Multi objective non-dominated sorting genetic algorithm (NSGA-II) for optimizing fuzzy rule base system

Multi-objective designs are genuine models for intricate combinatorial optimization problems. This paper presents a fast elitist non-dominated sorting multi objective genetic algorithm to develop smartly tuned fuzzy logic controllers with a finer trade-off between interpretability and exactitude in linguistic fuzzy modeling problems. The multi-objective genetic algorithm produces a group of non-dominated solutions defined FLCs for multi objective problem with satisfying objective at acceptance level without dominating to any other solution. In MO-GA, an integer encoding is used to indicate the linguistic level of fuzzy rule. Here, multi-objectives are transformed to a fitness function in order to initiate the NSGA-II, i.e. selection, crossover, and mutation. The intended approach generates an efficient and reliable fuzzy logic control system through the effective searching and self-learning adaptability of the NSGA-II. The simulation results based on multi-objective exhibits better performance than single objective while controlling car like robot.