A fuzzy logic based autonomous vehicle control system design in the TORCS environment

Abstract

In this study, a fuzzy logic based autonomous vehicle control system is designed and tested in The Open Racing Car Simulator (TORCS) environment. The aim of this study is that vehicle complete the race without to get any damage and to get out of the way. In this context, an intelligent control system composed of fuzzy logic and conventional control structures has been developed such that the racing car is able to compete the race autonomously. In this proposed structure, once the vehicle's gearshifts have been automated, a fuzzy logic based throttle/brake control system has been designed such that the racing car is capable to accelerate/decelerate in a realistic manner as well as to drive at desired velocity. The steering control problem is also handled to end up with a racing car that is capable to travel on the road even in the presence of sharp curves. In this context, we have designed a fuzzy logic based positioning system that uses the knowledge of the curvature ahead to determine an appropriate position. The game performance of the developed fuzzy logic systems can be observed from https://youtu.be/qOvEz3-PzRo.