A modified ant colony optimisation based approach to solve sub-tour constant travelling salesman problem

TLDR: This paper presents a modified form of the basic ACO algorithm, obtained by introducing a new operation known as restoring operation for solving the SCTSP problem, where the low-quality path/solution given by any ant is replaced by a nearest better solution.

Abstract: This paper presents an advanced ant colony optimisation approach to solve sub-tour constant travelling salesman problem (SCTSP) with time limit in fuzzy environment. In SCTSP, there is a set of sub-tours (two cities for each sub-tour). A traveller must complete his/her tour with predetermined sequence of distinct sub-tours. In this proposed model, the total time of travel must not exceed a predetermined time limit, i.e., the salesman should maintain a maximum time limit to complete his/her tour. In this paper, we present a modified form of the basic ACO algorithm, obtained by introducing a new operation known as restoring operation for solving the problem. In restoring operation, the low-quality path/solution given by any ant is replaced by a nearest better solution. The proposed problem is solved by considering fuzzy travel costs and time. Fuzzy credibility measure and graded mean integration method are used to obtain optimal decision. Computational results with different datasets are presented for illustration.