Design and Simulation of Integrated Lateral and Longitudinal Controller for Automated and Connected Vehicles

Abstract

This thesis is concerned with the design of an integrated longitudinal and lateral controller for a platoon of connected and automated vehicles executing an emergency lane change maneuver. The efficacy of the designed controller is shown using a simulation setup that has been built using an open-source simulator, called TORCS, in conjunction with Simulink, where the controller is designed. The longitudinal controller component consists of an Adaptive Cruise Control (ACC) System that employs a Constant Time Headway (CTH) policy. The lateral controller is based on the lead vehicle in a platoon broadcasting its GPS coordinates to all the following vehicles at regular intervals of time. This information is utilized by every following vehicle to construct a desired trajectory to track (in particular, it is a circular spline). Since the same information is used by every following vehicle, the issue of propagation of errors in the lateral direction is circumvented.

Stability robustness to variation in mass and inertia of the vehicle is crucial for ensuring that the controller can be implemented in practice. Vehicle mass and moment of inertia are most affected by the addition on passengers and cargo. The complicating part of studying the robustness stems from the characteristic polynomial being a rational function of the vehicle mass and is not readily amenable to analysis by the well-known root-locus method. A D-decomposition based method was adopt to find the maximum speed for safely/stably tracking a given reference trajectory.