Wheel speed reconstruction and multi-source collaborative brake control method for heavy multi-axle vehicle

Abstract

The application relates to a brake control method for heavy multi-axle vehicles, which is characterized in that the brake control method comprises a wheel speed sensor fault self-diagnosis and virtual reconstruction module, a brake lateral and longitudinal coordination distribution module and a multi-source brake optimization distribution module; under the premise of ensuring wheel speed measurement redundancy, the brake control method realizes layered coordination distribution of brake force among wheels and among multi-sources, and improves brake safety, stability and smoothness of heavy multi-axle vehicles under sensor fault, load transfer and complex working conditions.

Description

Technical Field

[0001] This invention relates to the field of vehicle braking control, and more particularly to a braking control method for wheel speed reconstruction and multi-source coordination for heavy-duty multi-axle vehicles. Background Technology

[0002] With the rapid development of the national economy and the continuous improvement of the transportation system, heavy-duty multi-axle vehicles are widely used in physical transportation, engineering construction, national defense equipment, and emergency rescue. Heavy-duty multi-axle vehicles have a large number of axles, varying wheelbases, complex load distribution, and diverse operating conditions. Their braking safety, stability, and controllability directly affect the overall safety performance and operational efficiency of the vehicle. Therefore, constructing a highly reliable, robust, and adaptable braking control system for heavy-duty multi-axle vehicles is a core technological guarantee for improving vehicle active safety and preventing rollovers and brake failures, and has significant engineering and application value.

[0003] However, existing braking control methods for heavy multi-axle vehicles based on traditional rules (such as anti-lock braking systems (ABS) and electronic braking systems (EBS) typically use wheel speed signal acquisition, braking force distribution decision-making, and actuator drive as their basic framework. On the one hand, the dynamic coupling relationship between each axle and the left and right wheels during the braking process of heavy multi-axle vehicles is strong and the nonlinear characteristics are significant, making it easy for dangerous conditions such as wheel lock-up, sideslip, and rollover to occur. On the other hand, the frequent transfer of vehicle load and the drastic changes in brake temperature lead to strong uncertainty in the key parameters of the braking system, making it difficult for traditional control or static distribution strategies to achieve precise and adaptive braking force regulation.

[0004] In recent years, although related patents have explored braking control, such as Chinese patent CN119370069A which discloses a method and device for anti-lock braking control of multi-axle vehicles, it switches between ABS depressurization / pressure holding / pressure boosting modes based on multiple conditions by acquiring slip degree, wheel speed deceleration, and wheel speed vector. Another example is Chinese patent CN116946113A, which proposes a lateral and longitudinal coupling integrated control method for multi-axle special vehicles, employing an upper-level LQR + sliding mode motion controller, and a lower-level optimal longitudinal force distribution with the goal of minimizing the total tire load rate. Yet another example is Chinese patent CN116588055A, which discloses a composite braking distribution method and device for multi-axle commercial vehicles, determining the braking mode based on driving behavior data and considering motor efficiency and battery charging efficiency to distribute regenerative braking and hydraulic braking force. While these methods offer some performance improvements compared to traditional braking control strategies, they still have some shortcomings and limitations, mainly manifested in the following ways:

[0005] (1) In the existing wheel speed measurement scheme, the hardware redundancy cost is high and the layout is difficult. Simple signal verification cannot provide an effective alternative signal after sensor failure. The wheel speed estimation method based on dynamic model has not yet established a fault self-diagnosis and hierarchical processing mechanism, and cannot realize a complete redundant link from fault identification to signal reconstruction.

[0006] (2) Existing braking force distribution methods for heavy multi-axle vehicles are mostly static distribution based on axle load ratio or ABS threshold control based on slip ratio. They only take wheel slip ratio as the control target and do not consider the dynamic changes in the center of gravity position with load transfer. The calculation of its lateral load transfer rate depends on static or simplified assumptions, which makes it difficult to accurately reflect the actual rollover risk. At the same time, due to the lack of a mechanism to actively generate the desired yaw moment, its differential braking capability is limited to preventing wheel lock-up and cannot coordinate the inherent contradiction between braking efficiency and rollover suppression under a unified framework.

[0007] (3) Existing multi-source braking control systems for heavy-duty multi-axle vehicles mostly adopt priority switching or fixed ratio allocation strategies, which are based solely on a simple comparison between braking demand and auxiliary braking availability. They do not introduce brake temperature as a decision variable and cannot predict the impact of heat fade on braking performance. At the same time, they lack multi-objective dynamic optimization capabilities. The timing and ratio of switching between auxiliary braking and air pressure braking are entirely determined by preset logic, making it difficult to adapt to complex working conditions such as long downhill slopes and continuous braking. This makes it difficult to avoid problems such as brake force decay, brake overheating, and retarder gear shock. Summary of the Invention

[0008] In view of the shortcomings of the prior art, the purpose of this invention is to provide a braking control method for wheel speed reconstruction and multi-source coordination for heavy multi-axle vehicles. This braking control method achieves hierarchical and coordinated distribution of braking force between wheels and among multiple sources while ensuring wheel speed measurement redundancy, thereby improving the braking safety, stability and ride comfort of heavy multi-axle vehicles under sensor failure, load transfer and complex working conditions.

[0009] The specific solution of the present invention is as follows:

[0010] A braking control method for heavy-duty multi-axle vehicles that integrates wheel speed reconstruction and multi-source coordination, characterized by comprising:

[0011] Wheel speed sensor fault self-diagnosis and virtual reconstruction module: acquires measured wheel speed signals characterizing the rotational speed of each wheel, calculates wheel speed estimates for each wheel based on the vehicle's longitudinal dynamics; generates a reliability evaluation index characterizing the reliability of the wheel speed signals based on the comparison between the measured wheel speed signals and the wheel speed estimates; when the reliability evaluation index meets preset fault judgment conditions, the corresponding wheel speed sensor is determined to be faulty, and fault level information is output; for the wheel determined to be faulty, the measured wheel speed signals of other wheels not determined to be faulty and the vehicle inertial measurement signal are used as inputs, and a reconstructed wheel speed signal for the faulty wheel is generated through a nonlinear mapping relationship reflecting the kinematic constraints between wheels, forming wheel speed measurement redundancy;

[0012] The braking lateral and longitudinal coordinated distribution module receives the reconstructed wheel speed signal and the fault level information; it fuses the suspension pressure signal and the signal characterizing the boom posture to calculate the real-time center of gravity position of the vehicle; based on the torque balance relationship and the real-time center of gravity position, it calculates the lateral load transfer rate and assesses the rollover risk; based on the deviation of the actual yaw rate from the reference expected value, it calculates the expected yaw moment through a feedback correction loop; with the goals of minimizing the lateral load transfer rate, minimizing the tracking deviation of the expected yaw moment, and balancing the longitudinal force load of each wheel tire, it solves the target longitudinal braking force of each wheel through numerical optimization under the conditions of meeting the total braking force requirement and the upper limit constraint of the braking force of each wheel.

[0013] Multi-source braking optimization allocation module: Receives the target longitudinal braking force of each wheel; based on the brake temperature dynamic model and engine braking torque characteristic function, performs feedforward estimation of brake temperature rise in the prediction time domain; constructs a cost function; wherein the first weight related to the proportion of auxiliary braking usage in the cost function is configured as an increasing function of continuous downhill duration, the second weight related to the highest temperature among all brakes is configured as an increasing function of brake temperature, and the third weight related to the degree of retarder gear change is configured as a decreasing function of rollover risk index; under the conditions of satisfying the total braking force balance constraint, the output capacity constraint of each braking source, the upper limit constraint of brake temperature, and the retarder shift smoothness constraint, solves the optimal control sequence, outputs engine braking control commands, hydraulic retarder gear commands, and axle air pressure power commands, realizing the coordinated allocation of multi-source braking forces.

[0014] Further, in step S1: In the wheel speed sensor fault self-diagnosis and virtual reconstruction module, based on the vehicle's longitudinal dynamics, the vehicle's longitudinal acceleration and historical vehicle speed information are used to calculate the estimated vehicle speed at the current moment, and the wheel speed estimates of each wheel are calculated based on the estimated vehicle speed and the rolling radius of each wheel; the estimated wheel speed is compared with the sensor's measured value, and a continuously changing reliability evaluation index is generated based on the deviation; a reliability threshold and a duration threshold are set, and when the reliability evaluation index is lower than the reliability threshold and the duration exceeds the duration threshold, the wheel speed sensor is determined to be faulty, and the fault level is output; for the wheel determined to be faulty, the measured wheel speed, vehicle longitudinal acceleration, yaw rate, and steering wheel angle of the other wheels not determined to be faulty are used as inputs, and a reconstructed wheel speed signal is generated through a nonlinear mapping model reflecting the kinematic constraints between wheels; based on the reliability evaluation index of each wheel, the normal wheel speed signal and the reconstructed wheel speed signal are weighted and synthesized, and the wheel speed signal finally used by each wheel is output;

[0015] Step S2: In the braking lateral and longitudinal coordinated distribution module, the final wheel speed signals and reliability evaluation indicators of each wheel output from step S1 are received; using the final wheel speed signals and reliability evaluation indicators, the longitudinal vehicle speed is calculated using a weighted average method, and the yaw rate and center of gravity sideslip angle are calculated through the speed difference between the left and right wheels; the suspension pressure signal and vehicle attitude angle signals including pitch and roll angles are acquired, and the center of gravity height and longitudinal position are dynamically calculated based on noise characteristic corrections associated with the reliability evaluation indicators of each wheel; the lateral load transfer rate is calculated based on the vertical load of each wheel, wherein the calculation of the vertical load of each wheel comprehensively considers the static load and longitudinal load. Load transfer caused by acceleration and lateral acceleration; the calculated actual yaw rate is compared with the expected yaw rate determined based on the driver's steering intention and vehicle speed, and the expected yaw moment is calculated through a feedback correction loop; using the braking force of each wheel as the optimization variable, a cost function is constructed to minimize the comprehensive lateral load transfer rate, minimize the yaw moment tracking deviation, and balance the longitudinal force load of each wheel tire. When the reliability evaluation index of a certain wheel is low, a penalty term for the change in the braking force distribution of the corresponding wheel is added; under the constraints that the total braking force meets the requirements, the braking force of each wheel does not exceed the upper limit, and the actuator change rate is limited, the target braking force of each wheel is solved through numerical optimization methods;

[0016] Step S3: In the multi-source braking optimization allocation module, the target braking force requirements for each wheel and vehicle operating status information output in step S2 are received. The vehicle operating status information includes wheel speed, brake temperature, engine speed, current retarder gear position, and rollover risk index. Based on the brake temperature dynamic model, the temperature evolution of each brake in the prediction time domain is predicted. A cost function is constructed with the objectives of maximizing the auxiliary braking usage ratio, minimizing the highest temperature in each brake, and smoothing the retarder gear position change. The weight of each objective term is dynamically adjusted according to the current operating conditions. Under the conditions of satisfying the total braking force balance constraint, engine braking capacity constraint, hydraulic retarder braking capacity constraint, air pressure braking capacity constraint for each wheel, brake temperature safety upper limit constraint, and retarder gear position change rate constraint, the optimal control sequence is solved by rolling time domain optimization, and the engine braking control command, hydraulic retarder gear position command, and air pressure power command for each axle are output.

[0017] Specifically, the weights related to brake temperature in the cost function are configured as a function of brake temperature, and the proportion of this weight in the total weight increases continuously and monotonically as the brake temperature rises from the normal range to the danger range; the weights related to the auxiliary braking ratio in the cost function are configured as a function of the duration of continuous downhill driving, and the proportion of this weight in the total weight increases continuously and monotonically as the duration of continuous downhill driving increases.

[0018] The innovative points and advantages of this invention are as follows:

[0019] (1) A method for self-diagnosis and virtual reconstruction of wheel speed sensor faults for heavy multi-axle vehicles is proposed. The wheel speed estimate is calculated by the vehicle longitudinal dynamics model and compared with the sensor measured value. A continuously changing reliability evaluation index is constructed based on the deviation. When the reliability index is lower than the threshold and lasts for a certain period of time, the sensor is judged to be faulty and the fault level is output. At the same time, based on the nonlinear mapping relationship reflecting the kinematic constraints between wheels, the wheel speed signal of the non-faulty wheel and the vehicle inertial measurement signal are used as input to reconstruct the wheel speed of the faulty wheel and output the reconstructed wheel speed and fault level, so as to realize the redundancy guarantee of wheel speed measurement of heavy multi-axle vehicles under complex working conditions.

[0020] (2) A method for coordinated lateral and longitudinal braking distribution for heavy multi-axle vehicles is proposed. By using a data fusion method based on signal quality weighting, the suspension pressure and boom attitude are fused to dynamically identify the real-time center of gravity position of the vehicle. The lateral load transfer rate is calculated based on the torque balance relationship to assess the rollover risk in real time. At the same time, the desired yaw moment is calculated based on the yaw rate deviation feedback. With minimizing the lateral load transfer rate, tracking the desired yaw moment, and balancing the longitudinal force load of each wheel tire as multiple objectives, and with the braking force of each wheel as the optimization variable, the target braking force of each wheel is solved by numerical optimization under the premise of meeting the total braking force requirement and the capability constraints of each wheel actuator, so as to achieve the coordinated distribution of braking force between the wheels.

[0021] (3) A multi-source braking optimization allocation method for heavy multi-axle vehicles is proposed. The total braking force demand of each wheel is received from the innovation point (2). Based on the brake temperature model and the engine braking torque characteristic function, a feedforward-predictive dynamic adjustment framework based on the thermal model is adopted to construct an optimization function with multiple objectives, namely maximizing the auxiliary braking usage ratio, minimizing the highest brake temperature, and smoothing the retarder gear shift. The weight of each objective is dynamically adjusted according to the current working condition. Under the conditions of satisfying the total braking force balance, the capacity constraints of each braking source, the upper limit constraint of brake temperature, and the smoothness constraint of retarder shift, the optimal control sequence is solved, and the engine braking force, hydraulic retarder gear and each axle air pressure power command are output to realize the coordinated optimization allocation of multi-source braking force. Attached Figure Description

[0022] The present invention will be further described below with reference to the accompanying drawings;

[0023] Figure 1 This is a schematic diagram of the principle of the present invention;

[0024] Figure 2 This is the characteristic curve of the retarder braking torque versus rotational speed in an embodiment of the present invention;

[0025] Figure 3 This is the engine braking power and speed characteristic curve of a heavy-duty multi-axle vehicle in an embodiment of the present invention;

[0026] Figure 4 This is a diagram showing the off-highway driving gears, braking torque, and vehicle speed of a heavy multi-axle vehicle in an embodiment of the present invention;

[0027] Figure 5 This is a diagram showing the road driving gears, braking torque, and vehicle speed of a heavy multi-axle vehicle in an embodiment of the present invention;

[0028] Figure 6 This is the vertical load transfer model in the embodiments of the present invention;

[0029] Figure 7 This is a diagram illustrating the effect of the longitudinal dynamics model in an embodiment of the present invention;

[0030] Figure 8 This is the Trucsim model in the embodiments of the present invention;

[0031] Figure 9 This is a diagram illustrating the wheel speed fault diagnosis and wheel speed reconstruction effect under the condition of single-wheel sensor failure in a heavy multi-axle vehicle according to an embodiment of the present invention. Detailed Implementation

[0032] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0033] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0034] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0035] The present invention provides a wheel speed reconstruction and multi-source cooperative braking control system for heavy-duty multi-axle vehicles, including a wheel speed sensor fault self-diagnosis and virtual reconstruction module, a braking lateral and longitudinal cooperative allocation module, and a multi-source braking optimization allocation module.

[0036] The specific working method of the wheel speed sensor fault self-diagnosis and virtual reconstruction module includes: acquiring measured wheel speed signals characterizing the rotational speed of each wheel, calculating wheel speed estimates for each wheel based on the vehicle's longitudinal dynamics; generating a reliability evaluation index characterizing the reliability of the wheel speed signals based on the comparison results between the measured wheel speed signals and the wheel speed estimates; determining that the corresponding wheel speed sensor has failed when the reliability evaluation index meets preset fault judgment conditions, and outputting fault level information; for the wheel determined to be faulty, using the measured wheel speed signals of other wheels not determined to be faulty and the vehicle inertial measurement signal as input, generating a reconstructed wheel speed signal for the faulty wheel through a nonlinear mapping relationship reflecting the kinematic constraints between wheels, thus forming wheel speed measurement redundancy.

[0037] The specific working method of the braking lateral and longitudinal coordinated distribution module includes: receiving the reconstructed wheel speed signal and the fault level information; fusing the suspension pressure signal and the signal characterizing the boom posture to calculate the real-time center of gravity position of the vehicle; calculating the lateral load transfer rate based on the torque balance relationship and the real-time center of gravity position to assess the rollover risk; calculating the expected yaw moment through a feedback correction loop based on the deviation of the actual yaw rate from the reference expected value; and, with the goals of minimizing the lateral load transfer rate, minimizing the tracking deviation of the expected yaw moment, and balancing the longitudinal force load of each wheel tire, solving the target longitudinal braking force of each wheel through numerical optimization under the conditions of meeting the total braking force requirement and the upper limit constraint of the braking force of each wheel.

[0038] The specific working method of the multi-source braking optimization allocation module includes: receiving the target longitudinal braking force of each wheel; based on the brake temperature dynamic model and the engine braking torque characteristic function, performing feedforward prediction of brake temperature rise in the prediction time domain; constructing a cost function; wherein the first weight related to the auxiliary braking usage ratio in the cost function is configured as an increasing function of the continuous downhill duration, the second weight related to the highest temperature among the brakes is configured as an increasing function of brake temperature, and the third weight related to the degree of retarder gear change is configured as a decreasing function of the rollover risk index; under the conditions of satisfying the total braking force balance constraint, the output capacity constraint of each braking source, the upper limit constraint of brake temperature, and the retarder shift smoothness constraint, solving the optimal control sequence, and outputting the engine braking control command, the hydraulic retarder gear command, and the axle air suppression power command to achieve the coordinated allocation of multi-source braking force.

[0039] All three modules are deployed in the vehicle main controller (VCU), which is equipped with an NXP MPC5644 chip. Vehicle network communication is based on the J1939 protocol, with various signals interacting in real time via the CAN bus. The sampling period for wheel speed signals, IMU signals, suspension pressure signals, and brake temperature signals is uniformly set to 10 ms. The modules are decoupled using standardized signal interfaces, supporting independent upgrades and replacements.

[0040] In terms of the overall vehicle control architecture, the VCU is responsible for upper-level coordination and functional control; the TCU is responsible for transmission control; the engine and retarder are managed by the ECU; and the service braking system is controlled by the EBS. This architecture can well adapt to the multi-controller collaboration requirements of 8-axle vehicles and has strong engineering feasibility and modular expansion capabilities.

[0041] The present invention relates to a specific control method for a braking control system for wheel speed reconfiguration and multi-source coordination of heavy-duty multi-axle vehicles, comprising the following steps.

[0042] Step S1: In the wheel speed sensor fault self-diagnosis and virtual reconstruction step, based on the vehicle's longitudinal dynamics, the vehicle speed estimate at the current moment is calculated using the vehicle's longitudinal acceleration and historical vehicle speed information. The wheel speed estimate for each wheel is then calculated based on the vehicle speed estimate and the rolling radius of each wheel. The wheel speed estimate is compared with the sensor's measured value, and a continuously changing reliability evaluation index is generated based on the deviation. A reliability threshold and a duration threshold are set. When the reliability evaluation index is lower than the reliability threshold and the duration exceeds the duration threshold, the wheel speed sensor is determined to be faulty, and the fault level is output. For the wheel determined to be faulty, the measured wheel speed, vehicle longitudinal acceleration, yaw rate, and steering wheel angle of the other wheels not determined to be faulty are used as inputs. A reconstructed wheel speed signal is generated through a nonlinear mapping model reflecting the kinematic constraints between wheels. Based on the reliability evaluation index of each wheel, the normal wheel speed signal and the reconstructed wheel speed signal are weighted and synthesized to output the final wheel speed signal used by each wheel.

[0043] The specific implementation of wheel speed sensor fault self-diagnosis and virtual reconstruction in step S1 includes:

[0044] Establishing a longitudinal dynamics model for the vehicle, neglecting air resistance and rolling resistance, the longitudinal motion equation can be expressed as:

[0045]

[0046] In the formula, For the overall vehicle quality, For longitudinal vehicle speed, For the first Axis No. Side tire longitudinal force, For total braking force, This refers to the number of axles in the vehicle.

[0047] Based on the above longitudinal dynamics model, using the vehicle's longitudinal acceleration Using historical vehicle speed information, calculate the estimated longitudinal vehicle speed at the current moment. :

[0048]

[0049] In the formula, Sampling time.

[0050] Based on the estimated vehicle speed and the rolling radius of each wheel Calculate the estimated wheel speeds for each wheel. :

[0051]

[0052] The residual calculation is the actual wheel speed measured by the sensor. With estimated wheel speed Difference:

[0053]

[0054] Map the residuals to continuous reliability weights:

[0055]

[0056] In the formula, For the reliability of wheel speed, To adjust the parameters.

[0057] Set threshold With duration ,when And the duration exceeds The system can determine sensor malfunctions in real time and output the malfunction level. The system uses a 0-point scale to represent normal operation, a 1-point scale to represent a warning, and a 2-point scale to represent failure. When a wheel is in a normal state, the reliability evaluation index is close to 1, and the final wheel speed is taken as the measured value. When a sensor fails, the reliability evaluation index tends to 0, and the final wheel speed smoothly switches to the reconstructed value, thus achieving redundancy protection for the wheel speed measurement signal.

[0058] For the first one that is determined to be faulty Axis No. To reconstruct the wheel speed, this embodiment uses a three-layer feedforward neural network, which includes an input layer, a hidden layer, and an output layer. The number of neurons in each layer is as follows: , and .

[0059] The input and output of the input layer are as follows:

[0060]

[0061] In the formula, For input variables, ,in , and These are the vehicle's longitudinal acceleration, yaw rate, and steering angle, respectively; the rest... , The sensors measure wheel speeds of the non-faulty wheels on each axle of a heavy-duty multi-axle vehicle. The sensor measures the number of faulty wheels.

[0062] The input and output of the hidden layer are represented as follows:

[0063]

[0064] In the formula, For input dimensions, For the hidden layer bias, The number of neurons in the hidden layer. The weight coefficients are the input variables for the hidden layer. is the activation function of the hidden layer, and its expression is an asymmetric piecewise linear function.

[0065]

[0066] In the formula, For output layer weights, For output layer bias, Let be the activation function of the output layer, and let its expression be a linear function. To reconstruct wheel speed output .

[0067] Wheel speed weighted fusion:

[0068]

[0069] In the formula, To output the final speed of each wheel, This is the reliability weight (the weight is 0 when the sensor fails).

[0070] Step S2: In the braking lateral and longitudinal coordinated distribution step, the final wheel speed signals and reliability evaluation indicators of each wheel output from step S1 are received; using the final wheel speed signals and reliability evaluation indicators, the longitudinal vehicle speed is calculated using a weighted average method, and the vehicle yaw rate and center of gravity sideslip angle are calculated through the speed difference between the left and right wheels; the suspension pressure signal and vehicle attitude angle signals including pitch and roll angles are acquired, and the center of gravity height and longitudinal position are dynamically calculated based on noise characteristic corrections associated with the reliability evaluation indicators of each wheel; the lateral load transfer rate is calculated based on the vertical load of each wheel, wherein the calculation of the vertical load of each wheel comprehensively considers the static load and longitudinal load. Load transfer caused by acceleration and lateral acceleration; the calculated actual yaw rate is compared with the expected yaw rate determined based on the driver's steering intention and vehicle speed, and the expected yaw moment is calculated through a feedback correction loop; using the braking force of each wheel as the optimization variable, a cost function is constructed to minimize the comprehensive lateral load transfer rate, minimize the yaw moment tracking deviation, and balance the longitudinal force load of each wheel tire. When the reliability evaluation index of a certain wheel is low, a penalty term for the change in the braking force distribution of the corresponding wheel is added; under the constraints that the total braking force meets the requirements, the braking force of each wheel does not exceed the upper limit, and the actuator change rate is limited, the target braking force of each wheel is solved through numerical optimization methods.

[0071] The method for implementing the lateral and longitudinal braking coordinated distribution in step S2 is as follows:

[0072] (1) Vehicle motion state estimation based on wheel speed

[0073] The final wheel speeds of each wheel obtained from step S1 With reliability weight Calculate the longitudinal speed of the vehicle yaw rate and centroid side slip angle :

[0074] Among them, the longitudinal vehicle speed estimation adopts a weighted average method, with reliability weights. As a basis for confidence level:

[0075]

[0076] In the formula, For heavy multi-axle vehicles, the rolling radius of each wheel. For heavy-duty multi-axle vehicles, this refers to the number of axles.

[0077] The yaw rate is calculated using the speed difference between the left and right wheels, and is also weighted using reliability factors.

[0078]

[0079] In the formula, For the first Wheelbase.

[0080] Among them, the centroid side slip angle Lateral velocity calculated based on kinematic relationships By measuring lateral acceleration (Measured by IMU) Integrated, and the yaw rate calculated based on wheel speed is used. Drift.

[0081] (2) Dynamic identification of the centroid position

[0082] Integrated suspension pressure With attitude angle (including pitch angle) and roll angle Based on noise characteristic corrections associated with reliability evaluation indicators for each round, the centroid height and longitudinal position are dynamically calculated. This example uses weighted least squares to estimate the real-time centroid height. With longitudinal position Perform dynamic calculation.

[0083]

[0084] In the formula, The observation matrix consists of suspension forces and attitude angles. The weight matrix values ​​take into account the reliability weights of each round. Regarding suspension pressure measurement noise, when the reliability weight of a certain wheel is low, the confidence level of the corresponding suspension pressure data decreases accordingly. This is the torque average residual vector.

[0085] (3) Calculation of lateral load transfer rate

[0086] Based on the vertical load of each wheel Calculate the lateral load transfer rate (LTR):

[0087]

[0088] Among them, vertical load The calculations consider static load, load transfer caused by longitudinal acceleration, load transfer caused by lateral acceleration, and load transfer caused by... The instantaneous load fluctuation reflected by the calculated wheel speed difference:

[0089]

[0090] When the LTR exceeds the preset rollover warning threshold, a rollover warning is triggered.

[0091] (4) Calculation of desired yaw moment

[0092] The actual yaw rate (Obtained from step (1)) and the expected value (Based on the driver's steering intention and vehicle speed) a comparison is made, and the desired yaw moment is calculated through a feedback correction process. The desired yaw moment is calculated through a feedback correction mechanism, which is configured to generate a correction amount based on the deviation between the actual yaw rate and the desired yaw rate.

[0093] (5) Multi-objective optimization allocation

[0094] The optimization variable is the braking force of each wheel. The cost function comprehensively considers the following: minimizing the target lateral load transfer rate (LTR); minimizing the tracking deviation of the actual yaw moment from the desired yaw moment; and balancing the longitudinal force load on each tire. This embodiment uses sequential quadratic programming:

[0095]

[0096] In the formula, This is the actual yaw moment. As a weighting factor, when the reliability weight of a certain wheel is low, the penalty for the braking force distribution of that wheel is increased to avoid over-reliance on the braking of that wheel when the signal is unreliable. This is the weighting factor.

[0097] Among the constraints is the balance of total braking force: Maximum braking force of each wheel: Actuator rate limit: The optimal solution of the cost function under the constraints is obtained by solving the numerical optimization method to obtain the target braking force for each round.

[0098] Step S3: In the multi-source braking optimization allocation step, the target braking force demand for each wheel and vehicle operating status information output from step S2 are received. The vehicle operating status information includes wheel speed, brake temperature, engine speed, current retarder gear position, and rollover risk index. Based on the brake temperature dynamic model, the temperature evolution of each brake in the prediction time domain is predicted. A cost function is constructed with the objectives of maximizing the auxiliary braking usage ratio, minimizing the highest temperature among each brake, and smoothing the retarder gear position changes. The weights of each objective item are dynamically adjusted according to the current operating conditions. Specifically, the weights related to brake temperature in the cost function are configured as functions of brake temperature, and when... When the brake temperature rises from the normal range to the danger range, the proportion of this weight in the total weight increases continuously and monotonically. The weight related to the auxiliary braking ratio in the cost function is configured as a function of the duration of continuous downhill driving, and the proportion of this weight increases continuously and monotonically as the duration of continuous downhill driving increases. Under the conditions of satisfying the total braking force balance constraint, engine braking capacity constraint, hydraulic retarder braking capacity constraint, each wheel air pressure braking capacity constraint, brake temperature safety upper limit constraint, and retarder gear change rate constraint, the optimal control sequence is solved by rolling time domain optimization, and the engine braking control command, hydraulic retarder gear command, and each axle air pressure power command are output.

[0099] The method for implementing multi-source braking optimization allocation in step S3 is as follows:

[0100] Receive the total braking force requirement of each wheel output in step S2 and vehicle operating status information, including wheel speeds. Brake temperature Engine speed Retarder current gear Side rollover risk index wait.

[0101] (1) Brake temperature model

[0102] The temperature dynamics of each brake are described using a lumped-parameter thermal model. Axis No. The temperature change rate of the side brake is:

[0103]

[0104] In the formula, Brake heat capacity (unit: J / K). It is the frictional heat distribution coefficient (the proportion of braking energy converted into heat). For the first Axis No. Braking force generated by the side brake (unit: N). For the first Axis No. Side wheel speed (unit: m / s) The convective heat transfer coefficient (unit: W / (m²)) K)), The brake cooling area (unit: m²). Ambient temperature (in K).

[0105] Discretizing the above equations allows for feedforward prediction of the temperature of each brake in the prediction time domain:

[0106]

[0107] (2) Engine braking characteristic model

[0108] Based on the engine braking characteristic curve, an engine braking torque relationship model is constructed, and the engine braking torque is expressed as:

[0109]

[0110] In the formula, This refers to the engine braking torque. Engine speed, For engine braking control input (engine braking gear). This represents the engine braking torque mapping relationship.

[0111] Convert engine braking torque into wheel braking force:

[0112]

[0113] In the formula, The total transmission ratio of the transmission system. The total transmission ratio of the transmission system. This is the wheel's rolling radius.

[0114] (3) Hydraulic retarder model

[0115] The braking force of a hydraulic retarder is determined by its operating gear and speed, and its braking torque is expressed as:

[0116]

[0117] In the formula, The retarder braking torque (N) m), For the retarder speed, This is the retarder gear control value. This is the characteristic function of the retarder.

[0118] The corresponding wheel braking force is:

[0119]

[0120] (4) Pneumatic braking model

[0121] The braking force provided by an air brake is related to the pressure in the brake chamber, which can be expressed as:

[0122]

[0123] In the formula, For the first Axis No. Side-mounted air pressure compression power (N). The pressure in the brake chamber is (Pa). This is the braking gain coefficient.

[0124] For the Axis No. The total braking force of the side wheels is formed by the combined action of different braking sources, satisfying the following:

[0125]

[0126] In the formula, , These are the distribution coefficients for engine braking and hydraulic retarder braking among the wheels, respectively. For wheels that do not participate in the corresponding auxiliary braking, the distribution coefficient is 0.

[0127] (5) Multi-source braking optimization allocation

[0128] The rolling time-domain optimization is achieved using a Model Predictive Control (MPC) framework. The MPC control inputs are:

[0129]

[0130] In the formula, For engine braking force, For the braking force of the hydraulic retarder, To provide the power for each wheel's air compression, control the input vector. Indicates at time All braking force sources are adjustable.

[0131]

[0132] in, For the wheel speeds of each wheel, For the temperature of each axle brake, Engine speed, Hydraulic retarder gear position, Side rollover risk index, state vector This indicates the vehicle's current operating conditions and constraint information.

[0133] The multi-objective optimization function for MPC is:

[0134]

[0135] In the formula, To maximize the proportion of assisted braking, To minimize the highest brake temperature, Minimize the rate of change of retarder gears. The weights for each objective are dynamically adjusted based on the current operating conditions.

[0136] The MPC constraints are:

[0137] Overall braking force balance: Capacities of each braking source: Brake temperature: Retarder shift smoothness constraint: .

[0138] The MPC output is as follows:

[0139]

[0140] This invention uses an eight-axle heavy-duty engineering vehicle as the implementation object. Parameters are shown in Table 1. Engine braking and retarder characteristics are shown in Table 1. Figure 2 , Figure 3 .

[0141] Table 1 Vehicle Platform Parameter Table of Embodiments

[0142]

[0143] The innovative points and advantages of this invention are as follows:

[0144] (1) A method for self-diagnosis and virtual reconstruction of wheel speed sensor faults for heavy multi-axle vehicles is proposed. The wheel speed estimate is calculated by the vehicle longitudinal dynamics model and compared with the sensor measured value. A continuously changing reliability evaluation index is constructed based on the deviation. When the reliability index is lower than the threshold and lasts for a certain period of time, the sensor is judged to be faulty and the fault level is output. At the same time, based on the nonlinear mapping relationship reflecting the kinematic constraints between wheels, the wheel speed signal of the non-faulty wheel and the vehicle inertial measurement signal are used as input to reconstruct the wheel speed of the faulty wheel and output the reconstructed wheel speed and fault level, so as to realize the redundancy guarantee of wheel speed measurement of heavy multi-axle vehicles under complex working conditions.

[0145] (2) A method for coordinated lateral and longitudinal braking distribution for heavy multi-axle vehicles is proposed. By using a data fusion method based on signal quality weighting, the suspension pressure and boom attitude are fused to dynamically identify the real-time center of gravity position of the vehicle. The lateral load transfer rate is calculated based on the torque balance relationship to assess the rollover risk in real time. At the same time, the desired yaw moment is calculated based on the yaw rate deviation feedback. With minimizing the lateral load transfer rate, tracking the desired yaw moment, and balancing the longitudinal force load of each wheel tire as multiple objectives, and with the braking force of each wheel as the optimization variable, the target braking force of each wheel is solved by numerical optimization under the premise of meeting the total braking force requirement and the capability constraints of each wheel actuator, so as to achieve the coordinated distribution of braking force between the wheels.

[0146] (3) A multi-source braking optimization allocation method for heavy multi-axle vehicles is proposed. The total braking force demand of each wheel is received from the innovation point (2). Based on the brake temperature model and the engine braking torque characteristic function, a feedforward-predictive dynamic adjustment framework based on the thermal model is adopted to construct an optimization function with multiple objectives, namely maximizing the auxiliary braking usage ratio, minimizing the highest brake temperature, and smoothing the retarder gear shift. The weight of each objective is dynamically adjusted according to the current working condition. Under the conditions of satisfying the total braking force balance, the capacity constraints of each braking source, the upper limit constraint of brake temperature, and the smoothness constraint of retarder shift, the optimal control sequence is solved, and the engine braking force, hydraulic retarder gear and each axle air pressure power command are output to realize the coordinated optimization allocation of multi-source braking force.

[0147] Those skilled in the art will understand that all specific algorithms used in the above embodiments have alternatives, and these alternatives should be considered to fall within the protection scope of this invention. It is understood that the technical features of the above modules of this invention are not limited to the specific algorithms listed in the above specific embodiments. For example, the nonlinear mapping model reflecting the kinematic constraints between wheels can be any data-driven model capable of learning and fitting the mapping relationship between the normal wheel motion state and the faulty wheel motion state (e.g., various artificial neural networks, support vector regression, Gaussian process regression, etc.), or it can be a state observer or Kalman filter based on the vehicle kinematic equations and the geometric constraint relationship between wheels. Similarly, the numerical optimization solution can include sequential quadratic programming, interior point method, particle swarm optimization, etc.; the rolling time-domain optimization can include model predictive control, dynamic programming, etc. Simple substitutions to the above embodiments by those skilled in the art without creative effort should fall within the protection scope of this invention.

Claims

1. A braking control method for wheel speed reconfiguration and multi-source coordination for heavy-duty multi-axle vehicles, characterized in that, include: Wheel speed sensor fault self-diagnosis and virtual reconstruction module: acquires measured wheel speed signals characterizing the rotational speed of each wheel, calculates wheel speed estimates for each wheel based on the vehicle's longitudinal dynamics; generates a reliability evaluation index characterizing the reliability of the wheel speed signals based on the comparison between the measured wheel speed signals and the wheel speed estimates; when the reliability evaluation index meets preset fault judgment conditions, the corresponding wheel speed sensor is determined to be faulty, and fault level information is output; for the wheel determined to be faulty, the measured wheel speed signals of other wheels not determined to be faulty and the vehicle inertial measurement signal are used as inputs, and a reconstructed wheel speed signal for the faulty wheel is generated through a nonlinear mapping relationship reflecting the kinematic constraints between wheels, forming wheel speed measurement redundancy; Braking lateral and longitudinal coordinated distribution module: receives the reconstructed wheel speed signal and the fault level information; By integrating suspension pressure signals with signals characterizing boom posture, the real-time position of the vehicle's center of gravity is calculated. Based on the moment balance relationship and the real-time center of gravity position, the lateral load transfer rate is calculated to assess the risk of rollover. Based on the deviation of the actual yaw rate from the reference expected value, the expected yaw moment is calculated through a feedback correction mechanism. With the goals of minimizing the lateral load transfer rate, minimizing the tracking deviation of the desired yaw moment, and balancing the longitudinal force load of each wheel tire, the target longitudinal braking force of each wheel is solved by numerical optimization under the conditions of meeting the total braking force requirement and the upper limit constraint of the braking force of each wheel. Multi-source braking optimization allocation module: Receives the target longitudinal braking force of each wheel; based on the brake temperature dynamic model and engine braking torque characteristic function, performs feedforward estimation of brake temperature rise in the prediction time domain; constructs a cost function; wherein the first weight related to the proportion of auxiliary braking usage in the cost function is configured as an increasing function of continuous downhill duration, the second weight related to the highest temperature among all brakes is configured as an increasing function of brake temperature, and the third weight related to the degree of retarder gear change is configured as a decreasing function of rollover risk index; under the conditions of satisfying the total braking force balance constraint, the output capacity constraint of each braking source, the upper limit constraint of brake temperature, and the retarder shift smoothness constraint, solves the optimal control sequence, outputs engine braking control commands, hydraulic retarder gear commands, and axle air pressure power commands, realizing the coordinated allocation of multi-source braking forces.

2. The braking control method according to claim 1, characterized in that, Includes the following steps: Step S1: In the wheel speed sensor fault self-diagnosis and virtual reconstruction module, based on the vehicle's longitudinal dynamics, the vehicle speed estimate at the current moment is calculated using the vehicle's longitudinal acceleration and historical vehicle speed information. The wheel speed estimate is then calculated based on the vehicle speed estimate and the rolling radius of each wheel. The wheel speed estimate is compared with the sensor's measured value, and a continuously changing reliability evaluation index is generated based on the deviation. A reliability threshold and a duration threshold are set. When the reliability evaluation index is lower than the reliability threshold and the duration exceeds the duration threshold, the wheel speed sensor is determined to be faulty, and the fault level is output. For the wheel determined to be faulty, the measured wheel speed, vehicle longitudinal acceleration, yaw rate, and steering wheel angle of the other wheels not determined to be faulty are used as inputs. A reconstructed wheel speed signal is generated through a nonlinear mapping model reflecting the kinematic constraints between wheels. Based on the reliability evaluation index of each wheel, the normal wheel speed signal and the reconstructed wheel speed signal are weighted and synthesized to output the final wheel speed signal used by each wheel. Step S2: In the braking lateral and longitudinal coordinated distribution module, the final wheel speed signals and reliability evaluation indicators of each wheel output from step S1 are received; using the final wheel speed signals and reliability evaluation indicators, the longitudinal vehicle speed is calculated using a weighted average method, and the yaw rate and center of gravity sideslip angle are calculated through the speed difference between the left and right wheels; the suspension pressure signal and vehicle attitude angle signals including pitch and roll angles are acquired, and the center of gravity height and longitudinal position are dynamically calculated based on noise characteristic corrections associated with the reliability evaluation indicators of each wheel; the lateral load transfer rate is calculated based on the vertical load of each wheel, wherein the calculation of the vertical load of each wheel comprehensively considers the static load and longitudinal load. Load transfer caused by acceleration and lateral acceleration; the calculated actual yaw rate is compared with the expected yaw rate determined based on the driver's steering intention and vehicle speed, and the expected yaw moment is calculated through a feedback correction loop; using the braking force of each wheel as the optimization variable, a cost function is constructed to minimize the comprehensive lateral load transfer rate, minimize the yaw moment tracking deviation, and balance the longitudinal force load of each wheel tire. When the reliability evaluation index of a certain wheel is low, a penalty term for the change in the braking force distribution of the corresponding wheel is added; under the constraints that the total braking force meets the requirements, the braking force of each wheel does not exceed the upper limit, and the actuator change rate is limited, the target braking force of each wheel is solved through numerical optimization methods; Step S3: In the multi-source braking optimization allocation module, the target braking force requirements for each wheel and vehicle operating status information output in step S2 are received. The vehicle operating status information includes wheel speed, brake temperature, engine speed, current retarder gear position, and rollover risk index. Based on the brake temperature dynamic model, the temperature evolution of each brake in the prediction time domain is predicted. A cost function is constructed with the objectives of maximizing the auxiliary braking usage ratio, minimizing the highest temperature in each brake, and smoothing the retarder gear position change. The weight of each objective term is dynamically adjusted according to the current operating conditions. Under the conditions of satisfying the total braking force balance constraint, engine braking capacity constraint, hydraulic retarder braking capacity constraint, air pressure braking capacity constraint for each wheel, brake temperature safety upper limit constraint, and retarder gear position change rate constraint, the optimal control sequence is solved by rolling time domain optimization, and the engine braking control command, hydraulic retarder gear position command, and air pressure power command for each axle are output.

3. The braking control method according to claim 2, characterized in that, The weights related to brake temperature in the cost function are configured as a function of brake temperature, and the proportion of this weight in the total weight increases continuously and monotonically as the brake temperature rises from the normal range to the danger range; the weights related to the auxiliary braking ratio in the cost function are configured as a function of the duration of continuous downhill driving, and the proportion of this weight in the total weight increases continuously and monotonically as the duration of continuous downhill driving increases.

4. The braking control method according to claim 2 or 3, characterized in that, The specific implementation of wheel speed sensor fault self-diagnosis and virtual reconstruction in step S1 includes: Establish a longitudinal dynamics model for the vehicle, neglecting air resistance and rolling resistance. The longitudinal equation of motion is expressed as: In the formula, For the overall vehicle quality, For longitudinal vehicle speed, For the first Axis No. Side tire longitudinal force, For total braking force, This refers to the number of axles in the vehicle. Based on the above longitudinal dynamics model, using the vehicle's longitudinal acceleration Calculate the estimated longitudinal vehicle speed at the current moment using historical vehicle speed information. : In the formula, Sampling time; Based on the estimated vehicle speed and the rolling radius of each wheel Calculate the estimated wheel speeds for each wheel. : The residual calculation is the actual wheel speed measured by the sensor. With estimated wheel speed Difference: Map the residuals to continuous reliability weights: In the formula, For the reliability of wheel speed, To adjust the parameters; Set threshold With duration ,when And the duration exceeds The system can determine sensor malfunctions in real time and output the malfunction level. The system is defined as follows: 0 indicates normal operation, 1 indicates warning, and 2 indicates failure. When a wheel is in a normal state, the reliability evaluation index is close to 1, and the wheel speed is taken as the measured value. When the sensor fails, the reliability evaluation index is taken as 0, and the wheel speed is smoothly switched to the reconstructed value, thus achieving redundancy protection of the wheel speed measurement signal.

5. The braking control method according to claim 4, characterized in that: For the first one that is determined to be faulty Axis No. The wheel speed is reconstructed using a three-layer feedforward neural network. This neural network includes an input layer, a hidden layer, and an output layer, with each layer containing a specific number of neurons. , and ; The input and output of the input layer are as follows: In the formula, For input variables, ,in , and These are the vehicle's longitudinal acceleration, yaw rate, and steering angle, respectively; the rest... , The sensors measure wheel speeds of the non-faulty wheels on each axle of a heavy-duty multi-axle vehicle. Measure the number of faulty wheels for the sensor; The input and output of the hidden layer are represented as follows: In the formula, For input dimensions, For the hidden layer bias, The number of neurons in the hidden layer. The weight coefficients are the input variables for the hidden layer. The activation function for the hidden layer is an asymmetric piecewise linear function. In the formula, For output layer weights, For output layer bias, Let be the activation function of the output layer, and let its expression be a linear function. To reconstruct wheel speed output ; Wheel speed weighted fusion: In the formula, To output the final speed of each wheel, For reliability weights.

6. The braking control method according to claim 2, characterized in that, The method for implementing the lateral and longitudinal braking coordinated distribution in step S2 is as follows: (1) Vehicle motion state estimation based on wheel speed The final wheel speeds of each wheel obtained from step S1 With reliability weight Calculate the longitudinal speed of the vehicle yaw rate and centroid side slip angle : Among them, the longitudinal vehicle speed estimation adopts a weighted average method, with reliability weights. As a basis for confidence level: In the formula, For heavy multi-axle vehicles, the rolling radius of each wheel. For heavy-duty multi-axle vehicles; The yaw rate is calculated using the speed difference between the left and right wheels, and is also weighted using reliability factors. In the formula, For the first Wheelbase; Among them, the centroid side slip angle Lateral velocity calculated based on kinematic relationships By measuring lateral acceleration The lateral acceleration is obtained by integration. The yaw rate is measured by the IMU and calculated based on wheel speed. Drifting; (2) Dynamic identification of the centroid position Integrated suspension pressure With including pitch angle and roll angle attitude angle Based on noise characteristic corrections associated with reliability evaluation indicators for each round, the centroid height and longitudinal position are dynamically calculated; specifically, the weighted least squares method is used to estimate the real-time centroid height. With longitudinal position Perform dynamic solution. In the formula, The observation matrix consists of suspension forces and attitude angles. The weight matrix values ​​take into account the reliability weights of each round. Regarding suspension pressure measurement noise, when the reliability weight of a certain wheel is low, the confidence level of the corresponding suspension pressure data decreases accordingly. This is the torque average residual vector; (3) Calculation of lateral load transfer rate Based on the vertical load of each wheel Calculate the lateral load transfer rate (LTR): Among them, vertical load The calculations consider static load, load transfer caused by longitudinal acceleration, load transfer caused by lateral acceleration, and load transfer caused by... The instantaneous load fluctuation reflected by the calculated wheel speed difference: When the LTR exceeds the preset rollover warning threshold, a rollover warning is triggered. (4) Calculation of desired yaw moment The actual yaw rate obtained from step (1) The expected value determined based on the driver's steering intention and vehicle speed The comparison is performed, and the expected yaw moment is calculated through the feedback correction process. The desired yaw moment is calculated through a feedback correction loop, which is configured to generate a correction amount based on the deviation between the actual yaw rate and the desired yaw rate. (5) Multi-objective optimization allocation The optimization variable is the braking force of each wheel. The cost function comprehensively considers the following objectives: minimizing the lateral load transfer rate (LTR); minimizing the tracking deviation of the actual yaw moment from the desired yaw moment; and balancing the longitudinal force load on each tire. Specifically, it employs sequential quadratic programming. In the formula, This is the actual yaw moment. As a weighting factor, when the reliability weight of a certain wheel is low, the penalty for the braking force distribution of that wheel is increased to avoid over-reliance on the braking of that wheel when the signal is unreliable. As a weighting factor; Among the constraints is the balance of total braking force: Maximum braking force of each wheel: Actuator rate limit: The optimal solution of the cost function under the constraints is obtained by solving the numerical optimization method to obtain the target braking force for each round.

7. The braking control method according to claim 2, characterized in that, The implementation method of multi-source braking optimization allocation in step S3 is as follows: Receive the total braking force requirement of each wheel output in step S2 And vehicle operating status information, including wheel speeds of each wheel. Brake temperature Engine speed Retarder current gear Side rollover risk index ; (1) Brake temperature model The temperature dynamics of each brake are described using a lumped-parameter thermal model. Axis No. The temperature change rate of the side brake is: In the formula, Brake heat capacity, unit J / K This is the frictional heat distribution coefficient, which is the proportion of braking energy converted into heat. For the first Axis No. The braking force generated by the side brake, in N. For the first Axis No. Side wheel speed, in m / s The convective heat transfer coefficient is expressed in W / (m²). K), The brake cooling area is expressed in m². Ambient temperature, in Kelvin (K). Discretize the above equations and perform feedforward prediction of the temperature of each brake in the prediction time domain: (2) Engine braking characteristic model Based on the engine braking characteristic curve, an engine braking torque relationship model is constructed, and the engine braking torque is expressed as: In the formula, This refers to the engine braking torque. Engine speed, For engine braking control input, This relates to the engine braking torque mapping relationship. Convert engine braking torque into wheel braking force: In the formula, The total transmission ratio of the transmission system. The total transmission ratio of the transmission system. The radius of the wheel's rolling radius; (3) Hydraulic retarder model The braking force of a hydraulic retarder is determined by its operating gear and speed, and its braking torque is expressed as: In the formula, The retarder braking torque (N) m), For the retarder speed, This is the retarder gear control value. This is the characteristic function of the retarder; The corresponding wheel braking force is: (4) Pneumatic braking model The braking force provided by an air brake is related to the pressure in the brake chamber, as expressed as: In the formula, For the first Axis No. Side-mounted air pressure compression power (N). The pressure in the brake chamber is (Pa). This is the braking gain coefficient; For the Axis No. The total braking force of the side wheels is formed by the combined action of different braking sources, satisfying the following: In the formula, , These are the distribution coefficients for engine braking and hydraulic retarder braking among the wheels, respectively. For wheels that do not participate in the corresponding auxiliary braking, the distribution coefficient is 0. (5) Multi-source braking optimization allocation The Model Predictive Control (MPC) framework is used to achieve rolling time-domain optimization. The MPC control input is: In the formula, For engine braking force, For the braking force of the hydraulic retarder, To provide the power for each wheel's air compression, control the input vector. Indicates at time All braking force sources are adjustable. in, For the wheel speeds of each wheel, For the temperature of each axle brake, Engine speed, Hydraulic retarder gear position, Side rollover risk index, state vector This indicates the vehicle's current operating status and constraint information; The multi-objective optimization function for MPC is: In the formula, To maximize the proportion of assisted braking, To minimize the highest brake temperature, Minimize the rate of change of retarder gears. The weights for each objective are dynamically adjusted based on the current operating conditions. The MPC constraints are: Overall braking force balance: Capacities of each braking source: Brake temperature: Retarder shift smoothness constraint: ; The MPC output is as follows: 。

Images