An Improved TS for the Open Vehicle Routing Problem with Soft Time Windows

In this paper, a problem is described and mathematical formulation is established on the basis of open vehicle routing problem with soft time windows. Meanwhile, an improved tabu search algorithm (TS) is proposed to optimize the OVRPSTW. in improved TS static, the initial solutions are obtained at random and then current one is selected from better solutions. Two paths are randomly selected on 2-opt to improve solution, ensuring the optimization between and inside the paths. Tabu list´s length is expressed by linear Representation to conduct a dynamic tabu list. Thus, the length and the struction of tabu list change with search process diversifying the search path, to expand the scope of neighborhood search, and to improve global optimization. Simulation experiment proved the feasibility and effectiveness of tabu search algorithm.