Research On Vehicle Route Planning Of Urban Express Distribution

With the acceleration of life rhythm,people are more and more inclined to fast and convenient access to services.In order to enhance customer service experience,express delivery enterprises began to provide personalized service of distribution door-to-door.The distribution time limit gradually changed from day-to-day calculation to minute level,and the distribution delivery scene gradually extended from outdoor to complex indoor environment.The existing distribution path planning model only considers the outdoor part of distribution,but ignores the time factor of indoor distribution,which leads to the unreasonable distribution path planning,delays distribution,decreases customer satisfaction,and increases the pressure of distribution personnel.Based on the integrated indoor and outdoor distribution scenario,this paper constructs an indoor and outdoor collaborative path planning model,which improves the rationality and effectiveness of distribution path planning.The model consists of two parts:indoor distribution point location and collaborative path planning.Aiming at the low accuracy of K-means algorithm for indoor fingerprint location and the premature problem of existing path planning algorithms,theoretical innovative research and simulation are carried out.The main research work is summarized as follows:1?Based on the urban indoor and outdoor integrated distribution scenario,this paper establishes an indoor and outdoor collaborative path planning model,which adds the acquired indoor distribution time data to the time constraints of the model to improve the time rationality of the path planning.According to the time window function of distribution services,the customer satisfaction function is established,and the distribution cost and customer satisfaction are taken as the optimization objectives of the model.Standard path planning model can reduce costs and improve delivery punctuality and customer satisfaction.The validity of the model is verified by simulation experiments.2?Aiming at the problem of low positioning accuracy of existing indoor fingerprint location K-means algorithm,an improved a positioning algorithm is proposed.Sample density function is added on the basis of mean-square deviation criterion function,and the actual class number of fingerprint database is obtained by using extended criterion function.The redundant clustering centers are removed effectively,the problem of solitary points is improved,and the convergence speed and positioning accuracy are improved.Experiments show that the positioning accuracy of the improved K-means algorithm is 23.1%higher than that of the existing K-means algorithm,and the positioning time is shortened.3?Aiming at the premature problem of model solving genetic algorithm,a hybrid genetic-simulated annealing heuristic algorithm is proposed in this paper.This method improves the premature problem of genetic algorithm by embedding simulated annealing algorithm into genetic algorithm and searching locally on the basis of global search,and improves the hybrid algorithm from fitness function,coding and other links according to distribution model.The simulation results show that the convergence speed of the hybrid genetic-simulated annealing heuristic algorithm is improved by 25%and the optimization effect is improved by 5.76%compared with the existing algorithms.