Research on FSAC trajectory tracking control based on optimized BP neural network algorithm

Zhiqiang Zhang, Gang Li, Zhixin Chen, Xing Zhang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

A longitudinal linear quadratic regulation LQR acceleration motion controller and a lateral linear model prediction (LTV-MPC) motion controller are designed to address the accuracy and stability of trajectory tracking with a four-wheel independent drive driverless formula car. Based on the BP neural network algorithm, the prediction step and control step parameters of the model prediction control are adaptively adjusted, and the genetic optimisation algorithm is used to optimise the BP neural network to improve the lateral trajectory tracking accuracy of the car. The simulation results show that the proposed lateral motion control strategy can control the unmanned racing car to track the lateral trajectory well during the trajectory tracking process.

Original languageEnglish
Title of host publicationProceedings of the 2023 7th CAA International Conference on Vehicular Control and Intelligence, CVCI 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798350340488
DOIs
Publication statusPublished - 2023
Externally publishedYes
Event7th CAA International Conference on Vehicular Control and Intelligence, CVCI 2023 - Changsha, China
Duration: 27 Oct 202329 Oct 2023

Publication series

NameProceedings of the 2023 7th CAA International Conference on Vehicular Control and Intelligence, CVCI 2023

Conference

Conference7th CAA International Conference on Vehicular Control and Intelligence, CVCI 2023
Country/TerritoryChina
CityChangsha
Period27/10/2329/10/23

Keywords

  • adaptive
  • formula car
  • four-wheel independent drive
  • model predictive control
  • trajectory tracking

Fingerprint

Dive into the research topics of 'Research on FSAC trajectory tracking control based on optimized BP neural network algorithm'. Together they form a unique fingerprint.

Cite this