Path planning with Rapid Exploring Random Tree for autonomous race vehicle

Autonomous vehicle research and development is emerging in recent years globally aimed to provide a highly reliable, secure, mobile, and intelligent transportation system. The procedure involves integration of multiple systems together to ensure safe driving. The functional reference standard architecture of the autonomous driving system is classified into three main categories: perception, decision & control, and actuation. Mainly the challenging part is decision making that requires high-level planning and vehicle control to accomplish the driving mission. This thesis work is a small contribution to design and develop an autonomous formula student racing vehicle, and it mainly focuses on the study of decision making strategies and vehicle control system. The experimentation aimed at the performance evaluation of a model-based control system and defining motion planning strategies for better vehicle navigation. In the first part of the thesis, a previously designed Model predictive based control system for lane following is validated on three different race track scenarios to evaluate the performance of model. In the second part of the thesis, development of vehicle motion planning and tracking control system using a model based algorithm to provide a highly safe navigation path to follow. For experimentation purpose a simplified vehicle model is used to validate the tracking control system, and suitable improvements are made based on Simulation results to reduce high-level tracking errors considering handling and safety limits. MATLAB® and Simulink® are the modern simulation tools that are used in this project work to perform all the analysis.