Longitudinal Trajectory Tracking Analysis for Autonomous Electric Vehicles Based on PID Control

The implementation of autonomous vehicles has huge potential for revolutionizing transportation as we currently know it. All use cases of autonomous vehicles require the vehicle to travel on a pre-specified path. Accurate tracking of this defined trajectory is a crucial aspect of the implementation of autonomous vehicles; a controller system is required to translate this pre- defined trajectory in the form of the throttle, brake, and steering inputs. This project covers the application of a Proportional-Integral-Derivative (PID) controller to achieve longitudinal trajectory tracking of autonomous electric vehicles with stability and accuracy in the CARLA autonomous driving simulation platform. The implemented controller\u27s performance is analyzed using a three-level iterative testing approach, which actively changes the controller\u27s error definition and proportional gain. The controller\u27s performance is assessed using a bundle of 10 increasingly oscillating trajectories that are designed to disrupt the control process. The results strongly favor the error term\u27s definition to include both location deviation and speed differences in the controller\u27s error term definition. An analysis is presented on the effects of different error ratio definitions, and a final specification is introduced. This study serves as a starting point for the implementation of more advanced trajectory control methods