4WIS CAV). Addressing the strong coupling of longitudinal and lateral dynamics and the competition for tire adhesion resources under steering and braking conditions, it proposes a corner module vehicle electromechanical braking system cornering braking coordinated control strategy. The effectiveness of this method is verified through MATLAB / Simulink and CarSim co-simulation under various operating conditions, as well as hardware-in-the-loop (HIL) testing.

[0143] (1) In both two-wheel steering (2WS) and four-wheel steering (4WS) modes, the method can suppress yaw rate fluctuations and stabilize lateral force changes. In emergency braking under variable adhesion conditions, the tracking error of the 4WS mode is reduced by up to 10.9% compared to the 2WS mode when the steering angle is fixed, and the yaw rate is always controlled within the expected range, meeting the steering and braking requirements of vehicles with different steering configurations.

[0144] (2) As a typical four-wheel steering platform, CAV is an ideal physical carrier for achieving precise longitudinal and lateral coordinated control. Combined with the three-intensity adaptive braking force distribution algorithm, the difference in braking torque fluctuation between the front and rear axles can be reduced by 19.7% in 4WS mode, giving full play to the advantages of the system's over-actuator.

[0145] (3) To address the problem of insufficient consideration of EMB dynamic characteristics in existing technologies, the proposed two-stage identification method accurately captures EMB hysteresis and torque delay characteristics. Combined with a three-closed-loop control structure including improved sliding mode control, the response delay is reduced to 55 milliseconds, achieving reliable execution of refined braking force distribution.

Claims

1. A corner module vehicle electromechanical braking system cornering braking coordination control strategy, characterized in that: Includes the following steps: S1. Establish a three-degree-of-freedom vehicle dynamics model that includes longitudinal, lateral, and yaw motions; a tire model that describes the coupling of longitudinal and lateral forces in the tire; and an EMB braking model. S2. Construct an upper-level controller. The upper-level controller is based on the Ackermann steering principle and adopts a "feedforward + feedback" composite control strategy to allocate the steering angle of the four wheels. It also combines the model predictive control (MPC) method and the tire friction ellipse constraint to solve the target lateral force. S3. Construct a lower-level controller. The lower-level controller receives the target lateral force sent by the upper-level controller. Based on braking strength, steering strength and real-time lateral stability, it dynamically solves the target clamping force of each wheel through an adaptive braking force distribution algorithm. Based on the EMB system model, it drives the EMB actuator to output the corresponding braking torque through closed-loop control. In S1, the EMB braking model employs a two-stage identification method, including: Offline identification stage: Based on EMB test bench data, the least squares method is used to identify key motor parameters and fit the nonlinear relationship between brake disc clamping force and ball screw axial displacement. Online adaptive correction stage: The recursive least squares (RLS) algorithm with forgetting factor is introduced to update the friction torque model parameters and transmission efficiency in real time based on armature current, motor speed and wheel vertical load; The specific operation of S2 is as follows: S2.1 Receives vehicle steering wheel angle input, real-time yaw rate and vehicle speed signals; S2.2 Based on the Ackermann steering principle, a "feedforward + feedback" composite control strategy is adopted to calculate and distribute the steering angles of the four wheels; S2.3 Using a three-degree-of-freedom vehicle dynamics model as the prediction model, and combining the allocated four-wheel steering angles, a multi-objective optimization function is constructed with tire force utilization rate as the optimization objective and tire friction ellipse as the hard constraint. S2.

4. Using the model predictive control method to solve the problem in a rolling manner, the optimal target lateral force that meets the steering requirements and does not exceed the tire adhesion limit is obtained. The specific operation of S3 is as follows: S3.1 Receive the target lateral force sent by the upper-level controller and obtain the real-time motion state of the vehicle; S3.2 Based on braking strength, steering strength, and the sideslip angle and yaw rate that characterize lateral stability, the target clamping force of each wheel is dynamically solved through an adaptive braking force distribution algorithm. S3.3 The adaptive braking force distribution algorithm includes a hierarchical adaptation function ψ(z,σ), which assigns different braking force distribution weights to the three working conditions of weak, medium and strong according to different combinations of braking intensity and steering intensity. S3.4 Based on the EMB system model, the EMB actuator is driven by closed-loop control to output a braking torque corresponding to the target clamping force.

2. The corner module vehicle electromechanical braking system cornering braking coordination control strategy according to claim 1, characterized in that: The specific steps in S2.2 include: S2.2.1 Convert the steering wheel angle input into a front axle angle reference; S2.2.

2. Based on the steady-state steering condition where the vehicle's center of gravity sideslip angle is zero, determine the feedforward coefficient of the rear axle steering angle relative to the front axle steering angle; S2.2.

3. Obtain the real-time yaw rate of the vehicle and determine the feedback coefficient in combination with the vehicle speed; S2.2.4 Summing the feedforward and feedback terms yields the rear axle rotation angle; S2.2.

5. Based on the Ackermann steering geometry, the independent steering angles of the four wheels are calculated from the front axle angle and the rear axle angle.

3. The corner module vehicle electromechanical braking system cornering braking coordination control strategy according to claim 1, characterized in that: The specific operation of S3.3 is as follows: the hierarchical adaptation function ψ(z,σ) According to braking intensity z and steering strength σ The system is divided into three operating conditions: weak, medium, and strong. Different weighting coefficients are set for different operating conditions. In the weak operating condition, the weights are evenly distributed. In the medium operating condition, the braking performance and tracking requirements are balanced. In the strong operating condition, lateral stability is given priority.

4. The corner module vehicle electromechanical braking system cornering braking coordination control strategy according to claim 3, characterized in that: In step S3.3, the adaptive braking force allocation algorithm further includes a lateral stability correction function. ξ(β, ω z ) When the centroid sideslip angle β or yaw rate ω z When the deviation from its expected value exceeds a preset threshold, the function reduces the braking force distribution weight of the corresponding wheel to prioritize lateral stability.

5. The corner module vehicle electromechanical braking system cornering braking coordination control strategy according to claim 1, characterized in that: In S3.4, the lower-level controller drives the EMB actuator using a position-speed-force three-closed-loop control, wherein: The force closed loop is the outer loop, which takes the deviation between the target clamping force and the clamping force estimated based on the EMB system model as input, and outputs the target motor speed through the PID controller; The speed closed loop is the intermediate loop, which takes the deviation between the target speed and the actual speed of the motor as input, and outputs the target current through the PI controller; The position closed loop is the inner loop. The target current and the motor rotor position feedback signals are used to generate the drive voltage through space vector pulse width modulation (SVPWM) to control the motor output torque.

Images