A mapping technique for better solution exploration: NSGA-II adaptation

A mapping method (MaM) for a better solution space exploration adapted to NSGA-II method is presented. The Mapping technique divides the solution space into several zones using a Hamming distance to a reference solution. We present a bijective mapping function from the search space to the binary representation space of solutions. For each zone, a mapping metric is used to evaluate the solution space exploration. According to this evaluation, a local search is performed. The mapping is adapted to the well known non-dominated sorting genetic algorithm-II (NSGA-II) method applied to solve the flexible job shop problem (FJSP) case. We present the comparison between the hybridization using the local search for the non-dominated solutions and the hybridization using the mapping metrics. The multi-objective metrics show the efficiency of mapping adaptation in terms of convergence and diversity.