Layout Option
Master of Science in Engineering
Formulation and Comparison of Coupled Path Tracking and Vehicle Dynamic Controls Using Co-Simulation
Date of Award
2024
Document Type
Thesis
Degree Name
Engineering
Department
Engineering
First Advisor
Dr. Diane Peters
Second Advisor
Dr. Craig Hoff
Third Advisor
Dr. Jennifer Bastiaan
Abstract
Autonomous vehicles are expected to increase in the future, with most if not all of them being electric vehicles possible with all-wheel drive layout. Autonomous vehicles utilize path-tracking algorithm which comprises lateral and longitudinal motion controls. Fundamentally the lateral controllers provide steering angle inputs which are general slip angles on front and rear axles, generating lateral force which finally produces lateral acceleration shifting the car laterally. This response is highly nonlinear as characterized by the understeer gradient characteristics of the vehicle. This non-linearity can cause deterioration in tracking performance at high lateral accelerations. With a wheel drive layout and the ability to distribute wheel torques, it may be possible to linearize vehicle response using vehicle dynamic controls which should improve the performance of the lateral control algorithms. Briefly summarizing this work, an architecture was developed where lateral control and torque vectoring controller were combined to test the aforementioned idea. Two opposite types of lateral control strategies were used – the first one was the Stanley controller which is highly linear in operation and another was state-of-the-art MPC control which can control nonlinear systems better given its optimization algorithm. Novel slip ratio control and torque vectoring strategies were developed to augment these lateral control algorithms. The controllers were then tested with a high-fidelity Carsim vehicle model in the loop. The test scenarios subjected the controllers to a variety of operating conditions such as low and high lateral acceleration, velocity variations, road camber, elevation, and friction variations. The metrics for evaluation were maximum lateral error and integral of lateral error. It was identified that in almost all the cases there was a forty to ninety percent improvement in tracking performance when torque vectoring augmented the lateral control algorithms except in the case of a double lane change simulation where MPC performed slightly worse by the addition of torque vectoring.
Recommended Citation
Panchal, Tanmay A., "Formulation and Comparison of Coupled Path Tracking and Vehicle Dynamic Controls Using Co-Simulation" (2024). Graduate Theses. 7.
https://digitalcommons.kettering.edu/graduate_theses/7
