Research On Optimization Of Low Carbon Cold Chain Distribution Network For Fresh Agricultural Products

With the adjustment of agricultural structure and the improvement of residents' consumption level,the output and circulation of fresh agricultural products in China are increasing year by year.The whole society is concerned about the quality and safety of fresh agricultural products at an unprecedented level.The logistics and distribution network of fresh agricultural products is of great importance to enterprises.At the same time,the logistics distribution network of many enterprises is in a chaotic and uneconomical state,resulting in environmental pollution,traffic congestion,energy consumption,ineffective occupation of land resources and other negative effects.With the country's increasingly strict control of carbon emissions,it is necessary to plan the location and distribution of fresh agricultural products,and build a low-carbon cold chain logistics distribution network by using scientific methods and following the low-carbon concept.This article focuses on the problem of carbon emission reduction in the cold chain logistics network of fresh agricultural products,and studies the optimization design of the low-carbon cold chain logistics network that is conducive to the sustainable development of cities.In order to maximize the overall benefits,this paper innovatively uses the Bi-level programming to study the cold chain low-carbon logistics network planning problem of fresh agricultural products,and solves the phenomenon of benefit contradictions in the location and route planning in the logistics network.Taking the low-carbon cold chain logistics network of fresh agricultural products as the research object,the main research contents and results of this paper are as follows:(1)The bi-level programming model of single-vehicle low-carbon cold chain logistics distribution network in static environment is constructed.The upper and lower models all consider the fixed cost,refrigeration cost,transportation cost,and the fresh agricultural product damage cost in the logistics network.The upper model also considers the operation cost and inventory cost of the distribution center,and proposes an alternative distribution center selection method based on hybrid clustering algorithm.At the same time,the negative impact of cold chain distribution network on the environment is quantified and taken into account as the external cost to explore the balance between economy and environment.(2)The bi-level programming model of single-vehicle low-carbon cold chain logistics distribution network considering real-time road conditions is constructed.By using Arc GIS and traffic information data,the time-varying traffic network of Guangzhou is established,so as to refine the vehicle speed and travel time and obtain relatively accurate vehicle fuel consumption.On this basis,comprehensively considering enterprise development,social environment and consumer demand,a low-carbon cold chain distribution model with the dual objectives of maximizing customer satisfaction and minimizing total cost is constructed to maximize the overall benefits.(3)Considering that the country is vigorously promoting new energy cold chain logistics vehicles,on the basis of the previous model,this paper constructs a bi-level programming model of multi-vehicle low-carbon cold chain logistics distribution network considering real-time road conditions,and explores the optimal combination of cold chain vehicles with different fuel consumption and new energy vehicle.Based on the comprehensive consideration of the characteristics of fresh agricultural products,demands of energy conservation and emission reduction,economic environment,infrastructure,customer satisfaction and other factors,this paper scientifically selects carbon emission factors to quantify the carbon emissions generated by the logistics network,and constructs three bi-level programming models of low-carbon cold chain distribution network of fresh agricultural products,which are progressive layer by layer.Their effectiveness was proved by case analysis respectively.Aiming at the three models proposed in this paper,from simple to complex,this paper respectively designs hybrid Genetic Algorithm,Genetic Algorithm-Simulated Annealing Algorithm and improved Genetic Algorithm-Simulated Annealing Algorithm to solve the model.The operators of genetic algorithm are improved,the adaptive crossover and mutation process are implemented,and a new detemperature function is designed.Finally,the advantages of the improved algorithm are proved by comparative analysis of cases.