Full-working-condition active stability control system and control method for tractor semitrailer

Abstract

The invention provides a tractor semitrailer full-working-condition active stability control system and method, and belongs to the technical field of vehicle engineering. The system comprises a kinetic model module, a reference quantity determination module and an active stability controller module. The dynamic model module constructs a linear three-degree-of-freedom tractor semitrailer model, a hub motor model and a longitudinal driver model. The reference quantity determination module generates an ideal state value according to the steady state condition of the system and corrects the ideal state value in combination with the road attachment condition. The active stability controller module is divided into three layers, an MIMO-IMFAC controller calculates additional yawing moment, a torque optimization controller optimizes tire load distribution through a quadratic programming algorithm, and a driving / braking torque distribution controller preferentially uses regenerative braking and dynamically adjusts braking torque of the semitrailer according to the yawing moment. By adopting the control system and the control method, the stability and the dynamic response capability of the vehicle under complex working conditions are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle engineering, and particularly to an active stability control system and control method for a tractor-semitrailer under all working conditions. Background Art

[0002] As a key equipment for road freight transportation, the tractor-semitrailer occupies an important position in the modern logistics industry due to its low transportation cost and high efficiency. It can transport a large amount of goods at one time, meeting the needs of large-scale material allocation, and playing a crucial role in promoting economic development and material circulation. However, compared with ordinary vehicles, the tractor-semitrailer has a large mass, a high center of gravity, a long body length, and there is a complex non-linear dynamic coupling effect between the tractor and the semitrailer. These factors pose severe challenges to the handling stability of the vehicle under conditions such as high-speed driving and emergency obstacle avoidance. In the common high-speed lane-changing scenario on highways, the tractor-semitrailer usually needs to perform lane-changing operations at a speed of not less than 60 m / h, and sometimes the semitrailer even needs to continuously change lanes at high speed. Such instantaneous steering operations cause a sharp change in the lateral force of the vehicle, greatly increasing the risk of vehicle instability and easily triggering serious traffic accidents. Once an accident occurs, it will not only cause cargo loss and vehicle damage, but may also lead to casualties and traffic congestion, bringing huge losses to society and the economy.

[0003] In the early stage, some studies explored the driving stability of tractor-semitrailers. For example, some studies conducted simulation analyses on the braking performance of commercial semitrailers during turning, and some studies developed mathematical methods for predicting the braking performance of trucks and semitrailers. These studies provided a certain theoretical basis for subsequent vehicle stability control. In recent years, typical stability control strategies include active steering of trailer wheels and differential braking, etc. Differential braking enhances the lateral stability of the vehicle by generating additional yaw moments. Compared with active steering of trailer wheels, it can play a role more quickly, directly and effectively, and does not require modification of the vehicle structure, so it has received extensive attention. However, existing studies have certain limitations. Many control methods rely on accurate vehicle models, but the dynamic characteristics of tractor-semitrailers are extremely complex, and they are affected by various factors such as vehicle load distribution, tire characteristics, and road conditions. Controllers designed based on simplified models often fail to effectively function under actual working conditions because the simplified models cannot accurately describe the dynamic behavior of the vehicle in complex situations, resulting in poor control effects. In addition, fuzzy control methods require powerful expert rules for configuration, which increases the complexity and uncertainty of control. The formulation of expert rules relies on experience and subjective judgment, and it is difficult to cover all possible working conditions. Moreover, in practical applications, the adjustment and optimization of rules are also relatively difficult. At the same time, some control methods lack sufficient adaptability and flexibility when dealing with multi-variable and strongly coupled systems, and cannot respond in real time to changes in vehicle states and external environments. Summary of the Invention

[0004] The object of the present invention is to provide a full-condition active stability control system and control method for a tractor-semitrailer, so as to solve the problem of poor stability of the tractor-semitrailer under working conditions such as emergency obstacle avoidance, improve the driving safety of the vehicle, and reduce the accident risk. By constructing a dynamic model, designing an advanced controller and a reasonable torque distribution strategy, the vehicle can maintain stable driving under various complex working conditions, and reduce the occurrence of dangerous situations such as sideslip, fishtailing and rollover. At the same time, the designed controller has strong real-time performance and low requirements for the dynamic model, and can adapt to the complex road environment of heavy-duty semitrailers.

[0005] To achieve the above object, the present invention provides a full-condition active stability control system for a tractor-semitrailer, including:

[0006] A dynamic model module, including a linear three-degree-of-freedom tractor-semitrailer model, a wheel hub motor model and a longitudinal driver model, which are respectively used to calculate the yaw angular velocity and the sideslip angle of the center of mass of the tractor-semitrailer, describe the dynamic response of the motor, and calculate the total driving torque;

[0007] A reference quantity determination module, which is used to calculate the steady-state value of the system state variable according to the condition that the derivative of the state variable of the system is zero at steady state, and correct it in combination with the road adhesion condition;

[0008] An active stability controller module, which has three layers. The upper layer is a MIMO-IMFAC controller, the middle layer is a torque optimization controller, and the lower layer is a drive / brake torque distribution controller, which are respectively used to calculate the additional yaw moment, optimize the tire load distribution and adjust the braking torque.

[0009] Preferably, the linear three-degree-of-freedom tractor-semitrailer model includes a tractor dynamic equation and a semitrailer dynamic equation;

[0010] The tractor dynamic equation is as follows:

[0011]

[0012] The semitrailer dynamic equation is as follows:

[0013]

[0014] Wherein, m1 and m2 respectively represent the masses of the tractor and the semitrailer, u1 and u2 respectively represent the longitudinal speeds of the tractor and the semitrailer, respectively represent the first-order derivatives of the yaw angular velocities of the tractor and the semitrailer, respectively represent the second-order derivatives of the yaw angular velocities of the tractor and the semitrailer, They respectively represent the first-order derivatives of the sideslip angles of the center of mass of the tractor and the semi-trailer. F1, F2, and F3 respectively represent the lateral forces of the three simplified axles in the model. F4 represents the lateral interaction force between the tractor and the semi-trailer at the fifth wheel. Γ represents the folding angle between the tractor and the semi-trailer at the saddle, I 1z ,I 2z respectively represent the yaw moments of inertia of the tractor and the semi-trailer. a, b, c, d, e represent geometric parameters. M1 and M2 respectively represent the additional yaw moments of the tractor and the semi-trailer.

[0015] Preferably, the in-wheel motor model uses a second-order transfer function to describe the motor dynamic response.

[0016] Preferably, the longitudinal driver model calculates the total driving torque required for the vehicle to reach the target speed based on the PI control algorithm.

[0017] Preferably, the MIMO-IMFAC controller is for a multi-input multi-output discrete-time system:

[0018]

[0019] where, y m (k) represents the m-th state variable at time k, u n (k) represents the n-th control variable at time k, and f represents the functional relationship between the input and output variables;

[0020] Under the assumption that the system satisfies the generalized Lipschitz continuity and the boundedness of the pseudo-Jacobian matrix PJM, the pseudo-Jacobian matrix is introduced to transform the system into a CFDL model:

[0021] y(k + 1) = y(k) + Φ c (k)Δu(k);

[0022] where, y(k) represents the state variable matrix at time k, Δu(k) represents the control increment matrix, and Φ c (k) represents the pseudo-Jacobian matrix;

[0023] The pseudo-Jacobian matrix Φ c (k) is expressed as:

[0024]

[0025] By designing the control input criterion function J[u(k)] and the pseudo-Jacobian matrix estimation criterion function J[Φ c (k)]:

[0026] J[u(k)] = |y * (k + 1) - y(k + 1)| 2 + λ|u(k) - u(k - 1)|2 ;

[0027]

[0028] Among them, represents the estimated value of the pseudo-Jacobian matrix at time k, and y * represents the desired state quantity matrix;

[0029] Based on the optimization principle, partial derivatives are solved for it, and then the control input u(k) and the estimated value of the pseudo-Jacobian matrix are expressed as follows:

[0030]

[0031] Among them, Δy(k) represents the state quantity increment at time k, λ > 0 represents the penalty factor for the change of the control input, μ > 0 represents the smoothing factor for the estimation of the pseudo-Jacobian matrix, η ∈ (0, 2] represents the step size factor, and ρ ∈ (0, 1] represents the step size factor.

[0032] Preferably, in the torque optimization controller, the tire load utilization coefficient is:

[0033]

[0034] Among them, represents the tire load utilization coefficient, β represents the road adhesion coefficient, F x represents the tire longitudinal force, F y represents the tire lateral force, F z represents the tire vertical force;

[0035] The objective function of the tire load utilization coefficient is optimized by the quadratic programming algorithm:

[0036]

[0037] Among them, represents the average tire load utilization rate, represents the weighting factor of the average tire load utilization rate, J ρ represents the optimization objective function, i represents the serial number of each driving wheel of the tractor, represents the tire load utilization coefficient of each driving wheel;

[0038] At the same time, it satisfies the comprehensive constraints of the tire longitudinal driving torque constraint, the motor torque constraint, the total driving torque and the additional yaw torque:

[0039]

[0040] F x,i ≤ αF z,i (17);

[0041] T min ≤T xi ≤T max (18);

[0042]

[0043] Wherein, r represents the tire radius, α represents the road friction coefficient, T xi represents the torque applied to each driven wheel, B1 represents the wheelbase of the tractor, T min and T max respectively represent the minimum and maximum values allowed for the drive wheel torque, T 2r , T 2l , T 3r , T 3l respectively represent the driving torque of the left and right wheels of the second axle and the left and right wheels of the third axle of the tractor, F xi represents the longitudinal force received by each drive wheel of the tractor, F yi represents the lateral force received by each drive wheel of the tractor, F zi represents the vertical force received by each drive wheel of the tractor.

[0044] Preferably, a braking torque distribution logic of the tractor is set in the drive / brake torque distribution controller: according to the output of the torque optimization controller, the drive and brake torques of the second and third axles are distinguished, the drive torque is provided by the motor, and the braking torque is preferentially distributed to the regenerative braking torque.

[0045] Preferably, a braking torque distribution logic of the semi-trailer is set in the drive / brake torque distribution controller: according to the additional yaw moment output by the MIMO-IMFAC controller, when the additional yaw moment is positive, a braking torque is applied to the left wheel, and when it is negative, a braking torque is applied to the right wheel.

[0046] Preferably, data interaction interfaces are used for communication between the three-layer structures of the active stability controller module. The MIMO-IMFAC controller transmits the calculated additional yaw moment to the torque optimization controller through the interface, the torque optimization controller transmits the optimized torque distribution result to the drive / brake torque distribution controller through the interface, and the drive / brake torque distribution controller feeds back the actual torque distribution execution situation and vehicle state information to the MIMO-IMFAC controller and the torque optimization controller through the interface, forming a closed-loop control loop.

[0047] The present invention also provides a full-condition active stability control method for a tractor semi-trailer, including the following steps:

[0048] Construct a linear three-degree-of-freedom tractor semi-trailer model, a hub motor model, and a longitudinal driver model;

[0049] Collect the actual state data of the vehicle in real time, calculate the reference quantity according to the steady-state conditions of the system, correct it in combination with the road adhesion conditions, compare the reference quantity with the actual state data, and obtain the state error as the control input;

[0050] The MIMO-IMFAC controller receives the sensor data, calculates the pseudo-Jacobian matrix in real time, and calculates the additional yaw moment according to the control input criterion function and the pseudo-Jacobian matrix estimation criterion function;

[0051] The torque optimization controller receives the additional yaw moment and the total driving torque, and optimizes the torque distribution by using the quadratic programming algorithm;

[0052] The drive / brake torque distribution controller specifically distributes the drive and brake torques of the tractor and the semi-trailer according to the distribution result. For the tractor, the regenerative braking torque is preferentially distributed. When the regenerative braking torque is insufficient, it is supplemented by the electric motor braking EMB torque. For the semi-trailer, according to the positive and negative of the additional yaw moment output by the MIMO-IMFAC controller, the braking torque is reasonably distributed to the wheels on the corresponding side;

[0053] Continuously monitor the vehicle state, compare the actual state with the target state, and adjust the controller parameters according to the deviation.

[0054] Therefore, the present invention adopts the above-mentioned active stability control system and control method for a tractor semi-trailer, and the beneficial technical effects are as follows:

[0055] (1) The overall vehicle stability is significantly improved.

[0056] The MIMO-IMFAC controller designed by the present invention can effectively enhance the stability of the tractor semi-trailer system under different road surface conditions such as dry land and wet land. Whether it is to cope with lateral instability situations such as vehicle sideslip and fishtailing, or to prevent dangers such as rollover, the controller can make the vehicle state change smoothly after control, and the control effect is significantly better than that of the traditional PID controller.

[0057] Under dangerous working conditions such as high-speed double-lane change (DLC) and J-turn, the MIMO-IMFAC controller can significantly reduce the sideslip angle, yaw angular velocity and roll angle of the vehicle, and effectively avoid vehicle instability. In the high-speed double-lane change working condition, the vehicle needs to quickly change the driving direction, which is easy to generate large sideslip and yaw. The upper-layer MIMO-IMFAC controller can timely adjust the vehicle attitude to make the vehicle pass smoothly. In the J-turn working condition, the steering angle of the vehicle is large and the requirement for stability is higher. The controller of the present invention can also ensure the stable driving of the vehicle.

[0058] (2) Optimize the tire load rate.

[0059] The effective intervention of the torque optimization controller significantly optimizes the tire load rate of the drive shaft. On low-adhesion road surfaces, it increases the wheel torque reserve of the vehicle, avoids the jitter phenomenon during the control process, not only enhances the safety of the vehicle but also improves driving comfort.

[0060] By reasonably distributing the driving and braking torques, the load pressure on the tires is reduced, and the service life of the tires is extended. In traditional control methods, the load distribution of the tires may be uneven, resulting in excessive wear of some tires. However, the torque optimization controller of the present invention can make the load on each tire more balanced, reduce tire wear, and lower the usage cost.

[0061] (3) Superior dynamic response performance.

[0062] The MIMO-IMFAC controller has a short dynamic response time and a high dynamic response peak. It can quickly adapt to complex working conditions such as low-adhesion road surfaces, demonstrating strong real-time control capabilities and stability, ensuring that the vehicle can respond correctly quickly in case of emergencies.

[0063] The output of the controller can quickly track the target value, enabling the vehicle to adjust its attitude in a timely manner and maintain stable driving. When encountering sudden steering or road condition changes, the MIMO-IMFAC controller can calculate the appropriate control quantity within a short time, drive the vehicle to make corresponding adjustments, and avoid instability caused by untimely response.

[0064] (4) Improved system safety.

[0065] The active lateral stability controller designed in the present invention can accurately control the output during the emergency obstacle avoidance process of the tractor-semitrailer, enabling the system to always operate within the working limit of the actuator, ensuring the safety and stability of the system operation.

[0066] The controller can adjust the control strategy in real time according to the actual state of the vehicle, avoiding system failures caused by improper control. At the same time, the system is also equipped with a fault diagnosis and fault tolerance control unit, which can monitor the working status of each module and the output data of the sensors in real time. When a fault in a certain module or sensor is detected, it can issue an alarm in a timely manner and make adjustments according to the preset fault tolerance strategy to ensure that the vehicle can still maintain basic driving safety in case of partial faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 is a structural diagram of an active stability control system for a tractor-semitrailer under all working conditions according to the present invention;

[0068] Figure 2 is an overall architecture diagram of the active stability controller;

[0069] Figure 3 It is a drive / composite braking torque distribution logic diagram. Specific implementation manners

[0070] The technical solutions of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0071] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention belongs.

[0072] Embodiment 1

[0073] As Figure 1 shown, it is a structural diagram of an active stability control system for a tractor-semitrailer under all working conditions of the present invention, including a dynamic model module, a reference quantity determination module, and an active stability controller module.

[0074] (1) Dynamic model module: Construct a linear three-degree-of-freedom tractor-semitrailer model, a hub motor model, and a longitudinal driver model.

[0075] Among them, the dynamic equation of the linear three-degree-of-freedom tractor-semitrailer model is derived based on the Lagrange method and is used to calculate the yaw angular velocity and the sideslip angle of the center of mass of the tractor-semitrailer.

[0076] When constructing the linear three-degree-of-freedom tractor-semitrailer model, the following simplified assumptions are made for the vehicle:

[0077] (1) Ignoring the influence of aerodynamic force and road gradient: In most cases, the influence of aerodynamic force and road gradient on the vehicle dynamic characteristics is relatively small. By ignoring these factors, the complexity of the model can be simplified, and at the same time, it will not have a significant impact on the calculation of the main state variables.

[0078] (2) Not considering the pitching and rolling motions of the tractor and the semitrailer: In the study of lateral stability control, the influence of pitching and rolling motions on vehicle yaw and sideslip can be ignored to a certain extent, so that the research focus can be concentrated on the lateral motion, improving the pertinence and practicality of the model.

[0079] (3) Assuming symmetric mass distribution: Although the actual mass distribution of the tractor-semitrailer is not completely symmetric, it can be approximately regarded as symmetric within a certain error range, which can reduce the number of parameters in the model and facilitate the establishment and analysis of the model.

[0080] (4) Ignoring the longitudinal speed difference: In the short term, the longitudinal speed difference between the tractor and the semitrailer is relatively small and has little influence on lateral stability. Therefore, ignoring this speed difference can simplify the model.

[0081] (5) Limit the articulation angle within the normal range: When the articulation angle is too large, the dynamic characteristics of the vehicle will change significantly, exceeding the applicable range of this model. Therefore, limiting the articulation angle within the normal range can ensure the effectiveness of the model.

[0082] (6) Adopt a small-angle approximation for the tire sideslip angle and use a constant lateral stiffness to linearly simulate the tire behavior: In the case of small sideslip angles, the lateral force of the tire is approximately linearly related to the sideslip angle. Using a constant lateral stiffness can simplify the tire model while meeting certain accuracy requirements.

[0083] The dynamic equations of the linear three-degree-of-freedom tractor-semitrailer model are as follows:

[0084] Dynamic equations of the tractor:

[0085]

[0086] Dynamic equations of the semitrailer:

[0087]

[0088] Among them, m1 and m2 represent the masses of the tractor and the semitrailer respectively, u1 and u2 represent the longitudinal speeds of the tractor and the semitrailer respectively, represent the first-order derivatives of the yaw angular velocities of the tractor and the semitrailer respectively, represent the second-order derivatives of the yaw angular velocities of the tractor and the semitrailer respectively, represent the first-order derivatives of the sideslip angles of the centers of mass of the tractor and the semitrailer respectively. F1, F2, and F3 represent the lateral forces of the three simplified axles in the model, F4 represents the lateral interaction force between the tractor and the semitrailer at the fifth wheel, Γ represents the folding angle between the tractor and the semitrailer at the saddle, I 1z ,I 2z represent the yaw moments of inertia of the tractor and the semitrailer respectively, a, b, c, d, e represent geometric parameters, and M1 and M2 represent the additional yaw moments of the tractor and the semitrailer respectively.

[0089] The in-wheel motor model simplifies to obtain the second-order transfer function G(s) of the motor drive model by determining the main poles of the closed-loop system as:

[0090]

[0091] Among them, T mi represents the actual input drive / brake electromagnetic torque of each tractor drive wheel, Denote the desired input drive / brake electromagnetic torque of each tractor drive wheel, ξ represents the damping ratio related to the drive motor parameters, which is usually determined according to the structural parameters of the drive motor, and s represents the input variable. This second-order transfer function takes into account the dynamic response characteristics of the motor and can accurately describe the dynamic process of the actual motor tracking the ideal state curve.

[0092] The longitudinal driver model uses a PI control algorithm to calculate the total drive torque, and the formula is:

[0093] T d =K u [k p (u - u d ) + k i ∫(u - u d )dt](6);

[0094] Among them, T d represents the desired total drive torque, u represents the actual vehicle speed, u d represents the desired vehicle speed, K u represents the speed factor set by humans, k p represents the proportional coefficient, k i represents the integral coefficient. The proportional coefficient k p is used to quickly respond to the vehicle speed deviation, and the integral coefficient k i is used to eliminate the steady-state error. The two work together to enable the vehicle to stably approach the target speed.

[0095] (2) Reference quantity determination module: According to the condition that the derivative of the state variable is zero at the steady state of the system, calculate the steady-state value of the system state variable, and use it as the ideal value of the subsequent control system. The difference between the ideal value and the actual value is used as the input of the control system, and it is ensured that the steady-state value meets the road adhesion limit.

[0096] (3) Active stability controller module: It is divided into three layers. The upper layer is the MIMO-IMFAC controller, which dynamically linearizes the input and output variables according to the input variables such as the yaw rate and the sideslip angle of the center of mass of the tractor and the semi-trailer, and calculates and outputs the additional yaw moment based on the extreme value search rule; the middle layer is the torque optimization controller, which uses the quadratic programming method to optimize and distribute the additional yaw moment output by the upper layer controller and the total drive torque output by the longitudinal speed controller to the drive motor according to the tire friction ellipse constraint and the tire load utilization coefficient; the lower layer is the drive / brake torque distribution controller, which, for the configuration of the distributed drive of the middle and rear axles of the tractor and the distributed electric braking system of the whole vehicle, preferentially distributes the regenerative braking torque, and the insufficient part is supplemented by the electric motor braking torque, and reasonably distributes the braking torque of the wheels of the semi-trailer axle according to the additional yaw moment of the semi-trailer output by the upper layer controller.

[0097] The MIMO-IMFAC controller is for multi-input multi-output discrete-time systems:

[0098]

[0099] where y m (k) represents the m-th state variable at time k, and u n (k) represents the n-th control variable at time k, and f represents the functional relationship between the input and output variables;

[0100] Under the assumption that the system satisfies generalized Lipschitz continuity and boundedness of the pseudo-Jacobian matrix, the pseudo-Jacobian matrix (pseudo-Jacobian matrix) is introduced to transform the system into a compact format (CFDL) model:

[0101] y(k + 1) = y(k) + Φ c (k)Δu(k) (8);

[0102] where y(k) represents the state variable matrix at time k, and Δu(k) represents the control increment matrix; Φ c (k) represents the pseudo-Jacobian matrix;

[0103] The pseudo-Jacobian matrix Φ c (k) is expressed as:

[0104]

[0105] By designing the control input criterion function and the pseudo-Jacobian matrix estimation criterion function:

[0106] J[u(k)] = |y * (k + 1) - y(k + 1)| 2 + λ|u(k) - u(k - 1)| 2 (10);

[0107]

[0108] where, represents the estimated value of the pseudo-Jacobian matrix at time k, and y * represents the desired state variable matrix;

[0109] Based on the optimization principle, the partial derivative is solved for it, and then the expressions of the control input u(k) and the estimated value of the pseudo-Jacobian matrix are as follows:

[0110]

[0111] Among them, Δy(k) represents the increment of the state variable at time k, λ > 0 is the penalty factor for controlling the change of the input, which is used to limit the drastic change of the control input and improve the stability of the system; μ > 0 represents the smoothing factor for estimating the pseudo-Jacobian matrix, making the estimation of the pseudo-Jacobian matrix smoother and avoiding drastic fluctuations, η ∈ (0, 2] represents the step size factor, which provides greater flexibility and generality for the control algorithm, and ρ ∈ (0, 1] represents the step size factor, enhancing the generality and versatility of the control algorithm.

[0112] In the torque optimization controller, the tire load utilization coefficient is:

[0113]

[0114] Among them, represents the tire load utilization coefficient, β represents the road adhesion coefficient, F x represents the longitudinal tire force, F y represents the lateral tire force, F z represents the vertical tire force;

[0115] The objective function for optimizing the tire load utilization coefficient through the quadratic programming algorithm:

[0116]

[0117] Among them, represents the average tire load utilization rate, represents the weighting factor of the average tire load utilization rate, represents the objective function, i represents the serial number of each driving wheel of the tractor, represents the tire load utilization coefficient of each driving wheel.

[0118] Simultaneously satisfy the comprehensive constraints of the longitudinal tire driving torque constraint, the motor torque constraint, the total driving torque and the additional yaw moment:

[0119]

[0120] F x,i ≤αF z,i (17);

[0121] T min ≤T xi ≤T max (18);

[0122]

[0123] Among them, r represents the tire radius, α represents the road friction coefficient, T xi represents the torque applied to each driven wheel, B1 represents the wheelbase of the tractor, T min and Tmax respectively represent the minimum and maximum values allowed for the driving wheel torque, T 2r , T 2l , T 3r , T 3l respectively represent the driving torques of the left and right wheels of the second axle and the left and right wheels of the third axle of the tractor, F xi represents the longitudinal force received by each driving wheel of the tractor, F yi represents the lateral force received by each driving wheel of the tractor, F zi represents the vertical force received by each driving wheel of the tractor.

[0124] The longitudinal forces of the four driving wheels are regarded as independent variables, and the objective function is transformed into a standard quadratic programming problem for solution. The standard form is expressed as follows:

[0125]

[0126] U = [T 2l T 2r T 3l T 3r T (21);

[0127] b eq = [T x M1] T (22);

[0128]

[0129] In the equation, H represents the coefficient matrix of the quadratic term, C represents the coefficient matrix of the linear term, U represents the state quantity matrix, A eq represents the coefficient matrix of the equality constraint, b eq represents the constant term matrix of the equality constraint, A represents the coefficient matrix of the inequality constraint, and b represents the constant term matrix of the inequality constraint.

[0130] This optimization process aims to make the load utilization of each tire more balanced, improving the driving stability of the vehicle and the service life of the tires.

[0131] ​In the drive / brake torque distribution controller, the brake torque distribution logic of the tractor is as follows: regenerative braking torque is preferentially used. When the regenerative braking torque is insufficient, it is supplemented by the electric motor braking (EMB) torque. Specifically, according to the output of the torque optimization controller, the drive and brake torques of the second and third axles are distinguished. The drive torque is provided by the motor, and the brake torque is preferentially distributed with the regenerative braking torque. The brake torque distribution logic of the semi-trailer is as follows: according to the additional yaw moment output by the MIMO-IMFAC controller, when the additional yaw moment is positive, brake torque is applied to the left wheels, and when it is negative, brake torque is applied to the right wheels, so as to adjust the attitude of the semi-trailer and enhance the overall stability of the vehicle.

[0132] Data interaction interfaces are used for communication between the three-layer structures of the active stability controller module. The MIMO-IMFAC controller transmits the calculated additional yaw moment to the torque optimization controller through the interface, and the torque optimization controller passes the optimized torque distribution result to the drive / brake torque distribution controller through the interface. At the same time, the drive / brake torque distribution controller feeds back the actual torque distribution execution situation and vehicle state information to the MIMO-IMFAC controller and the torque optimization controller through the interface, forming a closed-loop control loop to ensure the real-time performance and accuracy of the entire lateral control process.

[0133] Embodiment 2

[0134] A method for active stability control of a tractor semi-trailer includes the following steps:

[0135] 1. Establish a dynamic model: According to the actual structure and parameters of the tractor semi-trailer, accurately construct a linear three-degree-of-freedom tractor semi-trailer model, a hub motor model, and a longitudinal driver model, and determine the values of various parameters in the models to ensure that the models can accurately reflect the dynamic behavior of the vehicle;

[0136] In this step, for the determination of the parameters of the linear three-degree-of-freedom tractor semi-trailer model, multiple actual working condition tests and data fitting are required. By testing the vehicle under different road conditions, different vehicle speeds, and different load conditions, collecting the vehicle's motion data, and then using data fitting methods such as the least squares method to accurately determine the mass, moment of inertia, geometric parameters, etc. in the model to improve the accuracy and reliability of the model.

[0137] 2. Determine the reference quantity: Real-time collect the actual state data of the vehicle, including yaw rate, sideslip angle of the center of mass, vehicle speed, etc. Calculate the reference quantity according to the system steady-state conditions, that is, the steady-state value of the state variable, and correct it in combination with the road adhesion conditions. Compare the reference quantity with the actual state data to obtain the state error, which is used as the input signal for the subsequent control system;

[0138] In this step, for the correction of road adhesion conditions, it is necessary to combine the real-time meteorological data of the vehicle's environment and the road surface detection information. For example, when it is detected that there is water accumulation, snow accumulation or icing on the road surface, the road friction coefficient is adjusted according to the meteorological data and the road surface detection information, and then the steady-state value of the system state variable is corrected to make the reference quantity more in line with the actual driving conditions.

[0139] 3. Active stability controller control:

[0140] The MIMO-IMFAC controller receives signals such as the yaw rate and sideslip angle of the center of mass of the tractor and semi-trailer collected by the sensors. According to the preset control algorithm, it calculates the pseudo-Jacobian matrix in real time, and calculates the additional yaw moment required for the tractor and semi-trailer based on the control input criterion function and the pseudo-Jacobian matrix estimation criterion function. During the calculation process, by adjusting parameters such as the weight factor and step size factor, the performance of the controller is optimized to ensure that the output additional yaw moment can accurately track the target state variable;

[0141] The torque optimization controller receives the additional yaw moment output by the MIMO-IMFAC controller and the total drive torque calculated by the longitudinal driver model. According to the tire friction ellipse constraint and the tire load utilization coefficient, the quadratic programming algorithm is used to optimize the torque distribution of the drive motor. During the optimization process, the longitudinal force, lateral force of the tire and the torque limit of the motor are monitored in real time to ensure that the distributed torque meets the requirements of vehicle stability and safety, and at the same time improves the tire load utilization efficiency;

[0142] The lower-layer drive / brake torque distribution controller specifically distributes the drive and brake torques of the tractor and semi-trailer according to the distribution results of the torque optimization controller. For the tractor, the regenerative braking torque is preferentially distributed. When the regenerative braking torque is insufficient, it is supplemented by the electric motor braking (EMB) torque. During the distribution process, the drive / brake torque distribution logic is strictly followed to ensure that each wheel obtains an appropriate torque. For the semi-trailer, according to the positive and negative of the additional yaw moment output by the upper-layer controller, the braking torque is reasonably distributed to the wheels on the corresponding side to achieve precise control of the vehicle attitude.

[0143] 4. System feedback adjustment: During the vehicle operation, continuously monitor the actual state of the vehicle, such as yaw rate, sideslip angle of the center of mass, tire load rate, etc. Compare these actual state data with the target state, and adjust the parameters of the controller in real time according to the deviation magnitude and change trend. For example, when it is found that the sideslip angle of the vehicle exceeds the allowable range, appropriately increase the control gain of the MIMO-IMFAC controller to enhance the vehicle attitude adjustment ability; when the tire load rate is too high, optimize the torque distribution strategy to reduce the tire load pressure and ensure that the vehicle is always in a stable and safe operation state;

[0144] In this step, an adaptive adjustment strategy is adopted to adjust the controller parameters. According to factors such as the magnitude, change rate, and duration of the vehicle state deviation, the parameters of the controller are dynamically adjusted. For example, when the vehicle state deviation is large and the change rate is fast, the control gain is increased to quickly correct the deviation; when the deviation duration is long, the integral coefficient is appropriately adjusted to eliminate the steady-state error, thereby improving the adaptive ability and control effect of the system.

[0145] It should be noted that the content not elaborated in detail in the present invention is prior art and well-known to those skilled in the art.

[0146] Therefore, the present invention adopts the above-mentioned active stability control system and control method for a tractor-semitrailer under all working conditions to solve the problem of poor stability of the tractor-semitrailer under working conditions such as emergency obstacle avoidance, improve the driving safety of the vehicle, and reduce the accident risk. By constructing a dynamic model, designing an advanced controller, and a reasonable torque distribution strategy, the vehicle can maintain stable driving under various complex working conditions, reducing the occurrence of dangerous situations such as side slip, fishtailing, and rollover. At the same time, the designed controller has strong real-time performance and low requirements for the dynamic model, and can adapt to the complex road environment of heavy-duty semitrailers.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A full-condition active stability control system for a tractor semi-trailer, characterized in that, Including: A kinetic model module, including a linear three-degree-of-freedom tractor-semitrailer model, a hub motor model, and a longitudinal driver model, which are respectively used to calculate the yaw angular velocity and the sideslip angle of the center of mass of the tractor-semitrailer, describe the dynamic response of the motor, and calculate the total driving torque; A reference quantity determination module, which is used to calculate the steady-state values of the system state variables according to the condition that the derivative of the state variables of the system is zero at steady state, and make corrections in combination with the road adhesion conditions; An active stability controller module, which has three layers. The upper layer is a MIMO-IMFAC controller, the middle layer is a torque optimization controller, and the lower layer is a drive / brake torque distribution controller, which are respectively used to calculate the additional yaw moment, optimize the tire load distribution, and adjust the braking torque.

2. The active stability control system for a tractor-semitrailer under all working conditions according to claim 1, wherein, The linear three-degree-of-freedom tractor-semitrailer model includes a tractor dynamics equation and a semitrailer dynamics equation; The tractor dynamics equation is as follows: The semitrailer dynamics equation is as follows: where \(m_1\) and \(m_2\) represent the masses of the tractor and the semi-trailer respectively, and \(u_1\) and \(u_2\) represent the longitudinal speeds of the tractor and the semi-trailer respectively. represent the first-order derivatives of the yaw angular velocities of the tractor and the semi-trailer respectively. represent the second-order derivatives of the yaw angular velocities of the tractor and the semi-trailer respectively. represent the first-order derivatives of the sideslip angles of the centers of mass of the tractor and the semi-trailer respectively. \(F_1\), \(F_2\), and \(F_3\) represent the lateral forces of the three simplified axles in the model, \(F_4\) represents the lateral interaction force between the tractor and the semi-trailer at the fifth wheel, \(\Gamma\) represents the folding angle between the tractor and the semi-trailer at the saddle, \(I\) 1z , \(I\) 2z represent the yaw moments of inertia of the tractor and the semi-trailer respectively. \(a\), \(b\), \(c\), \(d\), \(e\) represent geometric parameters, and \(M_1\), \(M_2\) represent the additional yaw moments of the tractor and the semi-trailer respectively.

3. A full-condition active stability control system for a tractor semi-trailer according to claim 1, characterized in that The hub motor model uses a second-order transfer function to describe the dynamic response of the motor.

4. The active stability control system for a tractor-semitrailer under all working conditions according to claim 1, characterized in that The longitudinal driver model calculates the total driving torque required for the vehicle to reach the target speed based on the PI control algorithm.

5. The active stability control system for a tractor semi-trailer under all working conditions according to claim 1, characterized in that, The MIMO-IMFAC controller is for a multi-input multi-output discrete-time system: where y m (k) represents the m-th state variable at time k, and u n (k) represents the n-th control variable at time k, and f represents the functional relationship between the input and output quantities; Under the assumption that the system has generalized Lipschitz continuity and the pseudo-Jacobian matrix is bounded, the pseudo-Jacobian matrix is introduced to transform the system into a CFDL model: y(k + 1) = y(k) + Φ c (k)Δu(k); Among them, y(k) represents the state quantity matrix at time k, Δu(k) represents the control increment matrix, and Φ c (k) represents the pseudo-Jacobian matrix; Pseudo-Jacobian matrix Φ c is expressed as: By designing the control input criterion function J[u(k)] and the pseudo-Jacobian matrix estimation criterion function J[Φ c (k)]: J[u(k)] = |y * (k + 1)-y(k + 1)| 2 + λ|u(k)-u(k - 1)| 2 ; Among them, represents the estimated value of the pseudo-Jacobian matrix at time k, and y * represents the matrix of expected state variables; The partial derivative is solved based on the optimization principle, and then the control input u(k) and the estimated value of the pseudo-Jacobian matrix are obtained The expressions are as follows: Where, Δy(k) represents the state quantity increment at time k, λ>0 represents the penalty factor for the change of the control input, μ>0 represents the smoothing factor for the estimation of the pseudo-Jacobian matrix, η∈(0,2] represents the step factor, and ρ∈(0,1] represents the step factor.

6. The active stability control system for a tractor-semitrailer under all working conditions according to claim 1, wherein, In the torque optimization controller, the tire load utilization coefficient is: Among them, represents the tire load utilization coefficient, β represents the road adhesion coefficient, and F x represents the longitudinal force of the tire, and F y represents the lateral force of the tire, and F z represents the vertical force of the tire; The objective function of the tire load utilization coefficient is optimized by the quadratic programming algorithm: wherein, represents the average tire load utilization rate, represents the weighting factor of the average tire load utilization rate, J ρ represents the optimization objective function, i represents the serial numbers of the drive wheels of the tractor, represents the tire load utilization coefficient of each drive wheel; At the same time, it satisfies the comprehensive constraints of the longitudinal driving torque of the tire, the motor torque constraint, the total driving torque, and the additional yaw moment: T min ≤T xi ≤T max (18); Among them, r represents the tire radius, α represents the road friction coefficient, T xi represents the torque applied to each driven wheel, B1 represents the wheelbase of the tractor, T min and T max respectively represent the minimum and maximum values allowed for the drive wheel torque, T 2r , T 2l , T 3r , T 3l respectively represent the driving torques of the left and right wheels of the second axle and the left and right wheels of the third axle of the tractor, F xi represents the longitudinal force received by each drive wheel of the tractor, F yi represents the lateral force received by each drive wheel of the tractor, F zi represents the vertical force received by each drive wheel of the tractor.

7. The active stability control system for a tractor-semitrailer under all working conditions according to claim 1, wherein, In the drive / brake torque distribution controller, there is a braking torque distribution logic for the tractor: according to the output of the torque optimization controller, the drive and brake torques of the second and third axles are distinguished. The drive torque is provided by the motor, and the braking torque is preferentially distributed to the regenerative braking torque.

8. The active stability control system for a tractor semi-trailer under all working conditions according to claim 1, wherein In the drive / brake torque distribution controller, there is a braking torque distribution logic for the semitrailer: according to the additional yaw moment output by the MIMO-IMFAC controller, when the additional yaw moment is positive, the braking torque is applied to the left wheels, and when it is negative, the braking torque is applied to the right wheels.

9. The active stability control system for a tractor-semitrailer under all working conditions according to claim 1, characterized in that, The three-layer structure of the active stability controller module communicates through a data interaction interface. The MIMO-IMFAC controller transmits the calculated additional yaw moment to the torque optimization controller through the interface. The torque optimization controller transmits the optimized torque distribution result to the drive / brake torque distribution controller through the interface. The drive / brake torque distribution controller feeds back the actual torque distribution execution situation and vehicle state information to the MIMO-IMFAC controller and the torque optimization controller through the interface, forming a closed-loop control loop.

10. A full-condition active stability control method for a tractor semi-trailer, characterized in that Including the following steps: Construct a linear three-degree-of-freedom tractor-semitrailer model, a hub motor model, and a longitudinal driver model; Collect the actual state data of the vehicle in real time, calculate the reference quantity according to the steady-state conditions of the system, and correct it in combination with the road adhesion conditions. Compare the reference quantity with the actual state data to obtain the state error as the control input; The MIMO-IMFAC controller receives the sensor data, calculates the pseudo-Jacobian matrix in real time, and calculates the additional yaw moment according to the control input criterion function and the pseudo-Jacobian matrix estimation criterion function; The torque optimization controller receives the additional yaw moment and the total driving torque, and optimizes the torque distribution using the quadratic programming algorithm; The drive / brake torque distribution controller makes specific distributions of the drive and brake torques of the tractor and the semi-trailer according to the distribution results. For the tractor, the regenerative braking torque is preferentially distributed. When the regenerative braking torque is insufficient, it is supplemented by the electric motor braking EMB torque. For the semi-trailer, according to the positive and negative of the additional yaw moment output by the MIMO-IMFAC controller, the braking torque is reasonably distributed to the wheels on the corresponding side; Continuously monitor the vehicle state, compare the actual state with the target state, and adjust the controller parameters according to the deviation.

Images