Research And Application Of TSP Based On Greedy Random Adaptive Gray Wolf Optimization Algorithm

With the progress of science and technology and industry in our life,combinatorial optimization has been paid more and more attention as a set of classical and practical problems.Traveling Salesman Problem is a classical problem in combinatorial optimization,and it is a NP-Hard Problem.It is a popular research field in the fields of computer,mathematics and operations research,and TSP has a very high theoretical and application value.When the problem size increases with the city size,the time and space complexity will increase exponentially,and most algorithms cannot find a satisfactory solution.Therefore,the search for an effective algorithm to solve this problem has become the focus by many scholars.With the development of computer science and artificial intelligence,more and more methods can be used to solve TSP.Many heuristic algorithms are used to solve TSP,it's running time is relatively short,but the general result is not accurate enough.Nowadays,hybrid algorithm is used in solving TSP to improve the efficiency of solving TSP.In this paper,after analyzing various TSP algorithms,we focused on the Greedy Randomized Adaptive Search Procedures and combined it with other algorithms into a hybrid algorithm to solve TSP.GRASP algorithm is an algorithm often used to solve combinatorial optimization problems,it is a process of two iterations.In solving TSP,according to the first stage,construction stage,because of its greedy randomness,the solution is easily trapped into local optimum.For the reason to improve algorithm in solving TSP,according to the properties of the second iteration,in the first phase construct by controlling the length of the candidate list,to construct a feasible solution after random initial.Greed Wolf Optimization algorithm is an intelligent algorithm,introduced in the second stage to search a better solution.However,the initial solution of the algorithm is randomly and generated,it is largely affect the overall solution.That will be a feasible solution of phase structure as the initial solution of the algorithm.By using the strong global search characteristic of GWO and re-coding it,let it can be used to solve TSP and other discrete problems,and search the initial solution by its approximation search,so as to improve the solution that is trapped in a dilemma of falling into local optimum.For multiple data in TSPLIB,this improved algorithm is used to conduct a large-scale algorithm performance test,multiple parameter adjustments are made,and the results are analyzed from multiple aspects.In the second stage,the hybrid algorithm can improve the initial solution to search a better solution and reach the global optimum after searching,which can proves that the solution quality of the algorithm is greatly improved.Most of the examples are able to obtain the optimal solution,and the solution deviation of the large example issmall,so it can still be solved under the large example compared with other algorithms.Finally,this paper takes the routing problem of logistics distribution as an application scenario,which can be abstracted as the expansion of TSP,which is multi-traveling salesman problem.Logistics distribution requires many drivers to complete the distribution as quickly as possible in the shortest distance.Firstly,K-means is used to divide the distribution sites of each vehicle,and then the improved algorithm is used to optimize the distribution path.It is proved that the improved algorithm can be used in practical problems to raise the efficiency of the driver's distribution,shorten the distribution distance and save the time needed,so as to ensure the timely completion of tasks and reduce the losses to the company.