An efficient genetic algorithm for solving open multiple travelling salesman problem with load balancing constraint

• Published in: Decision Science Letters Vol. 10; no. 4; pp. 525 - 534
• Main Authors: Singamsetty, Purusotham, Thenepalle, Jayanth Kumar
• Format: Journal Article
• Language: English
• Published: Growing Science 01.10.2021
• ISSN: 1929-5804; 1929-5812; 1929-5812
• DOI: 10.5267/j.dsl.2021.5.003

The multiple travelling salesman problem (MTSP) is one of the widely studied combinatorial optimization problems with various theoretical and practical applications. However, most of the studies intended to deal with classical MTSP, very limited attention has been given to an open multiple travelling salesman problem and its variants. In this paper, an open multiple travelling salesman problem with load balancing constraint (OMTSPLB) is addressed. The OMTSPLB differs from the conventional MTSP, in which all the salesmen start from the central depot and need not come back to it after visiting the given number of cities by accomplishing the load balance constraint, which helps in fairly distributing the task among all salesmen. The problem aims to minimize the overall traversal distance/cost for operating open tours subject to the load balancing constraint. A zero-one integer linear programming (0-1 ILP) model and an efficient metaheuristic genetic algorithm (GA), is established for the OMTSPLB. Since no existing study on OMTSPLB, the proposed GA is tested on the relaxed version of the present model, comparative results are reported. The comparative results show that the proposed GA is competent over the existing algorithms. Furthermore, extensive experiments are carried out on OMTSPLB and the results show that proposed GA can find the global solution effectively.