Greedy Permuting Method for Genetic Algorithm on Traveling Salesman Problem

Abstract

Genetic algorithm (GA) is a very efficient algorithm for solving optimization problems such as traveling salesman problem (TSP), timetable scheduling and 0/1 knapsack problem. Among these problems, the TSP is a well-known combinatorial optimization problem and it is a NP-hard problem. Many literatures have discussed how to use different methods to make genetic algorithm more efficient on solving TSP and these methods include how to design the representation of solution, how to initialize the initial population, how to design crossover, mutation and selection operator and so on. In this paper, we present a new method to initialize an initial population for GA on TSP, which we call greedy permuting method (or GPM for short). We test GPM on some TSP benchmark problems and find that it is competitive with an already proposed method NF and it can make GA much more efficient than a randomly initializing method.