Multi-axis distributed driving vehicle trajectory tracking and stability cooperative control method

Abstract

The invention discloses a trajectory tracking and stability cooperative control method for a multi-axis distributed driving vehicle, and provides a hierarchical control architecture. The upper-layer trajectory tracking controller comprises a self-adaptive preview linear quadratic regulator based on fuzzy logic reasoning, and the self-adaptive preview linear quadratic regulator dynamically adjusts the preview distance according to the vehicle speed and the acceleration and outputs a steering angle instruction; and the seven-section S curve speed planner generates a smooth speed and acceleration curve. The lower-layer stability and torque distribution controller comprises an equivalent sliding mode controller for calculating an additional yaw moment based on the yaw velocity and the side slip angle error; and based on a torque distributor of quadratic programming, introducing an inter-axle yawing moment coefficient as a geometric constraint, tracking a required torque, and minimizing a tire attachment utilization rate and a slip rate as a target distribution torque. Through cooperation of the upper layer and the lower layer, high-precision trajectory tracking, high robust stability and longitudinal smoothness are considered, and the comprehensive performance of the multi-axle vehicle under the limiting working condition is remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of vehicle dynamics control technology, specifically a method for coordinated control of trajectory tracking and stability of multi-axle distributed drive vehicles, which is particularly suitable for heavy-duty vehicles with 8x8 independent drive and independent steering. Background Technology

[0002] With the widespread application of multi-axle distributed drive vehicles in military transportation, special engineering, and high-level autonomous driving, chassis integrated control faces the common technical challenge of coordinating trajectory tracking accuracy and yaw stability. These vehicles typically possess dynamic characteristics of multiple axles, multiple drive units, high loads, and complex center-of-gravity distributions. Therefore, under extreme driving conditions such as high-speed cruising, large-angle steering, or emergency obstacle avoidance on low-friction surfaces, the intensified coupling effect between longitudinal and lateral dynamics can easily lead to vehicle instability and increased tracking deviations, even resulting in loss of control risks. Traditional decoupled control architectures cannot coordinate the control of trajectory tracking accuracy and stability, making it difficult to effectively maintain vehicle yaw stability while ensuring high-precision trajectory tracking. Therefore, developing a cooperative control system that deeply integrates trajectory tracking and yaw stability control objectives and is applicable to complex multi-axle distributed drive architectures is crucial for improving the overall driving performance and handling safety of heavy vehicles.

[0003] Currently, research on coordinated control of vehicle trajectory tracking and stability largely focuses on four-wheel drive (4WID) passenger vehicles or light-duty platforms. The mainstream control algorithms employed include linear quadratic regulators (LQR), model predictive control (MPC), sliding mode control (SMC), and various improved strategies. For example, existing patents (such as CN115871641A) propose a coordinated control method that dynamically adjusts the weights of the trajectory tracking control model and the stability control model using an "adaptive coordination coefficient," aiming to balance tracking accuracy and stability under extreme conditions. However, such solutions still have significant limitations when dealing with multi-axle distributed drive vehicles such as 8×8. Their control models are based on simplified vehicle dynamics and do not adequately consider torque vector distribution strategies for lower-level actuators, failing to fully realize the control potential. Furthermore, while algorithms such as MPC offer superior performance, their computational complexity increases exponentially with system dimensions, making it difficult to meet the real-time requirements of controllers in multi-axle vehicles.

[0004] In summary, existing technologies suffer from numerous bottlenecks in multi-axis application scenarios, such as poor model adaptability, coarse execution layer allocation strategies, and real-time performance issues. There is an urgent need for an innovative, integrated solution that covers everything from top-level architecture design and mid-level adaptive decision-making to efficient execution at the bottom layer. Summary of the Invention

[0005] This invention proposes a multi-axis distributed drive vehicle trajectory tracking and stability cooperative control method. This method employs a hierarchical control architecture, including:

[0006] 1. Upper-level trajectory tracking controller, used to generate target motion commands for controlling the vehicle, including:

[0007] Lateral control module: dynamically adjusts the aiming distance based on the vehicle's real-time longitudinal speed and longitudinal acceleration, and calculates and outputs steering angle commands based on the aiming error;

[0008] Longitudinal control module: Constrains and smooths the longitudinal acceleration of the vehicle and its rate of change, and generates target velocity curve and target acceleration curve;

[0009] 2. Lower-level stability and torque distribution controller, used to execute upper-level commands and maintain vehicle yaw stability, including:

[0010] Yaw moment decision module: Calculates the additional yaw moment required for the vehicle body to maintain stability based on the vehicle's yaw rate error and center of gravity sideslip angle error;

[0011] Torque vector distribution module: Based on the driver's total required torque and the additional yaw moment, the drive torque is distributed to each wheel hub motor through multi-objective optimization calculation.

[0012] Compared with the prior art, the advantages of the present invention are:

[0013] 1. The multi-axis distributed drive vehicle trajectory tracking and stability collaborative control method described in this invention generates a desired state that balances tracking accuracy and motion smoothness through an adaptive preview linear quadratic regulator and a seven-segment S-curve planning at the upper layer; and ensures that the vehicle accurately tracks the upper layer commands while actively enhancing anti-interference capabilities and optimizing tire force utilization through equivalent sliding mode control and multi-objective optimization allocation at the lower layer.

[0014] 2. The multi-axle distributed drive vehicle trajectory tracking and stability cooperative control method described in this invention constructs the torque distribution problem as a multi-objective constrained optimization problem that includes minimizing tire adhesion utilization, minimizing tire slip ratio, and accurately tracking total driving force and additional yaw moment. By introducing an inter-axle yaw moment coefficient, the optimal inter-axle torque distribution ratio based on vehicle steering geometry is derived, significantly improving the system's dynamic response speed and stability limit on low-adhesion road surfaces. Attached Figure Description

[0015] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0016] Figure 1 This is a flowchart of a multi-axis distributed drive vehicle trajectory tracking and stability cooperative control method according to the present invention;

[0017] Figure 2 This is a schematic diagram of the multi-axis single-track two-degree-of-freedom model described in this invention;

[0018] Figure 3 This is a structural diagram of the adaptive fuzzy controller described in this invention;

[0019] Figure 4 This is a schematic diagram of the seven-segment S-curve acceleration planning described in this invention; Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0021] The invention will now be further described with reference to the accompanying drawings.

[0022] like Figure 1 As shown, a multi-axis distributed drive vehicle trajectory tracking and stability cooperative control method is proposed. This method adopts a hierarchical control architecture, including:

[0023] The upper-level trajectory tracking controller generates target motion commands for controlling the vehicle, including: a lateral control module that calculates the front wheel steering angle command based on the vehicle state and the desired trajectory using an adaptive preview linear quadratic adjuster. The longitudinal control module generates a smooth target velocity based on the desired velocity profile using a seven-segment S-curve planner. With acceleration curve;

[0024] The lower-level stability and torque distribution controller, used to execute upper-level commands and maintain vehicle yaw stability, includes: a yaw moment decision module based on the vehicle state and the desired yaw rate generated by the upper-level commands. The additional yaw moment required to maintain stability is calculated using equivalent sliding mode control. The torque vectoring module comprehensively considers the driver's total torque demand. Additional yaw moment Considering road surface adhesion conditions and motor external characteristics, the optimal driving torque is determined by solving a multi-objective quadratic programming problem. Distributed to 8 hub motors;

[0025] 1. Specific implementation of the upper-level trajectory tracking controller:

[0026] 1.1 Lateral Control Module: An adaptive pre-aiming linear quadratic regulator based on fuzzy logic inference is adopted. The fuzzy logic inference outputs dynamic pre-aiming time based on vehicle speed and acceleration. The specific steps are as follows:

[0027] Step A1. Construct a vehicle tracking error model, using... Figure 2 The multi-axis, single-track, two-DOF model shown is used as a reference model. The lateral tracking error is defined in the Frenet coordinate system. With heading error The following error state-space equation is established:

[0028] (1)

[0029] In the formula, the state vector Control input For the steering angle of each axis, This is a coefficient matrix related to vehicle parameters and longitudinal velocity;

[0030] Step A2. Construct an adaptive fuzzy logic reasoning module, such as... Figure 3 As shown, the input to this module is defined as the vehicle's longitudinal speed. and longitudinal acceleration The output is defined as the dynamic aiming time. The input and output are fuzzified using a triangular membership function, and the fuzzy subset partitioning is shown in Table 1. Based on the rule that "the higher the vehicle speed or the greater the acceleration, the longer the anticipation time is required to improve predictability", the fuzzy rules are shown in Table 2.

[0031] Table 1 Fuzzy Subset Partitioning

[0032] Fuzzy subsets NB NS ZO PS PB Speed ​​(km / h) 16 32 48 64 80 Acceleration (m / s²) -2 -1 0 1 2 Pre-aiming time (s) 0.05 0.1 0.15 0.2 0.25

[0033] Table 2 Fuzzy Rules

[0034] Pre-aiming time speed acceleration NB NS ZO PS PB NB NB NS ZO PS PB NS NB NS ZO PS PB ZO NB NS ZO PS PB PS NB NS ZO PS PB PB NS ZO PS PB PB

[0035] The centroid method is used to deblur the fuzzy output, and the accurate dynamic pre-aiming time is calculated. ;

[0036] Step A3. Based on dynamic aiming time and current vehicle status By transforming the Frenet coordinate system to the Cartesian coordinate system, the target point for the future trajectory is determined. The calculation formula is as follows:

[0037] (2)

[0038] In the formula, For vehicle location, For the longitudinal speed of the vehicle, For the vehicle's lateral speed, This refers to the vehicle's yaw angle;

[0039] Step A4. Calculate the aiming point With trajectory reference point Lateral error between aiming and heading error Construct the pre-aiming state vector The calculation formula is as follows:

[0040] (3)

[0041] In the formula, The heading angle of the trajectory reference point. This is the unit normal vector at the trajectory reference point;

[0042] Step A5. Pre-aiming state vector Substituting into the LQR framework, the optimal feedback control law is solved to obtain the reference steering angle. :

[0043] (4)

[0044] In the formula, The state feedback gain matrix is ​​obtained by LQR optimization. The pre-aiming state vector;

[0045] Step A6. Based on the vehicle steering geometry, the reference steering angle is calculated using the inter-axle steering ratio coefficient. Mapped to steering commands for each axis:

[0046] (5)

[0047] In the formula, For the first Shaft steering angle, For the first The steering ratio coefficient is set at the front steering axle, with the coefficient for the front steering axle being positive and the coefficient for the rear steering axle being negative, in order to achieve multi-axle coordinated steering.

[0048] 1.2 The longitudinal control module employs a seven-segment S-curve velocity planner to apply segmented constraints to longitudinal acceleration and jerk (Jerk). The specific steps are as follows:

[0049] Step B1. Determine the planning parameters, and input the initial state, including velocity. acceleration The target state includes speed. acceleration And set the maximum allowable acceleration. and maximum jerk ;

[0050] Step B2. Figure 4 As shown, the entire acceleration / deceleration process is divided into seven stages:

[0051] (6)

[0052] In the formula, Let be the maximum accelerometer, and be the derivative of the accelerometer.

[0053] Step B3. By solving the piecewise time integral equations, the duration of each stage is accurately calculated. to And the total time of the entire planning process. In each control cycle, the target velocity at the current moment is calculated in real time by interpolation based on the planned segmented time parameters and kinematic formulas. and target acceleration ;

[0054] 2. Specific implementation of the lower-level stability and torque distribution controller:

[0055] 2.1 Yaw Moment Decision Module: An equivalent sliding mode control algorithm is adopted. Its sliding mode surface function is constructed based on a linear combination of the yaw rate tracking error and the center of mass sideslip angle tracking error. The equivalent sliding mode control law of the algorithm consists of an equivalent control term and a switching control term. The specific steps are as follows:

[0056] Step C1. The desired yaw rate output by the upper-level adaptive pre-aiming linear quadratic regulator... As the tracking target, the expected centroid sideslip angle is... Set to 0 to maintain the vehicle's neutral steering characteristics;

[0057] Step C2. Define the sliding surface function. ,in These are weighting coefficients used to balance yaw rate tracking and sideslip suppression;

[0058] Step C3. Equivalent sliding mode control law It consists of equivalent control terms and switching control terms; let Assuming the system is undisturbed, the control quantity that keeps the system state on the sliding surface is derived. ; Ensure the system state approaches the sliding surface, The formula for calculating the additional yaw moment is:

[0059] (7)

[0060] In the formula, To add yaw moment, For equivalent control items, To switch control items, Given the vehicle's desired yaw rate, This refers to the vehicle's actual yaw rate. These are the vehicle's actual sideslip angle and the vehicle's expected sideslip angle, respectively. For control parameters, The moment of inertia of the vehicle body during yaw. For sliding switch gain, The yaw moment generated by the current tire force;

[0061] 2.2 Torque Vectoring Module: This module employs a quadratic programming-based algorithm. Its optimization objectives include minimizing tire adhesion utilization and tire slip ratio, and rationally distributing the total required torque and additional yaw moment to the eight wheels. The specific steps are as follows:

[0062] Step D1. Construct a multi-objective optimization problem, with the optimization variables being the drive torque of eight motors. The objective function and constraints are as follows:

[0063] (8)

[0064] In the formula, To achieve the tire adhesion utilization target, For tire slip dissipation targets, Weighting coefficients The driving torque of the wheels. This refers to the tire's longitudinal slip ratio. This refers to the tire's lateral slip ratio. This represents the total torque required by the entire vehicle. The wheelbase of the vehicle. The tire's rolling radius, For the longitudinal force of the tire, This refers to the lateral force of the tire. This represents the maximum road surface adhesion for the wheels. This represents the peak torque of the motor.

[0065] Step D2. Construct equality constraints based on the inter-axis yaw moment coefficient, and define the distribution rules for the inter-axis yaw moment coefficient.

[0066] (9)

[0067] During the optimization process, the torque distribution of each axis must meet the following requirements. proportional relationship;

[0068] In the formula, For the first Torque distribution weight of the shaft, For the first The driving torque of the shaft, For the first Distance from the axle to the vehicle's center of gravity The turning radius;

[0069] Step D3. Rearrange the objective function and constraints into standard QP form, and solve them online using the QP algorithm to obtain the optimal torque distribution vector. ;

[0070] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-axle distributed drive vehicle trajectory tracking and stability collaborative control method, characterized in that, A hierarchical control architecture is adopted, including: The upper-level trajectory tracking controller is used to generate target motion commands for controlling the vehicle, including: Lateral control module: dynamically adjusts the aiming distance based on the vehicle's real-time longitudinal speed and longitudinal acceleration, and calculates and outputs steering angle commands based on the aiming error; Longitudinal control module: Constrains and smooths the longitudinal acceleration of the vehicle and its rate of change, and generates target velocity curve and target acceleration curve; The lower-level stability and torque distribution controller executes commands from the upper level and maintains vehicle yaw stability, including: Yaw moment decision module: Calculates the additional yaw moment required for the vehicle body to maintain stability based on the vehicle's yaw rate error and center of gravity sideslip angle error; Torque vector distribution module: Based on the driver's total required torque and the additional yaw moment, the drive torque is distributed to each wheel hub motor through multi-objective optimization calculation.

2. The trajectory tracking and stability coordination control method for a multi-axle distributed drive vehicle according to claim 1, characterized in that, The lateral control module employs an adaptive pre-aiming linear quadratic regulator based on fuzzy logic inference. The fuzzy logic inference outputs a dynamic pre-aiming time based on vehicle speed and acceleration. The specific steps are as follows: Step 1. Output dynamic preview time by fuzzy inference rules according to vehicle longitudinal speed and longitudinal acceleration ; Step 2. Based on the dynamic pre-aiming time and current vehicle status By transforming the Frenet coordinate system to the Cartesian coordinate system, the target point for the future trajectory is determined. The calculation formula is as follows: （1） In the formula, For vehicle location, For the longitudinal speed of the vehicle, For the vehicle's lateral speed, This refers to the vehicle's yaw angle; Step 3. Calculate the aiming point With trajectory reference point Lateral error between aiming and heading error Construct the pre-aiming state vector The calculation formula is as follows: （2） In the formula, The heading angle of the trajectory reference point. This is the unit normal vector at the trajectory reference point; Step 4. Pre-aiming state vector Substituting into the LQR framework, the reference steering angle is obtained by solving the optimal feedback control law: （3） In the formula, The state feedback gain matrix is ​​obtained by LQR optimization. The pre-aiming state vector; Step 5. Based on the vehicle steering geometry, adjust the reference steering angle using the inter-axle steering ratio coefficient. Mapped to steering commands for each axis: （4） In the formula, For the first Shaft steering angle, For the first The steering ratio coefficient is set at the front steering axle, with a positive coefficient for the front steering axle and a negative coefficient for the rear steering axle, to achieve multi-axle coordinated steering.

3. The method for coordinated control of trajectory tracking and stability of a multi-axis distributed drive vehicle according to claim 1, characterized in that, The longitudinal control module employs a seven-segment S-curve velocity planner to impose segmented constraints on longitudinal acceleration and jerk.

4. The method for coordinated control of trajectory tracking and stability of a multi-axis distributed drive vehicle according to claim 1, characterized in that, The yaw moment decision module employs an equivalent sliding mode control algorithm. Its sliding mode surface function is constructed based on a linear combination of the yaw angular velocity tracking error and the center of mass sideslip angle tracking error. The equivalent sliding mode control law of the algorithm consists of an equivalent control term and a switching control term. The calculation formulas for the sliding mode surface function and the additional yaw moment are as follows: （5） In the formula, To add yaw moment, For equivalent control items, To switch control items, Given the vehicle's desired yaw rate, This refers to the vehicle's actual yaw rate. and These are the vehicle's actual sideslip angle and the vehicle's expected sideslip angle, respectively. For control parameters, The moment of inertia of the vehicle body during yaw. For sliding switch gain, This is the yaw moment generated by the current tire force.

5. The method for coordinated control of trajectory tracking and stability of a multi-axis distributed drive vehicle according to claim 1, characterized in that, The torque vector allocation module employs a quadratic programming-based algorithm, whose optimization objectives include minimizing tire adhesion utilization and tire slip ratio. The torque allocation optimization problem based on the quadratic programming algorithm is expressed as follows: （6） In the formula, To achieve the tire adhesion utilization target, For tire slip dissipation targets, Weighting coefficients For the first Axis No. The driving torque of the side wheels, This represents the total torque required by the entire vehicle. The wheelbase of the vehicle. The tire's rolling radius, For the first Axis No. Side tire longitudinal force, For the first Axis No. Maximum road surface adhesion of the side wheels This represents the peak torque of the motor.

6. The method for coordinated control of trajectory tracking and stability of a multi-axis distributed drive vehicle according to claim 5, characterized in that, The torque distribution optimization problem, and the multi-objective optimization algorithm, are designed to address actuator redundancy in multi-axis drive systems. The constraints also include factors based on the inter-axis yaw moment coefficient. The equality constraint is specifically defined as follows: the distribution ratio of the driving torque of each shaft is determined by the inter-shaft yaw moment coefficient. The decision, calculation formula, and constraint relationships are as follows: （7） During the optimization process, the torque distribution of each axis must meet the following requirements. proportional relationship; In the formula, For the first Torque distribution weight of the shaft, For the first The driving torque of the shaft, For the first Distance from the axle to the vehicle's center of gravity This is the turning radius.

Images