Chen, Haowen (Jilin University), Liu, Qifang (Jilin University), Sun, Dazhen (Jilin University), Yu, Shuyou (Jilin University), Chen, Hong (Tongji University)

Fast Model Predictive Control Trajectory Tracking for Autonomous Vehicles Based on BFGS Interior Point Method

Scheduled for presentation during the Invited session "Modeling, control and optimization of vehicle and renewable energy system" (FrB4), Friday, November 1, 2024, 11:10−11:30, Room T4

7th IFAC Conference on Engine and Powertrain Control, Simulation and Modeling, Oct 30 - Nov 1, 2024, Dalian, China

This information is tentative and subject to change. Compiled on January 2, 2025

Abstract

This paper proposes a fast solver based on the quasi-Newton interior point method for trajectory tracking in model predictive control of autonomous driving. Traditional interior point convex optimization uses Newton’s iterations to solve a series of linear systems generated during the optimization process. In this paper, a quasi-Newton algorithm (BFGS) is proposed to replace the Newton algorithm, leveraging the specific structure of trajectory tracking problems where equality constraints are eliminated when converted to a quadratic programming problem. Simulation results show that the proposed BFGS interior point method achieves a solution time that is only one-tenth of that of the traditional interior point method, greatly improving computational speed.