Braking energy recovery predictive control method and system for multi-axle electric heavy trucks

Abstract

The present invention discloses a brake energy recovery predictive control method and system for a multi-axle electric heavy-duty truck, the method comprising the following steps: determining an initial predictive control algorithm and its initial cost function based on a vehicle braking dynamics model; optimizing the initial cost function based on an instantaneous response mechanism; predicting a real-time road adhesion coefficient based on a Kalman filter and optimizing the first-stage cost function based on a weight coefficient; performing soft-constraint optimization on the constraints of the initial predictive control algorithm based on a control barrier function; integrating a final predictive control model based on the second-stage cost function and smoothing constraints; solving and controlling the brake energy recovery predictive control problem of the multi-axle electric heavy-duty truck based on the final predictive control model; the present invention can adopt an improved model predictive control strategy for the braking system of the multi-axle electric heavy-duty truck, ultimately achieving efficient energy recovery, precise vehicle speed tracking, articulation angle stability, and a smooth deceleration process, thereby adapting to complex road conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle subsystem joint control. Specifically, the present invention is applied to the field of multi-axle electric heavy-duty trucks, and in particular to a brake energy recovery predictive control method and system for multi-axle electric heavy-duty trucks. Background Art

[0002] In recent years, with the rapid development of the new energy vehicle industry, research on electric vehicle brake energy recovery technology has achieved remarkable results. In the process of transitioning from traditional internal combustion engine vehicles to new energy vehicles, brake energy recovery, as a key technology to improve vehicle range and energy utilization, has become an important topic in the design of electric vehicle control systems. However, compared with ordinary passenger electric vehicles, electric heavy-duty trucks have higher requirements for brake energy recovery technology due to their complex structure, highly coupled power system, and large dynamic load variation range. Currently, the application of existing technologies still has many limitations and cannot meet the engineering requirements of multi-axle electric heavy-duty trucks under actual complex road conditions. The specific limitations are as follows:

[0003] Chinese patent CN114368369A discloses an integrated control method for a chassis braking system that adapts to the road friction coefficient. Specifically, the method achieves coordinated control of the vehicle's lateral stability, anti-lock braking, and anti-rollover protection by online adjustment of the objective function weights and desired state values. However, the method mainly focuses on the dynamic balance between the vehicle's longitudinal and lateral dynamic characteristics, lacks specificity in optimizing the efficiency of braking energy recovery, and fails to design a dynamic priority allocation strategy for the energy recovery system. In particular, in scenarios where the road adhesion coefficient suddenly changes, the control system is prone to target conflict, resulting in a coordinated imbalance between energy recovery efficiency and braking stability, making it difficult to cope with the complex and changing operating conditions of electric heavy-duty trucks.

[0004] China's published patent CN118810453A discloses a reinforcement learning-based brake energy recovery control method for multi-axis electric heavy-duty trucks. Specifically, it uses machine learning technology to achieve a dynamic balance between energy recovery efficiency and braking stability. Although this method improves energy recovery efficiency to a certain extent in a simulation environment, its online inference latency is relatively high, and it cannot meet the instantaneous response requirements of electric heavy-duty trucks, especially in dynamic scenarios. In addition, the reinforcement learning model requires a large amount of calibration data support, and there are problems of high computational complexity and hardware resource consumption in actual engineering applications. Moreover, the reliability of the algorithm stability under variable working conditions needs further verification.

[0005] China's published patent CN117445682A discloses a brake energy recovery control method, device, equipment and storage medium. The key to this method is that when the braking signal or vehicle parameters exceed a preset threshold, the system adjusts the motor torque to exit energy recovery to reduce the sense of frustration; however, this method relies on static calibration value trigger control, lacks the ability to actively adapt to dynamic working conditions, and has difficulty in achieving high-precision speed tracking and dynamic control; especially in sudden acceleration / deceleration scenarios, fixed threshold judgment may lead to torque step changes and lack a progressive and smooth constraint mechanism, which not only affects driving comfort, but may also cause power system instability due to violent fluctuations in motor torque, and urgently needs improvement.

[0006] In summary, the existing electric heavy truck brake energy recovery technology has the following problems that need to be solved:

[0007] (1) Insufficient collaborative optimization: Existing technologies mostly optimize a single performance indicator (such as stability or energy recovery efficiency) and fail to achieve dynamic synergy between the two. In particular, the efficiency is low in scenarios with multi-axis coupling and complex load changes.

[0008] (2) Limited real-time and robustness: Some algorithms rely on high computing resources and cannot meet the instantaneous response requirements of vehicle dynamic scenarios; while fixed threshold methods lack adaptive capabilities and are prone to control failure under complex road conditions.

[0009] (3) Rigid control strategy: Traditional rigid constraints can easily lead to severe fluctuations in motor torque, affecting vehicle smoothness and reliability.

[0010] Therefore, in view of the dynamic characteristics of multi-axle electric heavy-duty trucks, an intelligent control strategy that takes into account both braking energy recovery efficiency and braking stability is urgently needed. Summary of the Invention

[0011] The object of the present invention is to provide a method and system for predictive control of brake energy recovery of a multi-axle electric heavy truck, thereby solving all or one of the above-mentioned problems existing in the prior art.

[0012] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0013] In one aspect, the present invention provides a method for predictive control of brake energy recovery for a multi-axle electric heavy truck, comprising the following steps:

[0014] Model building steps:

[0015] Construct a vehicle braking dynamics model based on the characteristics of a multi-axle electric heavy-duty truck;

[0016] Steps to build a predictive control algorithm:

[0017] Determining an initial predictive control algorithm for the multi-axle electric heavy truck and an initial cost function of the initial predictive control algorithm based on the vehicle braking dynamics model;

[0018] Cost function optimization steps:

[0019] performing a first optimization operation on the initial cost function based on a transient response mechanism to obtain an optimized first-stage cost function;

[0020] Weight coefficient optimization steps:

[0021] predicting a real-time road adhesion coefficient based on a Kalman filter, and performing a second optimization operation based on a weight coefficient on the first-stage cost function based on the real-time road adhesion coefficient to obtain a second-stage cost function;

[0022] Constrained optimization steps:

[0023] performing soft constraint optimization on the constraint conditions of the initial predictive control algorithm based on the control barrier function to obtain smooth constraint conditions;

[0024] Predictive control steps:

[0025] integrating a final predictive control model based on the second-stage cost function and the smoothness constraint;

[0026] The braking energy recovery predictive control problem of the multi-axle electric heavy-duty truck is solved based on the final predictive control model, and the first item of the optimal control sequence obtained is used as the control instruction of the multi-axle electric heavy-duty truck for braking energy recovery control, and the solution process at the next moment is entered based on the final predictive control model.

[0027] Furthermore, the construction of a vehicle braking dynamics model based on the characteristics of a multi-axle electric heavy-duty truck includes:

[0028] Establish a global coordinate system;

[0029] Based on the global coordinate system, a first dynamic model of the tractor and a second dynamic model of the trailer are established by taking into account the tire forces, air resistance, and the force of the trailer acting on the tractor through the hinge point in the multi-axle electric heavy-duty truck, and the tire forces on the trailer in the multi-axle electric heavy-duty truck and the reaction force of the tractor acting on the trailer through the hinge point;

[0030] The first dynamic model and the second dynamic model are used as the vehicle braking dynamic model.

[0031] Furthermore, the determining of the initial predictive control algorithm of the multi-axle electric heavy truck based on the vehicle braking dynamics model includes:

[0032] Performing first-order approximate linearization processing on the vehicle braking dynamics model to obtain a time-varying linear equation;

[0033] The time-varying linear equation is subjected to Euler discretization processing to obtain a corresponding state space expression as the initial predictive control algorithm.

[0034] Furthermore, the initial cost function includes:

[0035] A first cost function is about energy recovery efficiency, a second cost function is about articulation angle error penalty, and a third cost function is about speed tracking condition.

[0036] Furthermore, the first optimization operation includes:

[0037] Introducing a transient response mechanism into the initial predictive control algorithm, and setting a corresponding first-order dynamic response for the control target corresponding to the initial cost function;

[0038] Based on the first-order dynamic response, the cost function in the first optimization stage is set to a weighted sum of the dynamic error and the control input to obtain the first-stage cost function.

[0039] Furthermore, the real-time road adhesion coefficient prediction based on Kalman filtering includes:

[0040] Estimate preliminary road adhesion coefficient based on tire slip ratio;

[0041] Define the state vector of the Kalman filter;

[0042] Determining a road adhesion coefficient observation equation based on a Kalman filter based on the state vector and the preliminary road adhesion coefficient estimation formula;

[0043] The real-time road adhesion coefficient is predicted based on the road adhesion coefficient observation equation.

[0044] Furthermore, the second optimization operation includes:

[0045] Setting an error weight adjustment strategy corresponding to a control target in the first-stage cost function based on the high and low changes in the real-time road adhesion coefficient;

[0046] Setting an adjustment coefficient for the error weight coefficient in the first-stage cost function based on the error weight adjustment strategy;

[0047] The first-stage cost function is updated based on the adjustment coefficient to obtain the second-stage cost function.

[0048] Furthermore, the soft constraint optimization includes:

[0049] The constraints of the initial predictive control algorithm are sorted into smoother constraints based on the general form of the inequality constraints of the control barrier function, thereby obtaining the smooth constraints.

[0050] Furthermore, the constraints of the initial predictive control algorithm include:

[0051] Constraints on road adhesion, battery power, and battery SOC.

[0052] On the other hand, the present invention also provides a brake energy recovery prediction control system for a multi-axle electric heavy truck, comprising:

[0053] Model building module, used to: build a vehicle braking dynamics model based on the characteristics of a multi-axle electric heavy-duty truck;

[0054] A predictive control algorithm building module is used to: determine an initial predictive control algorithm for the multi-axle electric heavy truck and an initial cost function of the initial predictive control algorithm based on the vehicle braking dynamics model;

[0055] A cost function optimization module is used to: perform a first optimization operation on the initial cost function based on a transient response mechanism to obtain an optimized first-stage cost function;

[0056] a weight coefficient optimization module, configured to: predict a real-time road adhesion coefficient based on a Kalman filter, and perform a second optimization operation based on the weight coefficient on the first-stage cost function based on the real-time road adhesion coefficient to obtain a second-stage cost function;

[0057] A constraint optimization module is used to perform soft constraint optimization on the constraint conditions of the initial predictive control algorithm based on a control barrier function to obtain smooth constraint conditions;

[0058] The predictive control module is used to: integrate a final predictive control model based on the second-stage cost function and the smoothness constraint condition; solve the brake energy recovery predictive control problem of the multi-axle electric heavy-duty truck based on the final predictive control model, use the first item of the solved optimal control sequence as the control instruction of the multi-axle electric heavy-duty truck to perform brake energy recovery control, and enter the solution process at the next moment based on the final predictive control model.

[0059] The beneficial effects of the technical solution of the present invention are:

[0060] 1. The brake energy recovery predictive control method for a multi-axle electric heavy-duty truck described in the present invention can be implemented for the braking system of a multi-axle electric heavy-duty truck. It adopts an improved model predictive control strategy and combines it with a transient response mechanism to improve computational efficiency and vehicle speed tracking accuracy. It dynamically adjusts and optimizes target weights based on the road adhesion coefficient to balance energy recovery and stability. It achieves progressive constraint satisfaction by controlling the barrier function to ensure smooth motor torque and avoid system oscillation. Ultimately, it achieves efficient energy recovery, precise vehicle speed tracking, stable articulation angle, and a smooth deceleration process to meet the needs of complex road conditions.

[0061] 2. The braking energy recovery prediction control system of the multi-axle electric heavy-duty truck described in the present invention can realize the braking energy recovery prediction control method of the multi-axle electric heavy-duty truck described in the present invention through the mutual cooperation of system modules. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0063] Figure 1 Schematic diagram of the XOY coordinate system in the predictive control method for braking energy recovery of a multi-axle electric heavy truck according to Example 1 of the present invention;

[0064] Figure 2 1 is a flow chart of a method for predictive control of braking energy recovery for a multi-axle electric heavy truck according to embodiment 1 of the present invention;

[0065] Figure 3 Schematic diagram of the architecture of the brake energy recovery prediction control system of the multi-axle electric heavy truck described in Example 2 of the present invention. DETAILED DESCRIPTION

[0066] The preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby making a clearer and more precise definition of the protection scope of the present invention.

[0067] In the description of the present invention, it should be noted that the embodiments described in the present invention are only part of the embodiments of the present invention, rather than all of the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative work are within the scope of protection of the present invention.

[0068] The terms "first," "second," and the like in the specification and claims herein and in the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate so that the embodiments of the present invention described herein can be implemented in orders other than those illustrated or described herein. In addition, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or device comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product, or device.

[0069] Example 1: This example provides a method for predicting and controlling the braking energy recovery of a multi-axle electric heavy truck. Figure 1 and Figure 2 As shown, the following steps are included:

[0070] S100, model construction steps:

[0071] In this step, a braking dynamics model considering the articulation angle between the multi-axle heavy truck tractor and the trailer is established as follows:

[0072] S101, establish the XOY global coordinate system, such as Figure 1 As shown, based on Figure 1 The coordinate system describes the position of the vehicle in absolute space and defines the vehicle coordinate system of the tractor and trailer respectively. and ,in:

[0073] 1) is the tractor vehicle coordinate system, the origin is the center of mass of the tractor, is the trailer vehicle coordinate system, the origin is the center of mass of the trailer; The articulation point between the tractor and the trailer ( and In actual vehicles, they are overlapping. Figure 1 In order to clearly show the forces, the two points are shown separately in the front and rear). The hinge point is located between the tractor's center axle and the rear axle, and is located on the axis of the tractor and trailer;

[0074] 2) For the The longitudinal force on the wheels of the group, For the The lateral force on the wheels of the group, , , , is the interaction force between the tractor and trailer at the articulation point, is the articulation angle between the tractor and trailer, is the front wheel turning angle; is the longitudinal speed of the tractor, is the lateral speed of the tractor, is the yaw angle of the tractor, is the longitudinal speed of the trailer, is the trailer's lateral velocity, is the trailer yaw angle, is the distance between the center of each axle and the center of mass of the vehicle body, is the distance from the center of mass of the vehicle to the hinge point, The axle track of the tractor.

[0075] S102. Based on the coordinate system XOY, considering the tire forces on the tractor, air resistance, and the force of the trailer acting on the tractor through the hinge point, and considering the tire forces on the trailer and the reaction force of the tractor acting on the trailer through the hinge point, the dynamic models of the tractor and the trailer are established as follows:

[0076] 1) The dynamic model of the tractor is expressed as:

[0077] ;

[0078] in, is the total mass of the tractor, is the total moment of inertia of the tractor; since the longitudinal braking force of the tractor is equal to the sum of the motor braking force and the air pressure braking force on the tractor, the following relationship can be obtained:

[0079] ;

[0080] in, is the transmission ratio of each motor on the tractor, is the torque of each motor on the tractor, is the air pressure of the pneumatic brake, is the wheel radius of the tractor, Air pressure Conversion factor into braking force;

[0081] 2) The dynamic model of the trailer is expressed as:

[0082] ;

[0083] in, is the total mass of the trailer, is the total moment of inertia of the trailer;

[0084] 3) Since the tractor and trailer are connected by a saddle at a hinge point, the constraints at the hinge point of the tractor and trailer need to be considered when establishing the dynamic model of the multi-axle vehicle. The constraints on the longitudinal, lateral, and yaw motions of the tractor and trailer at the hinge point are expressed as follows:

[0085] .

[0086] It should be noted that in order to achieve coordinated control of the vehicle and efficient recovery of braking energy, this method is based on the above model and constraints on the motor torque. and brake pressure Carry out precise control to ensure the smooth reduction of vehicle speed, optimize energy recovery efficiency, maintain vehicle driving stability and prevent tail-swinging; optimize motor torque and brake pressure Key variables such as the braking force can be reasonably distributed, which is not only conducive to achieving stable and controllable speed reduction, but also effectively improving the recovery rate of braking energy. In addition, combined with the above-mentioned constraint relationship between the articulated point between the tractor and the trailer, it ensures that all parts of the vehicle move in a coordinated manner during braking to maintain vehicle stability.

[0087] S200, predictive control algorithm construction steps:

[0088] In this step, based on the dynamic model constructed in the above steps, the corresponding predictive control algorithm is constructed, which includes the state iteration equation and cost function, as follows:

[0089] S201, based on the dynamic model of the tractor and the dynamic model of the trailer, setting a prediction model for energy recovery and defining an initial cost function;

[0090] 1) Select the method used to represent the vehicle's motion state As a state variable, As control input;

[0091] 2) Perform first-order approximate linearization on the dynamic models of the tractor and trailer, and obtain the time-varying linear equations as follows:

[0092] ;

[0093] ;

[0094] ;

[0095] 3) Perform Euler discretization on the above time-varying linear equation to obtain the corresponding state space expression as the prediction model, as follows:

[0096] ;

[0097] ;

[0098] in, is the air resistance coefficient, is the longitudinal coupling coefficient between the tractor and trailer, is the tire cornering stiffness, is the yaw damping coefficient, is the yaw coupling coefficient of the tractor and trailer.

[0099] S202. According to the characteristics of the electric multi-axle heavy truck, a cost function of the predictive control algorithm is set. The cost function includes three control objectives, as follows:

[0100] 1) The first control goal: to ensure high braking energy recovery efficiency and to increase energy recovery efficiency It is defined as the ratio of the energy recovered by the regenerative braking system to the total braking energy, that is:

[0101] ;

[0102] in, For the Shaft motor braking torque, For the The rotational speed of each wheel; For tractors The speed of time, For trailers the speed of the moment;

[0103] based on , the cost function corresponding to the first control objective is as follows:

[0104] ;

[0105] in To theoretically maximize energy recovery efficiency, is the weight coefficient of energy recovery efficiency;

[0106] 2) Second control objective: articulation angle error penalty; this control objective is to prevent the vehicle from folding or losing stability due to excessive articulation angles. The corresponding cost function is as follows:

[0107] ;

[0108] in, for The size of the hinge angle at any moment; is the reference articulation angle, =0, which means the truck's desired state is to drive in a straight line; is the weight coefficient regarding the stability of the hinge angle;

[0109] 3) The third control objective: Ensure that the actual speed of the vehicle reaches the desired speed. The corresponding cost function is as follows:

[0110] ;

[0111] in, is the actual vehicle speed of the towing vehicle; is the actual speed of the trailer vehicle; is the desired speed of the tractor; is the expected speed of the trailer; is the weight coefficient for speed tracking.

[0112] S203, based on the prediction model of S201 and the cost function of S202, the predictive control (MPC) algorithm of the multi-axle electric heavy truck braking model is integrated as follows:

[0113] ;

[0114] Among them, T max is the maximum torque of the motor, P max is the maximum brake pressure;

[0115] It should be noted that the real-time response speed and computational efficiency of the above-mentioned MPC algorithm still have room for improvement. Since the weight coefficients in the cost function are fixed, the control instructions calculated according to the fixed cost function under dynamic and complex road conditions are not the optimal solution for the current state. Moreover, the constraints of the aforementioned model only consider the motor torque and air brake, and use the maximum value as the constraint. Sudden changes in the control instructions may occur, and a smooth and comfortable driving state cannot be guaranteed. In addition, the battery state is not constrained in the above-mentioned constraints, so battery overcharging may occur. This method is precisely aimed at the above problems. In the subsequent steps, the real-time response speed and computational efficiency of the MPC algorithm are optimized through the transient response mechanism. By dynamically adjusting the weight coefficients in the cost function according to the road adhesion coefficient, the vehicle can balance driving stability and energy recovery efficiency under dynamic road conditions. The constraints are optimized by the barrier control function (CBF) so that the constraints are asymptotically satisfied at a finite speed, ultimately ensuring the safety and comfort of vehicle driving.

[0116] S300, cost function optimization step:

[0117] In this step, the cost function of the MPC algorithm is optimized based on the transient response mechanism, as follows:

[0118] S301. To improve the real-time response speed and computational efficiency of the MPC algorithm, a transient response mechanism is introduced into the MPC algorithm. First-order dynamic responses are set for each of the three control objectives in the braking energy recovery (energy recovery efficiency deviation, articulation angle deviation, and speed tracking), as follows:

[0119] ;

[0120] in , , are the error attenuation coefficients under the corresponding control targets, which can directly determine the speed of dynamic response.

[0121] S302: Based on the expected dynamic response, the stage cost function is set to the weighted sum of the dynamic error and the control input, that is, the cost function is optimized for the first time. The first stage cost function obtained is as follows:

[0122] ;

[0123] in, , , are the weight coefficients corresponding to the control objectives, is the regularization coefficient used to balance the dynamic error and actuator load.

[0124] S400, weight coefficient optimization steps:

[0125] In this step, the road adhesion coefficient is estimated in real time based on the Kalman filter, and the weight coefficient in the cost function is dynamically optimized according to the estimated road adhesion coefficient, as follows:

[0126] S401. Considering the need to maintain driving stability during vehicle braking, in order to achieve optimal energy recovery efficiency while ensuring driving stability, it is necessary to incorporate a road adhesion coefficient constraint into the optimization process:

[0127] 1) Make a preliminary estimate of the road adhesion coefficient based on the tire slip rate as follows:

[0128] ;

[0129] in, is the tire slip rate, is the tire longitudinal stiffness, is the road adhesion coefficient, F zi is the vertical load of the tire;

[0130] 2) Define the state vector of the Kalman filter as: , assuming that the adhesion coefficient changes smoothly in a short time, the state and observation equations of the Kalman filter are:

[0131] ;

[0132] in, is the process noise, is the measurement noise, which obeys the Gaussian distribution with mean 0 and variance Q, R;

[0133] 3) Based on the above formula, the road adhesion coefficient is estimated using Kalman filtering. The specific prediction and update steps are as follows:

[0134] ;

[0135] ;

[0136] in, is the calculated real-time estimated road adhesion coefficient.

[0137] S402, based on Dynamically adjust the weight coefficients in MPC (the following is the weight adjustment strategy):

[0138] 1) In response to the vehicle traveling on a road with low adhesion coefficient, the error weights of vehicle speed and articulation angle in the objective function (i.e., the first-stage cost function) are increased;

[0139] 2) In response to the vehicle traveling on a road surface with a high adhesion coefficient, increasing the error weight of the energy recovery efficiency in the objective function;

[0140] 3) The weight coefficient in the objective function is expressed as:

[0141] ;

[0142] in, , , are the weight adjustment coefficients corresponding to the three control objectives, which are used to control the change range of the weights; , are the preset minimum and maximum values ​​of the road adhesion coefficient respectively.

[0143] S403. Based on the optimization in S402, the first-stage cost function is optimized for the second time. The final second-stage cost function is as follows:

[0144] .

[0145] S500, constraint optimization steps:

[0146] In this step, the constraints in the MPC algorithm are optimized based on the control barrier function (CBF). The constraints are set for road adhesion, which affects vehicle driving stability, battery power, which affects energy recovery, and battery SOC, which is added as a constraint to prevent battery overcharging. The details are as follows:

[0147] S501. Determine the change formula of battery SOC:

[0148] ;

[0149] in, is the initial SOC of the battery, is the total energy that the battery can store, is the braking process start time, is the end time of the braking process, is the number of motors involved in braking energy recovery, is the vehicle speed, For the The transmission ratio of each motor, is the tire radius.

[0150] S502. Determine the general form of the inequality constraint of CBF:

[0151] .

[0152] S503: For the three constraints (battery power constraint, road adhesion constraint, and battery SOC constraint), define , then:

[0153] ;

[0154] Among them, h1 is the constraint condition on battery power, h2 is the constraint condition on road adhesion, h3 is the constraint condition on battery SOC, SOC max is the maximum SOC of the battery.

[0155] S504. Based on the above formula, substituting into CBF, we have:

[0156] .

[0157] S505. Arrange the above formula into smoother constraints as follows:

[0158] .

[0159] S600, predictive control steps:

[0160] In this step, based on all the above steps, the final cost function, constraints, and initial MPC model are integrated to obtain an optimized MPC algorithm model. The online braking system control instructions are solved based on the optimized MPC model, and online periodic control is performed based on the obtained control instructions, as follows:

[0161] S601, the optimized MPC algorithm is as follows:

[0162] ;

[0163] in:

[0164] is the attenuation coefficient of battery power change, which can determine the relaxation speed of battery power constraint. When the power is too low, the control power will slowly approach the upper limit to prevent the battery from overloading due to sudden power changes.

[0165] is the attenuation coefficient of the road adhesion change, which is reduced on low adhesion roads. It can extend the restraint buffer time and improve the To improve response speed;

[0166] The attenuation coefficient of the battery SOC change is used to control the dynamic boundary of the battery state of charge. This may cause the SOC to quickly approach the threshold, limiting the energy recovery potential; This will cause overcharging risk due to hysteresis response;

[0167] It should be noted that in actual working conditions, the magnitude of each attenuation coefficient is adaptively adjusted according to the test conditions.

[0168] S602, when actually controlling, solve the predictive control problem based on the optimized MPC model, and use the obtained optimal control sequence The first item is used as the control instruction of the braking system at the current moment to perform brake energy recovery control of the multi-axis electric heavy truck, and S602 is performed periodically.

[0169] In summary, this method can improve the real-time response speed and computational efficiency of the vehicle based on the optimized MPC control algorithm, take into account the vehicle's driving stability and brake energy recovery efficiency under dynamic and complex road conditions, and improve the smoothness and safety of multi-axle electric heavy-duty trucks during driving.

[0170] It should be noted that the above examples are only for explaining the present invention and are not intended to limit the scope of protection of the present invention.

[0171] Example 2: This example is based on the same inventive concept as the method for predictive control of braking energy recovery of a multi-axle electric heavy truck described in Example 1, and provides a predictive control system for braking energy recovery of a multi-axle electric heavy truck, such as Figure 3 Shown, including:

[0172] Model building module, used to: build a vehicle braking dynamics model based on the characteristics of a multi-axle electric heavy-duty truck;

[0173] A predictive control algorithm building module is used to: determine an initial predictive control algorithm for the multi-axle electric heavy truck and an initial cost function of the initial predictive control algorithm based on the vehicle braking dynamics model;

[0174] A cost function optimization module is used to: perform a first optimization operation on the initial cost function based on a transient response mechanism to obtain an optimized first-stage cost function;

[0175] a weight coefficient optimization module, configured to: predict a real-time road adhesion coefficient based on a Kalman filter, and perform a second optimization operation based on the weight coefficient on the first-stage cost function based on the real-time road adhesion coefficient to obtain a second-stage cost function;

[0176] A constraint optimization module is used to perform soft constraint optimization on the constraint conditions of the initial predictive control algorithm based on a control barrier function to obtain smooth constraint conditions;

[0177] The predictive control module is used to: integrate a final predictive control model based on the second-stage cost function and the smoothness constraint condition; solve the brake energy recovery predictive control problem of the multi-axle electric heavy-duty truck based on the final predictive control model, use the first item of the solved optimal control sequence as the control instruction of the multi-axle electric heavy-duty truck to perform brake energy recovery control, and enter the solution process at the next moment based on the final predictive control model.

[0178] Different from the existing technology, the present application adopts a brake energy recovery prediction control method and system for a multi-axle electric heavy-duty truck. It can adopt an improved model predictive control strategy for the braking system of the multi-axle electric heavy-duty truck, combined with the transient response mechanism to improve computing efficiency and vehicle speed tracking accuracy; dynamically adjust the optimization target weight based on the road adhesion coefficient to balance energy recovery and stability; achieve progressive constraint satisfaction by controlling the barrier function to ensure smooth motor torque and avoid system oscillation; and ultimately achieve efficient energy recovery, precise vehicle speed tracking, stable articulation angle and smooth deceleration process to adapt to complex road conditions.

[0179] It should be understood that in the various embodiments of this document, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this document.

[0180] It should also be understood that in the embodiments herein, the term "and / or" merely describes an association between associated objects, indicating that three possible relationships exist. For example, "A and / or B" could represent: A alone, A and B simultaneously, or B alone. Furthermore, the character " / " in this document generally indicates an "or" relationship between the associated objects.

[0181] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the composition and steps of each example according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this document.

[0182] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0183] In the several embodiments provided herein, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices, or units, or can be an electrical, mechanical, or other form of connection.

[0184] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments herein.

[0185] In addition, the functional units in the various embodiments herein may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0186] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this article is essentially or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of this article. The aforementioned storage medium includes: various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0187] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention's description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A predictive control method for braking energy recovery of a multi-axle electric heavy truck, characterized in that: The following steps are involved: Model building steps: Construct a vehicle braking dynamics model based on the characteristics of a multi-axle electric heavy-duty truck; Steps to build a predictive control algorithm: Determining an initial predictive control algorithm for the multi-axle electric heavy truck and an initial cost function of the initial predictive control algorithm based on the vehicle braking dynamics model; Cost function optimization steps: performing a first optimization operation on the initial cost function based on a transient response mechanism to obtain an optimized first-stage cost function; Weight coefficient optimization steps: predicting a real-time road adhesion coefficient based on a Kalman filter, and performing a second optimization operation based on a weight coefficient on the first-stage cost function based on the real-time road adhesion coefficient to obtain a second-stage cost function; Constrained optimization steps: performing soft constraint optimization on the constraint conditions of the initial predictive control algorithm based on the control barrier function to obtain smooth constraint conditions; Predictive control steps: integrating a final predictive control model based on the second-stage cost function and the smoothness constraint; The braking energy recovery predictive control problem of the multi-axle electric heavy-duty truck is solved based on the final predictive control model, and the first item of the optimal control sequence obtained is used as the control instruction of the multi-axle electric heavy-duty truck for braking energy recovery control, and the solution process at the next moment is entered based on the final predictive control model.

2. The method for predictive control of brake energy recovery for a multi-axle electric heavy truck according to claim 1, characterized in that: The vehicle braking dynamics model based on the characteristics of a multi-axle electric heavy-duty truck is constructed, including: Establish a global coordinate system; Based on the global coordinate system, a first dynamic model of the tractor and a second dynamic model of the trailer are established by taking into account the tire forces, air resistance, and the force of the trailer acting on the tractor through the hinge point in the multi-axle electric heavy-duty truck, and the tire forces on the trailer in the multi-axle electric heavy-duty truck and the reaction force of the tractor acting on the trailer through the hinge point; The first dynamic model and the second dynamic model are used as the vehicle braking dynamic model.

3. The predictive control method for braking energy recovery of a multi-axle electric heavy truck according to claim 1 is characterized in that: The initial predictive control algorithm for the multi-axle electric heavy truck is determined based on the vehicle braking dynamics model, including: Performing first-order approximate linearization processing on the vehicle braking dynamics model to obtain a time-varying linear equation; The time-varying linear equation is subjected to Euler discretization processing to obtain a corresponding state space expression as the initial predictive control algorithm.

4. The method for predictive control of brake energy recovery for a multi-axle electric heavy truck according to claim 1, characterized in that: The initial cost function includes: A first cost function is about energy recovery efficiency, a second cost function is about articulation angle error penalty, and a third cost function is about speed tracking condition.

5. The method for predictive control of braking energy recovery for a multi-axle electric heavy truck according to claim 1, characterized in that: The first optimization operation includes: Introducing a transient response mechanism into the initial predictive control algorithm, and setting a corresponding first-order dynamic response for the control target corresponding to the initial cost function; Based on the first-order dynamic response, the cost function in the first optimization stage is set to a weighted sum of the dynamic error and the control input to obtain the first-stage cost function.

6. The method for predictive control of brake energy recovery for a multi-axle electric heavy truck according to claim 1, characterized in that: The method of predicting the real-time road adhesion coefficient based on Kalman filtering includes: Estimate preliminary road adhesion coefficient based on tire slip ratio; Define the state vector of the Kalman filter; Determining a road adhesion coefficient observation equation based on a Kalman filter based on the state vector and the preliminary road adhesion coefficient estimation formula; The real-time road adhesion coefficient is predicted based on the road adhesion coefficient observation equation.

7. The method for predictive control of brake energy recovery for a multi-axle electric heavy truck according to claim 1, characterized in that: The second optimization operation includes: Setting an error weight adjustment strategy corresponding to a control target in the first-stage cost function based on the high and low changes in the real-time road adhesion coefficient; Setting an adjustment coefficient for the error weight coefficient in the first-stage cost function based on the error weight adjustment strategy; The first-stage cost function is updated based on the adjustment coefficient to obtain the second-stage cost function.

8. The method for predictive control of brake energy recovery for a multi-axle electric heavy truck according to claim 1, characterized in that: The soft constraint optimization includes: The constraints of the initial predictive control algorithm are sorted into smoother constraints based on the general form of the inequality constraints of the control barrier function, thereby obtaining the smooth constraints.

9. The method for predictive control of brake energy recovery for a multi-axle electric heavy truck according to claim 8, characterized in that: The constraints of the initial predictive control algorithm include: Constraints on road adhesion, battery power, and battery SOC.

10. A brake energy recovery prediction control system for a multi-axle electric heavy truck, characterized in that: include: Model building module, used to: build a vehicle braking dynamics model based on the characteristics of a multi-axle electric heavy-duty truck; A predictive control algorithm building module is used to: determine an initial predictive control algorithm for the multi-axle electric heavy truck and an initial cost function of the initial predictive control algorithm based on the vehicle braking dynamics model; A cost function optimization module is used to: perform a first optimization operation on the initial cost function based on a transient response mechanism to obtain an optimized first-stage cost function; a weight coefficient optimization module, configured to: predict a real-time road adhesion coefficient based on a Kalman filter, and perform a second optimization operation based on the weight coefficient on the first-stage cost function based on the real-time road adhesion coefficient to obtain a second-stage cost function; A constraint optimization module is used to perform soft constraint optimization on the constraint conditions of the initial predictive control algorithm based on a control barrier function to obtain smooth constraint conditions; The predictive control module is used to: integrate a final predictive control model based on the second-stage cost function and the smoothness constraint condition; solve the brake energy recovery predictive control problem of the multi-axle electric heavy-duty truck based on the final predictive control model, use the first item of the solved optimal control sequence as the control instruction of the multi-axle electric heavy-duty truck to perform brake energy recovery control, and enter the solution process at the next moment based on the final predictive control model.

Images