An OSGI-MPH Algorithm for Solving Combinatorial Optimization Problem

Abstract

Travelling Salesman Problem(TSP) is a classical combinatorial optimization problem. There are numerous algorithms to solve it in the field, like Genetic Algorithm, Any Colony Algorithm, Simulated Annealing Algorithm. However, these algorithms usually have the weak global seach ability, premature converging capability, weak parallel ability and poor accuracy. Particularly, when facing large-scale combinatorial optimization problems, the algorithms above often require a great time cost. In this paper, we propose OSGI-MPH(Multi Parent Hybrid Algorithm based on Optimal Selection and Gene Insertion) by integrating a variety of heuristic seaerch algorithms and intelligent optimization algorithms. The algorithms and intelligent optimization algorithms. The algorithms has excellent global search ability and convergence ability. We test the performance of the algorithm in large-scale TSPLIB instances, and the results show that the algorithm has excellent performance compared with the traditional heuristic algorithm and intelligent optimization algorithm.