Research Of Vehicle Routing Problem In Logistics Based On Particle Swarm Optimization And Ant Colony Algorithm

Under the context of economic globalization and information, logistics as "the third profit source" has become an important factor in stimulating GDP growth and the development of third industry in our country. The distribution is an important part which derives from the logistics system. To optimize the distribution route can improve the efficiency of distribution, reduce distribution costs, but also can effectively reduce traffic congestion, air and noise pollution caused by logistics distribution. Therefore, it is a great practical significance how to effectively use the existing resources to establish a reasonable distribution scheme to improve the economic efficiency of enterprises.Vehicle routing problem (VRP) is the key issue in the optimization of logistics distribution. In recent years, VRP has become the focus of operations research, applied mathematics, graph theory, computer science and other disciplines. Since it is a complex combinatorial optimization problem, many intelligent heuristic optimization algorithms are developed to solve it. Particle swarm optimization and ant colony optimization are intelligent optimization algorithm which are inspired by the foraging behavior of birds and ants in nature. With studying the advantages of particle swarm optimization and ant colony algorithm, this paper proposed a fusion algorithm (PSO-MMA) which can effectively combine the advantages of two algorithms. It is showed PSO-MMA is an excellent algorithm in the precision of the search solution by applying this algorithm to the traveling salesman problem (TSP). Furthermore, the algorithm is applied to the VRP problem and verified its effectiveness.In this paper, the research content and the completed work are as follows:1 This paper proposed adaptive mutation strategy and divided multi-particle swarm to improve the deficiency that the particles are easily fall into local optimal solution.2 Because the pheromone of ant colony algorithm is evenly distributed in the initial search, it has the shortcoming of searching blindness. This paper used the suboptimal solution of the optimized PSO to initial the pheromone of the max-min ant system. Then the paper used the ACO and took the standard cases of the international library of TSPLIB as experimental data to search the optimal solutions. TSP is a classic combinatorial optimization problem; because of its NP-complete it has become a standard which can measure the merits of algorithm. Results show that the accuracy of searching solution in PSO-MMA is significantly better than the particle swarm optimization (PSO), max min ant system (MMAS) and the improved particle swarm optimization (IPSO). PSO-MMA searched the optimal solution of ei151 and berlin52 and indicated the superiority of the algorithm proposed in this paper.3 The paper established the mathematical model of vehicle routing problem and used the instances of the VRP database and other references to test the fusion algorithm. The simulation results verify the effectiveness of the fusion algorithm in solving the vehicle routing problem with capacity constraints.