Applying m-Mutation Operator in Genetic Algorithm to Solve Permutation Problems

Abstract

Routing problems, scheduling problems, transportation scheduling problems are interesting problems and variants of Permutation problems. The intention of solving these problems is to find a better path (solution) among enormous, feasible, available solutions. The better (or best) solution should provide a cost effective path which would enable a anyone (or device) to travel to all the given cities (location) one and only once and finally return to the starting city(location). Approaches like Brute Force would not work since it is not feasible to calculate cost for all the possible paths (because rise in the number of cities would exponentially increase the number of all possible permutation of cities). Approaches like heuristics can be trusted reasonably as it uses reduced amount of computing power. Heuristic techniques are used because it gives quick and better solution even if it does not guarantee the best solution. In this research work, the authors propose a mutation operator called (m - mutation) that could be applied in genetic algorithm to solve permutation problems. The efficiency of the proposed mutation operator is compared with the efficiency of existing mutation operators in solving the same permutation problem and the results are encouraging.