SAC-based vehicle chassis longitudinal, transverse and vertical integrated control method

By adopting a vehicle chassis longitudinal, transverse and vertical integrated control method based on SAC, and combining deep reinforcement learning and model predictive control, the problems of low integration and insufficient stability in chassis integrated control are solved, and efficient stability control of vehicles under complex working conditions is achieved.

CN122402491APending Publication Date: 2026-07-17HEFEI UNIV OF TECH
0 Cites 0 Cited by

Patent Information

Application Number
CN202610748544.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-05-28
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing integrated chassis control methods are difficult to effectively coordinate multiple subsystems in actual vehicle operation, resulting in low integration, insufficient collaborative control capabilities, and difficulty in improving vehicle stability under complex operating conditions.

Method used

A vehicle chassis longitudinal, transverse and vertical integrated control method based on SAC is adopted, which combines deep reinforcement learning, sliding mode control and model predictive control. By constructing a safety envelope and dynamic weights, the control quantity is allocated to the steer-by-wire, distributed drive and active suspension systems to realize the stability index calculation and safety constraints of the vehicle state parameter set.

Benefits of technology

It significantly improves vehicle stability under complex operating conditions, simplifies the control process, enhances the timeliness and robustness of control, and strengthens adaptability and reliability in complex vehicle dynamics control.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a vehicle chassis longitudinal, transverse and vertical integrated control method based on SAC, which is applied to the technical field of intelligent drive-by-wire chassis control and comprises the following steps: 1, defining a vehicle state parameter set and an action control set of a deep reinforcement learning algorithm SAC; 2, designing a multi-dimensional stability evaluation module MDSS to calculate vehicle multi-dimensional risk characteristic values, inputting the vehicle multi-dimensional risk characteristic values into a vehicle state parameter set, and generating a dynamic control weight adjustment action control set; 3, designing an SMC-MPC security envelope to carry out security constraint on the action control set; 4, designing a bottom layer distribution mechanism to decouple the safety control quantity to steering, driving, braking and suspension execution systems; and 5, designing an SAC network structure, and training an SAC intelligent agent to obtain an optimal strategy network. According to the method, the problems of strong nonlinear coupling of chassis multidirectional dynamics and actuator interference under the limiting working condition are effectively solved, the adaptability of data driving and the safety of traditional control are both considered, and the handling stability of the vehicle is remarkably improved.
Need to check novelty before this filing date? Find Prior Art