Research On Vehicle Routing Problem And Its Intelligent Algorithm

With the rapid expansion of economic globalization and the information technology industry soared, as an emerging logistics services, broad prospects and value-added features for all to see, there is a leap of progress worldwide. In many aspects of logistics, reduce transportation costs, improve transport efficiency will help accelerate the development of the logistics industry. Therefore, as the core issue of the transport vehicle routing problem has been plenty of research, and made a wealth of research results. It is through the organization, optimizing cargo transport routes, to meet certain constraints under the premise of the lowest transportation costs, the shortest distance transport, the minimum transit time as the target, the goods reach the destination.This article is based on a mathematical model of vehicle routing problem, researchers using a variety of rich before combining algorithms designed to improve the PSO algorithm is applied to vehicle routing problem, and at the same time with a genetic algorithm task delivery and pick up random vehicle path problem. Work done as follows:(1) A brief description of the research process and current situation of vehicle routing problem, and research significance, introduces the mathematical model of the vehicle routing problem, the exact algorithm, heuristic algorithms and intelligent optimization algorithms.(2) An overview of the genetic algorithm, the basic idea of??the particle swarm optimization algorithm step processes and applications.(3) The main particle swarm algorithm capable constrained vehicle routing problem, due to the particle velocity previous affect the speed of the subject, thereby affecting the algorithm search capability, this paper particle swarm optimization inertia weight selection for using linear and non-linear combination of values??of the way. Experimental results show that the improved success rate improved searching optimal solutions, the global search ability to make progress, and accuracy is also improved. For large-scale vehicle routing problem, particle swarm algorithm and genetic algorithm combination of programs, genetic algorithm-specific crossover operator thought. Experimental results show that the improved, better to avoid the premature convergence, while improving the convergence speed and accuracy.(4) For random vehicle routing problem at the same time delivery and pick up the task, due to the uncertainty of information, taking into account not dependent on the use of genetic algorithms to solve specific problems, combined with the problem using natural number coding method, and algorithm Select the operator to improve and ensure the best individual preservation, and the base case were tested to obtain good results.