Vehicle Longitudinal and Lateral Dynamics Control with Model Predictive Control

Abstract

This work presents a model predictive control (MPC) to perform a double lane change maneuver. The proposed model uses the steering commands with constant velocity to quickly track the desired reference path. The prediction of future movements can help improve how the planar vehicle dynamics are controlled in physical scenarios. Different acceleration profiles are tested to verify the model performance. This paper explains the basics of MPC in the described context. It compares and contrasts the performance of the MPC-based and a PID controller while executing the same maneuver. The results were analyzed using the root-mean-square and maximum error from time-domain signals. It is shown that in some instances, the model predictive control strategy has key benefits when compared to its conventional control counterpart.