With the rapid development of China’s fresh market and logistics enterprises,people have also put forward new requirements for fresh food distribution.In the face of market opportunities,a multi-depot fresh food distribution model is established to meet customer distribution needs.The perishable and timely distribution of fresh products has been the difficulty faced by logistics enterprises.Reducing the delivery time and ensuring the freshness of fresh products can meet the nutritional value needs of people.At the same time,reasonable planning of multi-depot paths is also the focus of logistics enterprises.In view of the perishable characteristics of fresh products and the high requirements for the timeliness of distribution,a multi-depot fresh food distribution vehicle path planning model is established.The K-means clustering is used to divide the customer group in the depot example,and then the genetic algorithm combined with variable neighborhood and simulated annealing algorithm(VND-SAGA)is adopted.After continuous iteration,the optimized vehicle distribution route is obtained.Compared with the traditional adaptive genetic algorithm,the cost is reduced,and the total cost is reduced.The specific research contents are as follows :(1)The multi-depot path planning problem(MDVRP)and the fresh cold chain distribution problem are considered at the same time,and the optimization research on the fresh path of multi-depot is carried out.The existing research results and theories of vehicle routing planning are analyzed and summarized,and then the mathematical model of multi-depot fresh food distribution is established.A multi-objective function with the minimum refrigeration cost,fuel cost,loss cost,time penalty cost,fixed transportation cost of products and fixed vehicle cost is constructed,and the time window constraints of customers and vehicle load constraints are considered to ensure the freshness of fresh products and improve customer satisfaction.(2)An improved algorithm based on traditional genetic algorithm is proposed.Firstly,the convergence speed is improved dynamically by combining adaptive crossover and mutation operators.Then,the simulated annealing algorithm(SA)is introduced to follow the Metropolis criterion to avoid falling into local optimum and generate new solutions,which make the individual population better.Finally,the variable neighborhood descent search algorithm(VND)is used to generate new solutions through the descent search of different neighborhood structures to improve search accuracy.(3)In different cases,the genetic annealing algorithm combined with the variable neighborhood(VND-SAGA)and adaptive genetic algorithm are used to solve different cases.The VND-SAGA algorithm reduces the distribution costs by 9.82 %,13.8 % and 8.7 % respectively on the optimization results of the adaptive genetic algorithm.The comparison and analysis of the optimization results of the two algorithms under different depots prove that the genetic annealing algorithm combined with the variable neighborhood has a better optimal solution,thus verifying the advantages of the improved algorithm.In the environment of the new challenges posed by the development of the modern fresh market to logistics distribution,this paper studies the multi-depot fresh food distribution problem builds a multi-depot fresh food distribution model and designs an algorithm.Through different cases,the relevant research methods and models were proved to be effective in improving the efficiency of logistics distribution and lowering costs,and meeting the distribution requirements of fresh products in multiple depots. |