MPC-based automatic driving stability rolling optimization control method and system

Abstract

The invention relates to an MPC-based automatic driving stability rolling optimization control method and system, and belongs to the field of automatic driving. The method comprises the steps that automatic driving state parameters are detected, automatic driving state sensing is conducted on the basis of the automatic driving state parameters, and a vehicle state model and actual tire lateral force are obtained; when the actual lateral force of the tire exceeds a preset threshold range, a stable expected path is generated according to the automatic driving state parameters; performing hierarchical rolling optimization control based on the stability expected path, and outputting an execution instruction; and regulating and controlling an automatic driving strategy of the vehicle based on the execution instruction, updating the vehicle state model and the actual lateral force of the tire, and repeating the steps until automatic driving is finished. Layered rolling optimization control of automatic driving is realized, and stable driving of the vehicle under different working conditions is guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of autonomous driving, and particularly relates to a rolling optimization control method and system for autonomous driving stability based on MPC. Background Art

[0002] Autonomous driving technology is a popular field that has developed rapidly in recent years. It integrates technologies from multiple fields such as machine learning, computer vision, path planning, and control theory. It identifies the road environment through computer vision, sensor data, and other information, and then realizes the autonomous control of the vehicle. With the continuous progress of technologies such as artificial intelligence, autonomous driving technology has made remarkable progress and has also been applied to certain scenarios such as intelligent parking systems. However, at present, this technology still faces many challenges, and its stability and reliability need to be further improved. Previously, the industry mainly used LQR and PID to control autonomous driving stability. Due to their simple calculation, mature parameter adjustment, and ability to meet the needs of most scenarios, currently, model predictive control (MPC) has gradually replaced traditional PID and LQR control methods in the field of autonomous driving. MPC significantly improves the trajectory tracking accuracy and robustness by predicting the future state of the vehicle and solving the optimal control sequence. However, the control variable rate constraints of traditional MPC (such as the steering wheel angle change rate) are mostly fixed values and do not consider the dynamic changes of different vehicle speeds, curvatures, or road surface adhesion coefficients, resulting in a decline in control effects under high-speed sharp turns or low-adhesion road surfaces. Moreover, the simplification of the vehicle dynamics model (such as ignoring the nonlinear characteristics of tires or air resistance) will lead to the accumulation of prediction errors, especially under high speeds or low-adhesion road surfaces, where the errors are significantly amplified. Summary of the Invention

[0003] To solve the above problems existing in the prior art, the present invention provides a rolling optimization control method and system for autonomous driving stability based on MPC.

[0004] The object of the present invention can be achieved by the following technical solutions:

[0005] A rolling optimization control method for autonomous driving stability based on MPC. The implementation of the rolling optimization control method for autonomous driving stability includes the following steps:

[0006] S1: Detect the autonomous driving state parameters, perform autonomous driving state perception based on the autonomous driving state parameters, and obtain the vehicle state model and the actual lateral force of the tire;

[0007] S2: When the actual lateral force of the tire exceeds the preset threshold range, generate a stable desired path according to the autonomous driving state parameters;

[0008] S3: Perform hierarchical rolling optimization control based on the stable desired path and output an execution instruction;

[0009] S4: Regulate the vehicle's autonomous driving strategy based on the execution instruction, and update the vehicle state model and the actual lateral force of the tire. Repeat steps S1 - S4 until the autonomous driving ends.

[0010] Preferably, step S1 specifically includes:

[0011] S101: Obtain the autonomous driving state parameters, which include lane obstacle parameters and vehicle state parameters;

[0012] S102: Sense the vehicle state based on the autonomous driving state parameters and construct the vehicle state model;

[0013] S103: Sense the tire state based on the autonomous driving state parameters to obtain the actual lateral force of the tire.

[0014] Preferably, the construction of the vehicle state model in step S102 specifically includes:

[0015] S102 - 1: Obtain the vehicle vertical load based on the autonomous driving state parameters, where the vehicle vertical load includes the front wheel vertical load and the rear wheel vertical load;

[0016] S102 - 2: Construct the vehicle state model based on the vehicle vertical load and the autonomous driving state parameters.

[0017] Preferably, step S102 - 2 specifically includes:

[0018] Describe the longitudinal and lateral accelerations and the vehicle yaw angular acceleration during the vehicle's autonomous driving process through the vehicle vertical load and the autonomous driving state parameters to achieve vehicle state perception. The mathematical description is Where, is the vehicle longitudinal acceleration, is the vehicle lateral acceleration, is the vehicle yaw angular acceleration, m is the vehicle mass, F x is the longitudinal driving force, ρ is the air density, C d is the wind resistance coefficient, A is the vehicle frontal area, v x is the vehicle longitudinal speed, v y is the vehicle lateral speed, is the vehicle yaw angular speed, F zf is the front wheel vertical load, F zr is the rear wheel vertical load, I Z is the yaw moment of inertia, l f is the vehicle front wheel wheelbase, l r is the vehicle rear wheel wheelbase.

[0019] Preferably, the selection of the stable desired path in step S2 specifically includes:

[0020] S201: Generate an obstacle avoidance path set based on the lane obstacle parameters;

[0021] S202: Set the strategic layer optimization goal and the tactical layer optimization goal, traverse the obstacle avoidance path set, and select the stable desired path based on the strategic layer optimization goal and the tactical layer optimization goal.

[0022] Preferably, the mathematical description of the strategic layer optimization goal is where T is the required duration of the obstacle avoidance path, ω1, ω2, and ω3 are the safety distance weight, energy consumption weight, and acceleration change rate weight respectively, d w is the safety distance violation value, E is the energy consumption, △a is the acceleration change rate, d safe is the standard safety distance, E ref is the reference energy consumption, △a ref is the reference acceleration change rate; the mathematical description of the tactical layer optimization goal is where K is the number of operations required to complete the obstacle avoidance path, δ k is the steering angle of the kth operation, δ ref is the reference steering angle, △a k is the acceleration change rate of the kth operation.

[0023] Preferably, the generation of the execution instruction in step S3 specifically includes:

[0024] S301: The vehicle travels according to the stable desired path and generates a first weight matrix and a second weight matrix;

[0025] S302: Based on the first weight matrix and the second weight matrix, optimize and solve the tactical layer MPC problem in real time to obtain a control sequence, and generate an execution instruction based on the control sequence.

[0026] Preferably, step S301 specifically includes:

[0027] S301-1: Obtain the longitudinal control weight and the lateral control weight and generate the first weight matrix;

[0028] S301-2: Set the second weight matrix.

[0029] Preferably, step S302 specifically includes:

[0030] The mathematical description of the tactical layer MPC problem is where q lat is the lateral control weight, q lon is the longitudinal control weight, y kThe lateral displacement for the k-th operation, y ref,k The reference value of the lateral displacement for the k-th operation, v k The vehicle speed for the k-th operation, δ k The steering angle for the k-th operation, △a k The acceleration change rate for the k-th operation, r δ The steering angle weight, r Δa The acceleration change weight, v ref,k The reference vehicle speed for the k-th operation.

[0031] An MPC-based stability rolling optimization control system for autonomous driving, which is used to execute the above-mentioned autonomous driving stability rolling optimization control method, includes a vehicle state perception module, a path generation module, a rolling optimization control module, and an autonomous driving regulation module;

[0032] The vehicle state perception module is used to detect the autonomous driving state parameters, perceive the autonomous driving state based on the autonomous driving state parameters, and obtain the vehicle state model and the actual lateral force of the tire;

[0033] The path generation module is used to generate a stable desired path according to the autonomous driving state parameters when the actual lateral force of the tire exceeds the preset threshold range;

[0034] The rolling optimization control module is used to perform hierarchical rolling optimization control based on the stable desired path and output an execution instruction;

[0035] The autonomous driving regulation module is used to regulate the vehicle autonomous driving strategy based on the execution instruction, and update the vehicle state model and the actual lateral force of the tire, and repeat the above steps until the autonomous driving ends.

[0036] The beneficial effects of the present invention are:

[0037] (1) An adaptive constraint for autonomous driving is proposed to avoid understeering or oversteering, and improve the stability in wet road surface and sharp turn scenarios.

[0038] (2) By introducing a dynamic weight adjustment mechanism, the priority of the control target is dynamically adjusted according to the lateral error, vehicle speed, road surface friction coefficient, etc.

[0039] (3) By adopting a strategic layer - tactical layer double-loop architecture, the calculation efficiency is improved through hierarchical optimization. The strategic layer plans the global reference trajectory, considering energy consumption and comfort; the tactical layer performs local optimization to handle the vehicle body dynamics constraints.

[0040] (4) A dual-mode tire lateral force formula is proposed to distinguish the linear and non-linear working ranges (i.e., small side slip angle conditions and large side slip angle conditions), reducing the calculation error on wet road surfaces. Description of the Drawings

[0041] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with the accompanying drawings.

[0042] Figure 1 It is a flowchart of the steps of a method for rolling optimization control of the stability of an autonomous vehicle based on MPC according to the present invention. Specific embodiments

[0043] To better understand the present invention, more detailed descriptions of various aspects of the present invention will be made with reference to the accompanying drawings. It should be understood that these detailed descriptions are only descriptions of the exemplary embodiments of the present invention, and do not limit the scope of the present invention in any way. Throughout the specification, the expression "and / or" includes any and all combinations of one or more of the associated listed items. As used herein, terms such as "substantially", "about" and similar terms are used as terms indicating approximation, rather than terms indicating degree, and are intended to account for the inherent deviations in measured or calculated values that would be recognized by those of ordinary skill in the art. Additionally, in the present invention, the order of description of the steps of each process does not necessarily represent the order in which these processes occur in actual operation, unless otherwise clearly defined or derivable from the context.

[0044] It should also be understood that expressions such as "including", "comprising", "having", "containing" and / or "comprising of" are open-ended rather than closed-ended expressions in this specification, which means that the stated features, elements and / or components exist, but do not exclude the existence of one or more other features, elements, components and / or their combinations. In addition, when an expression such as "at least one of..." appears after a list of listed features, it modifies the entire list of features, rather than just individual elements in the list. In addition, when describing the embodiments of the present invention, the use of "may" means "one or more embodiments of the present invention". And, the term "exemplary" is intended to refer to an example or illustration.

[0045] Unless otherwise defined, all terms used herein (including engineering terms and scientific and technical terms) have the same meaning as the ordinary understanding of those of ordinary skill in the art to which the present invention pertains. It should also be understood that, unless clearly stated in the present invention, words defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art, and should not be interpreted in an idealized or overly formal sense.

[0046] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0047] Example 1:

[0048] Please refer to Figure 1 , a stability rolling optimization control method for autonomous driving based on MPC, including:

[0049] S1: Detect the autonomous driving state parameters, perform autonomous driving state perception based on the autonomous driving state parameters, and obtain the vehicle state model and the actual lateral force of the tire;

[0050] S2: When the actual lateral force of the tire exceeds the preset threshold range, generate a stable desired path according to the autonomous driving state parameters;

[0051] S3: Perform hierarchical rolling optimization control based on the stable desired path and output an execution instruction;

[0052] S4: Regulate the autonomous driving strategy of the vehicle based on the execution instruction (i.e., adjust the vehicle speed, steering angle, etc. in stages), and update the vehicle state model and the actual lateral force of the tire, and repeat steps S1 to S4 until the autonomous driving ends.

[0053] In this embodiment, performing autonomous driving state perception based on the autonomous driving state parameters can be specifically implemented through the following steps:

[0054] S101: Obtain the autonomous driving state parameters through sensors. The autonomous driving state parameters include lane obstacle parameters and vehicle state parameters. The lane obstacle parameters include obstacle coordinates and relative speeds of obstacles. The vehicle state parameters include vehicle frontal area, vehicle mass, vehicle front wheelbase, vehicle rear wheelbase, vehicle yaw rate, vehicle acceleration, vehicle center of mass height, vehicle front wheelbase, vehicle rear wheelbase, tire side slip angle, tire side slip stiffness, maximum lateral force of the tire, longitudinal driving force, vehicle longitudinal speed, and vehicle lateral speed;

[0055] S102: Perceive the vehicle state based on the autonomous driving state parameters and construct the vehicle state model;

[0056] S102-1: Obtain the vehicle vertical load based on the autonomous driving state parameters. The vehicle vertical load includes the front wheel vertical load and the rear wheel vertical load. The mathematical description of the front wheel vertical load is The mathematical description of the rear wheel vertical load is where m is the vehicle mass, g is the acceleration due to gravity, l f is the vehicle front wheelbase, l r is the vehicle rear wheelbase, a is the vehicle acceleration, and h is the vehicle center of mass height;

[0057] S102-2: Construct the vehicle state model based on the vehicle vertical load and the autonomous driving state parameters, describe the longitudinal and lateral accelerations and the vehicle yaw angular acceleration during the vehicle's autonomous driving, realize vehicle state perception, and determine whether the vehicle is slightly deviated due to factors such as wind during the vehicle's autonomous driving. The mathematical description is as follows where is the vehicle longitudinal acceleration (m / s 2 ), is the vehicle lateral acceleration (m / s 2 ), is the vehicle yaw angular acceleration (rad / s 2 ), m is the vehicle mass (kg), F x is the longitudinal driving force (N), generated by the engine / braking system, ρ is the air density (kg / m 3 ), C d is the drag coefficient (dimensionless), A is the vehicle frontal area (m 2 ), v x is the vehicle longitudinal speed (m / s), v y is the vehicle lateral speed (m / s), is the vehicle yaw angular speed (rad / s), F zf is the front wheel vertical load (N), F zr is the rear wheel vertical load (N), I Z is the yaw moment of inertia (kg / m 2 ), l f is the vehicle front wheelbase (m), l r is the vehicle rear wheelbase (m).

[0058] S103: Perceive the tire state based on the autonomous driving state parameters, solve the deficiency that the traditional model assumes that the tire grips the ground completely like an eraser (i.e., assume that the vehicle always has four wheels on the ground), and there will be slippage on a wet road surface, correct the possible deviation in the tire force prediction, obtain the actual lateral force of the tire, and estimate the effective grip force in real time. The mathematical description of the actual lateral force of the tire is as follows where F y is the actual lateral force of the tire (N), C α is the tire cornering stiffness (N / rad), α is the tire slip angle (rad), α max is the maximum effective slip angle of the tire (about 0.14 rad for dry ground and about 0.05 rad for wet ground), λ is the attenuation rate coefficient (1 / rad), is the maximum lateral force of the tire (N).

[0059] In this embodiment, a stable desired path is generated according to the autonomous driving state parameters, and it can be specifically implemented through the following steps:

[0060] S201: Generate an obstacle avoidance path set based on the lane obstacle parameters (that is, through the position and relative speed of the obstacles ahead, combined with the vehicle's own driving conditions and road conditions, intelligently plan multiple automatic obstacle avoidance paths);

[0061] S202: Set the strategic layer optimization goal and the tactical layer optimization goal. The mathematical description of the strategic layer optimization goal is where T is the required duration of the obstacle avoidance path, ω1, ω2, and ω3 are the safety distance weight, energy consumption weight, and acceleration change rate weight respectively, d w is the safety distance violation value, E is the energy consumption, △a is the acceleration change rate, d safe is the standard safety distance, E ref is the reference energy consumption, △a ref is the reference acceleration change rate; the mathematical description of the tactical layer optimization goal is where K is the number of operations required to complete the obstacle avoidance path, δ k is the steering angle of the k-th operation, δ ref is the reference steering angle, △a k is the acceleration change rate of the k-th operation. Traverse the obstacle avoidance path set and select the stable expected path based on the strategic layer optimization goal and the tactical layer optimization goal.

[0062] In this embodiment, hierarchical rolling optimization control is performed based on the stable expected path, which can be specifically implemented through the following steps:

[0063] S301: The vehicle drives according to the stable expected path and generates a first weight matrix and a second weight matrix;

[0064] S301-1: Obtain the longitudinal control weight and the lateral control weight and generate the first weight matrix. The first weight matrix is used to select the autonomous driving control mode. When the longitudinal control weight is large, it is similar to gently holding the steering wheel to maintain the vehicle speed during manual driving. When the lateral control weight is large, it is similar to suddenly turning the steering wheel to brake sharply or slowly turning the steering wheel and cooperating with point braking during manual driving. The expression of the longitudinal control weight is where is the basic longitudinal weight (default value 0.8 - 1.2), γ1 and γ2 are adjustment coefficients, v is the real-time vehicle speed, v ref is the target vehicle speed, d actual is the actual distance from the vehicle in front; the expression of the lateral control weight is where u1, u2, and u3 are the centrifugal force influence coefficient, deviation correction demand coefficient, and road surface safety coefficient respectively, μ is the road surface friction coefficient, R is the turning radius, e lat is the lateral error (m), e refis the reference error (m); the expression of the first weight matrix is S301-2: Set the second weight matrix, and the mathematical description is where R is the second weight matrix, and r δ is the steering angle weight, and r Δa is the acceleration change weight.

[0065] S302: Based on the first weight matrix and the second weight matrix, optimize and solve the tactical layer MPC problem in real time to obtain a control sequence, and generate an execution instruction based on the control sequence. The mathematical description of the tactical layer MPC problem is where y k is the lateral displacement of the kth operation, and y ref,k is the reference value of the lateral displacement of the kth operation, v k is the vehicle speed of the kth operation, and v ref,k is the reference vehicle speed of the kth operation.

[0066] Embodiment 2:

[0067] An MPC-based stability rolling optimization control system for autonomous driving includes a vehicle state perception module, a path generation module, a rolling optimization control module, and an autonomous driving regulation module;

[0068] The vehicle state perception module is used to detect autonomous driving state parameters, and based on the autonomous driving state parameters, perform autonomous driving state perception to obtain a vehicle state model and the actual lateral force of the tire. Specifically:

[0069] Obtain the autonomous driving state parameters through sensors. The autonomous driving state parameters include lane obstacle parameters and vehicle state parameters. The lane obstacle parameters include obstacle coordinates and relative speeds of obstacles. The vehicle state parameters include vehicle frontal area, vehicle mass, vehicle front wheelbase, vehicle rear wheelbase, vehicle yaw rate, vehicle acceleration, vehicle center of mass height, vehicle front wheelbase, vehicle rear wheelbase, tire side slip angle, tire cornering stiffness, maximum lateral force of the tire, longitudinal driving force, vehicle longitudinal speed, and vehicle lateral speed;

[0070] Perceive the vehicle state based on the autonomous driving state parameters and construct the vehicle state model: Obtain the vehicle vertical load based on the autonomous driving state parameters. The vehicle vertical load includes the front wheel vertical load and the rear wheel vertical load; construct the vehicle state model based on the vehicle vertical load and the autonomous driving state parameters to describe the longitudinal and lateral accelerations and the vehicle yaw angular acceleration during the vehicle's autonomous driving process, realize vehicle state perception, and can judge whether the vehicle is slightly deviated due to factors such as wind during the vehicle's autonomous driving process.

[0071] Perceive the tire state based on the autonomous driving state parameters, solve the deficiency that the traditional model assumes that the tire grips the ground completely like an eraser (i.e., assumes that the vehicle always has all four wheels on the ground), and there will be skidding on a slippery road surface in reality, correct the possible deviation in tire force prediction, obtain the actual lateral force of the tire, and estimate the effective grip force in real time.

[0072] The path generation module is used to generate a stable expected path according to the autonomous driving state parameters when the actual lateral force of the tire exceeds the preset threshold range. Specifically:

[0073] Generate an obstacle avoidance path set based on the lane obstacle parameters (that is, through the position and relative speed of the front obstacle, etc., combined with the vehicle's own driving conditions and road conditions, intelligently plan multiple automatic obstacle avoidance paths);

[0074] Set the strategic layer optimization goal and the tactical layer optimization goal, traverse the obstacle avoidance path set, and select the stable expected path based on the strategic layer optimization goal and the tactical layer optimization goal.

[0075] The rolling optimization control module is used to perform hierarchical rolling optimization control based on the stable expected path and output an execution instruction. Specifically:

[0076] The vehicle drives according to the stable expected path and generates a first weight matrix and a second weight matrix: obtain the longitudinal control weight and the lateral control weight and generate the first weight matrix. The first weight matrix is used to select the autonomous driving control mode. When the longitudinal control weight is large, it is similar to gently holding the steering wheel to maintain the vehicle speed during manual driving. When the lateral control weight is large, it is similar to suddenly turning the steering wheel and braking sharply or slowly turning the steering wheel and coordinating with point braking during manual driving; set the second weight matrix.

[0077] Based on the first weight matrix and the second weight matrix, optimize and solve the tactical layer MPC problem in real time to obtain a control sequence, and generate an execution instruction based on the control sequence.

[0078] The autonomous driving regulation module is used to regulate the vehicle's autonomous driving strategy based on the execution instruction (that is, adjust the vehicle speed, steering angle, etc. in batches), and update the vehicle state model and the actual lateral force of the tire, and repeat steps S1 - S4 until the autonomous driving ends.

[0079] The above are only the preferred embodiments of the present invention and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above in the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments by using the above-disclosed technical content without departing from the technical solution of the present invention. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. An MPC-based stability rolling optimization control method for autonomous driving, characterized in that, The implementation of the above-mentioned roll optimization control method for autonomous driving stability includes the following steps: S1: Detect the autonomous driving state parameters, perform autonomous driving state perception based on the autonomous driving state parameters, and obtain the vehicle state model and the actual lateral force of the tires; S2: When the actual lateral force of the tires exceeds the preset threshold range, generate a stable desired path according to the autonomous driving state parameters; S3: Perform hierarchical roll optimization control based on the stable desired path and output an execution instruction; S4: Regulate the autonomous driving strategy of the vehicle based on the execution instruction, and update the vehicle state model and the actual lateral force of the tires. Repeat steps S1 - S4 until the autonomous driving ends.

2. The roll optimization control method for autonomous driving stability according to claim 1, wherein The specific content of step S1 includes: S101: Obtain the autonomous driving state parameters, where the autonomous driving state parameters include lane obstacle parameters and vehicle state parameters; S102: Perceive the vehicle state based on the autonomous driving state parameters and construct the vehicle state model; S103: Perceive the tire state based on the autonomous driving state parameters to obtain the actual lateral force of the tires.

3. The automatic driving stability rolling optimization control method according to claim 2, characterized in that, The construction of the vehicle state model in step S102 specifically includes: S102 - 1: Obtain the vehicle vertical load based on the autonomous driving state parameters, where the vehicle vertical load includes the front wheel vertical load and the rear wheel vertical load; S102 - 2: Construct the vehicle state model based on the vehicle vertical load and the autonomous driving state parameters.

4. The roll optimization control method for autonomous driving stability according to claim 3, characterized in that The specific content of step S102 - 2 includes: Describe the longitudinal and lateral accelerations and the vehicle yaw angular acceleration during the vehicle's autonomous driving process through the vehicle vertical load and the autonomous driving state parameters, realizing vehicle state perception. The mathematical description is as follows where is the vehicle longitudinal acceleration, is the vehicle lateral acceleration, is the vehicle yaw angular acceleration, m is the vehicle mass, F x is the longitudinal driving force, ρ is the air density, C d is the drag coefficient, A is the vehicle frontal area, v x is the vehicle longitudinal speed, v y is the vehicle lateral speed, is the vehicle yaw angular speed, F zf is the front wheel vertical load, F zr is the rear wheel vertical load, I Z is the yaw moment of inertia, l f is the vehicle front wheel wheelbase, l r is the vehicle rear wheel wheelbase.

5. The automatic driving stability rolling optimization control method according to claim 2, characterized in that, The selection of the stable desired path in step S2 specifically includes: S201: Generate an obstacle avoidance path set based on the lane obstacle parameters; S202: Set the strategic layer optimization target and the tactical layer optimization target, traverse the obstacle avoidance path set, and select the stable desired path based on the strategic layer optimization target and the tactical layer optimization target.

6. The roll optimization control method for autonomous driving stability according to claim 5, characterized in that The mathematical description of the optimization objective at the strategic level is where T is the time required for the obstacle avoidance path, ω1, ω2, and ω3 are the safety distance weight, energy consumption weight, and acceleration change rate weight respectively, d w is the safety distance violation value, E is the energy consumption, △a is the acceleration change rate, d safe is the standard safety distance, E ref is the reference energy consumption, △a ref is the reference acceleration change rate; the mathematical description of the optimization objective at the tactical level is where K is the number of operations required to complete the obstacle avoidance path, δ k is the steering angle of the k-th operation, δ ref is the reference steering angle, △a k is the acceleration change rate of the k-th operation.

7. The automatic driving stability rolling optimization control method according to claim 1, characterized in that The generation of the execution instruction in step S3 specifically includes: S301: The vehicle drives according to the stable desired path and generates a first weight matrix and a second weight matrix; S302: Real - time optimize and solve the tactical layer MPC problem based on the first weight matrix and the second weight matrix to obtain a control sequence, and generate an execution instruction based on the control sequence.

8. The automatic driving stability rolling optimization control method according to claim 7, characterized in that, The specific content of step S301 includes: S301 - 1: Obtain the longitudinal control weight and the lateral control weight and generate the first weight matrix; S301 - 2: Set the second weight matrix.

9. The roll optimization control method for autonomous driving stability according to claim 7, wherein The specific content of step S302 includes: The mathematical description of the tactical layer MPC problem is as follows where q lat is the lateral control weight, q lon is the longitudinal control weight, y k is the lateral displacement at the k-th operation, y ref,k is the reference value of the lateral displacement at the k-th operation, v k is the vehicle speed at the k-th operation, δ k is the steering angle at the k-th operation, △a k is the acceleration change rate at the k-th operation, r δ is the steering angle weight, r Δa is the acceleration change weight, v ref,k is the reference vehicle speed at the k-th operation.

10. An MPC-based stability rolling optimization control system for autonomous driving, characterized in that, The system is applied to the roll optimization control method for autonomous driving stability as described in any one of claims 1 - 9, and includes a vehicle state perception module, a path generation module, a roll optimization control module, and an autonomous driving regulation module; The vehicle state perception module is used to detect the autonomous driving state parameters, perform autonomous driving state perception based on the autonomous driving state parameters, and obtain the vehicle state model and the actual lateral force of the tires; The path generation module is used to generate a stable desired path according to the autonomous driving state parameters when the actual lateral force of the tires exceeds the preset threshold range; The rolling optimization control module is used to perform hierarchical rolling optimization control based on the stable expected path and output execution instructions; The automatic driving regulation module is used to regulate the vehicle's automatic driving strategy based on the execution instructions, update the vehicle state model and the actual lateral force of the tire, and repeat the above steps until the automatic driving ends.

Images