Multi-Hybrid Algorithm for Searching Solutions to Optimization Problems: Case of Traveling Salesman Problem (TSP)

EasyChair Preprint 4491

Abstract

 In this paper, we are interested in finding solutions to the travelling salesman problem(
TSP) by proposing a multi-hybrid algorithm between a multi-colony ant colony algorithm (OCF)
and the Optimization Towers heuristic (LKH). The proposed approach combines both insertional hybridization
which is a new and very efficient approach and parallel hybridization. The goal of this
hybridization is to reduce the computation time as well as the improvement of the optimal solution.
The two algorithms move in parallel synchronously while communicating with each other. To assess the
effectiveness of the proposed metaheuristic, we examined its performance on instances of the TSPLIB
online instance library for the TSP. The experimental results indicate that the hybrid heuristic that we
call (MOCFLKH - TSP) competes with the other existing methods in the literature by being illustrated
by a great improvement in the minimization of computation time and computation costs.

Keyphrases: Travelling Salesman Problem, ants, colony, heuristic, multi-colony

@booklet{EasyChair:4491,
  author    = {Nguimeya Tsofack Baudoin},
  title     = {Multi-Hybrid Algorithm for Searching Solutions to Optimization Problems: Case of Traveling Salesman Problem (TSP)},
  howpublished = {EasyChair Preprint 4491},
  year      = {EasyChair, 2020}}