Research On Vehicle Routing Optimization Algorithm With Hard Time Windows

In the past two decades,vehicle routing problem has been the focus of research in the fields of mathematics application,computer application,transportation and management science.Among them,the vehicle routing problem with hard time windows is proved to be NP-hard and difficult to be solved by traditional methods,which is usually solved by heuristic algorithm.However,the heuristic algorithm has some defects such as slow convergence speed and easy to fall into local optimality.In addition,the multiple objectives of the vehicle routing problem with a hard time windows often contradict each other,and the balance between the objectives needs to be considered.Therefore,it is of great significance to study the multi-objective vehicle routing problem with hard time windows.In this paper,a hybrid variable neighborhood tabu search algorithm for solving binocular objects is proposed by combining neighborhood search strategy and intelligent algorithm.Meanwhile,a hybrid neighborhood evolutionary algorithm for solving multi-objective is proposed to further meet the practical needs,and good results are obtained.The details are as follows:1)Aiming at the vehicle routing problem with hard time windows,a bi-objective nonlinear optimization model was established to minimize the number of vehicles and the total driving distance,and a hybrid variable neighborhood tabu search algorithm was proposed.On the one hand,an improved saving algorithm is used to generate the initial solution,and three deletion operators and an insertion operator are designed to optimize the initial solution,which provides excellent initial solutions for subsequent tabu search.On the other hand,tabu iterative search is carried out based on four kinds of neighborhood construction operators,and the flexible storage structure of tabu search,tabu criterion of avoiding roundabout search and amnesty criterion of enhancing diversity search are used to get rid of the local optimal solution effectively and finally achieve global optimization.The experimental results on the Solomon and Homberger datasets show that the algorithm is superior to the two similar search algorithms in the literature,and has good convergence and stability.2)Based on the two-objective nonlinear optimization model,adding the goal of minimizing the longest path distance,a hybrid neighborhood evolution algorithm is proposed.Evolutionary algorithm is used for global search and neighborhood search algorithm for local optimization.In each iteration,the hybrid neighborhood evolution algorithm first generates a set of populations by using the evolutionary algorithm,and then optimizes the individuals in the population by using local search operators based on specific objective functions to obtain a set of local optimal solutions.Then,the global optimal solution and local optimal solution are compared by fast non-dominated sorting and congestion calculation,and a group of optimal solutions is selected as the next generation initial population.At the same time,speed is introduced and Solomon data set is adjusted.Experimental results show that the convergence and distribution of the proposed method are better than that of the fast non-dominated sorting genetic algorithm with elite solution.