Optimal path estimation and tracking for an automated vehicle using GA optimized fuzzy controller

Proper path estimation and then tracking it accurately is challenging task for any Automated Vehicle(AV) on industry floor. The paper proposes an algorithm which uses computer vision along with fuzzy control to estimate and track the path for any orientation and position. The computer vision techniques are used to detect and locate the target in the neighborhood environment. The optimal path between initial and final target with specific orientation requirements of vehicle is considered to be governed by cubic curve fitting technique. The fuzzy inference based control system effectively tracks down the estimated path which eliminates iterative updates and heavy computations. Genetic algorithm is implemented to select optimal input and output membership functions. The path tracking uses Mamdani fuzzy inference model which performs smoothly with very less computational complexity. The results obtained from experiments demonstrate that the algorithm is simple to implement yet effective in target tracking in complex environment.