Optimization Solution Of Vehicle Routine Problem And Its Algorithm Design

With the development of the social economy, the importance of the physical distribution is increasingly visible. Transportation is a link that is directly connecting consumers, so how to arrange transportation routine has become a important study subject about cutting costs and increasing benefits.Author of the article studies VRPHTW and VRPSTW firstly, and gives the aim function of the two kinds of questions. Author also points that VRPSTW is more in line with the actual situation and more general than VRPHTW. Hence, Author focuses on analyzing and improving model of the VRPSTW and applies the new model into actual question better.Physical distribution transportation belongs to NP question in optimize question, fruits will be produced by solving the question with optimize algorithm. However, these algorithms all have their own defects. To some extent, the defects constrain the application of the algorithm. After studying Taboo Search Algorithm, Genetic Algorithm, and And Immune Algorithm adequately, author brings up two kinds of new mixed algorithm which is used to solve vehicle distribution problem.Algorithm 1: Solving the small-scale nodes problem by the characteristic of Hopfield network having fast convergence speed, we use Taboo Search Algorithm and Immune Algorithm to construct the new mixed algorithm because the network couldn't get the global optimize solution; algorithm 2: to get routine solution because of the characteristic of the global search of Genetic Algorithm. Author brings up the theory of parallel computation to solve the problem of the slow research speed of Genetic Algorithm , and adds the diversity of population by modifying genetic operation to take control of early mature.Finally, the result of the algorithm that is brought up by me is testified by test data in the thesis, which shows that the two algorithms could cost the price of vehicle transportation and improve the economy revenues of the cooperation. In fact, the two algorithms both have the high practice value.