Research On Vehicle Scheduling Problem Based On Distribution Estimation Algorithm

With the increasingly fierce of market competition,the rapid development of science and technology and the continuous improvement of the professional level of logistics,a large number of enterprises have leaded logistics theory and technology into the production and management and considered the logistics as an important means of market competitiveness and the core level of competition.How to design reasonable routes for the vehicles in order to reduce the transport cost is the main research contents of the problem.Vehicle scheduling problem of logistics is a typical NP-hard problem,when the scale of the problem is large,the computing time needed for the solution will increase,so we will be difficult to get the exact solution of the problem.Therefore,most present research scholars mainly use intelligent optimization algorithm to solve the vehicle routing problem.The Estimation of Distribution Algorithm(EDA)is an evolutionary algorithm based on probability distribution model.In recent years,it has also been widely concerned and developed,and has been successfully applied in many industrial development fields.Moreover it has achieved good results.In this regard,based on different constraint conditions of vehicle routing problem,this paper improve Estimation of distribution algorithm in three ways,and the improved algorithms are simulated to verify the effectiveness.At first,aiming at the classical vehicle routing problem of capacity constraints,the appropriate coding mechanism and probability model are designed.The traditional binary code is changed to decimal coding mode,which reduces the cumbersome process of conversion between codes.According to the feature of Vehicle routing problem,the ordinary two-dimensional matrix converted into three-dimensional matrix,that is,each vehicle corresponds to a separate two-dimensional matrix.Finally joined the local search for high-quality individuals to more detailed search,and then put forward a solution of improved distribution estimation algorithm(IEDA).According to simulate the problem of CVRP,show that the proposed algorithm can improve the global search ability,effectively avoid the local optimization of the algorithm,and reduce the total distribution cost(distance),thus verifying the effectiveness of the algorithm.Secondly,for random demand Multi-vehicle VRP,the random vehicle scheduling model based on time taxes is established.The random distribution problem is transformed into a series of static problem.In addition,for the multi-vehicle model considered the load rate as the basis for the selection of vehicles,established vehicle scheduling problem that consider the integrated cost of a load rate and fuel consumption as the optimal target.On the basis of the previous algorithm,the hybrid estimation algorithm(HEDA)is proposed by mixing the distribution estimation algorithm with the parallel saving algorithm.Once more,this paper proposed an adaptive distribution algorithm(AEDA)to solve the problem of low cost vehicle scheduling.The initial probability model is improved,so that the probability model can accumulate more high quality information and the initial search range of the algorithm is more extensive.The adaptive update mechanism based on information entropy is designed to update the learning rate and mutation rate.This can enhanced the search ability of the algorithm.