An efficient hybrid genetic algorithm for solving truncated travelling salesman problem

Abstract

This paper considers a practical truncated traveling salesman problem (TTSP), in which the salesman is only required to cover a subset of out of given cities (rather than covering all the given cities as in conventional travelling salesman problem (TSP)) with minimal traversal distance. Thus, every feasible solution tour contains exactly cities including the starting city. However, extensive research on TSP has been received and various efficient solution techniques including exact, heuristic, and metaheuristic algorithms are devoted, a very limited attention has been given to TTSP models because of its solution structure. The TTSP model comprises two types of problems including city selection i.e. as a salesman's trip need not include all the cities, the challenge is to identify which combination of cities are to be visited and which sequence of cities will constitute minimal traversal distance. A hybrid genetic algorithm (GA) comprising sophisticated mutation operators is developed to tackle this problem efficiently. Comparative computational findings suggest that the proposed GA has capability to outperform existing approaches in terms of TTSP results. In addition, the proposed GA report improved results and will serve as a basis for forthcoming TTSP studies.