An improved approach to resolve a combinatorial optimization problem based CoronaVirus Optimization Algorithm

Abstract

Combinatorial optimization problems refer to intractable problems that can’t be performed using exact methods. The resolution of combinatorial problems geared towards the application of heuristics, metaheuristics also matheuristics, in order to provide good enough approximations. As exact methods provide resolution corresponding to small problem scale, the approximation methods target large scale of complex problems. Metaheuristics are used to deploy intelligent methods to solve complex problems in a reasonable amount of time. The performance of a metaheuristic is improved by means of parameters adjustment as well as, hybridization within heuristics, iterative improvement methods or various metaheuristics. The cooperation of several optimization algorithms leads to improve resolution, also to overcome the limitations reported in resolving NP-hard problems. The resolution of complex problems, is thus constrained by stagnation on local optimums, as the optimization process is possibly stagnant on a specific search space region. In fact, traveling salesman problem is a combinatorial problem, that arises problematics related to the efficiency of its resolution methods. The aim of this work is to investigate on the improvement of a new bio-inspired method so-called coronavirus optimization algorithm in order to provide improved resolutions to traveling salesman problem. Various intelligent approaches are investigated and hybridized within coronavirus optimization algorithm, namely random replicate operator, elitist selective replicate operator, iterated local search, stochastic hill-climbing also improved self-organizing map. The numerical results are obtained using symmetric TSPLIB benchmarks.