Study on Dual Bi-Directional Heuristic Development Framework

Abstract

The study creates a multi-start heuristic framework. The heuristic development approach accelerates heuristic convergence to high-quality optimum solutions. The Dual Bi-Directional (DBD) heuristic is a multi-start heuristic that searches for the global optimum solution (GOS) in four concurrent directions using a pair search starting from both the beginning and ending nodes. When the four optimum tours connect to form an optimum round tour of all the search space nodes, the search ends. The DBD heuristic then begins to improve the discovered optimum round tour in a unidirectional approach by employing a global search metaheuristic. By introducing the unidirectional heuristic to improve the multi-start heuristic constructed optimum round tour, the multi-start heuristic framework reduces the non-convergence of the bi-directional approach. The approach of the multi-start heuristic framework will allow the convergence of the bi-directional heuristic.