Research On Vehicle Routing Problem (VRP) Optimization In Cold Chain Logistics Based On Partheno-genetic Algorithm

Driven by the Internet, the demand for cold-chain logistics involved in agricultural products is growing, and the cost has been the key factor restricting the development of cold chain logistics. Thus, how to reduce the cost of cold-chain logistics effectively is a common concern of nation, business and consumer. At the same time, as the expansion of cold chain logistics market and consumer demands for service quality improvement, companies have to improve their service quality and customer satisfaction to increase customer stickiness and have a place in a highly competitive market. Therefore, the study aims to make reasonable vehicle scheduling and route optimization, lowest overall cost while ensuring a high level of satisfaction, and support the daily scheduled work of cold chain logistics enterprises. This paper broke the limitations of the traditional model which only considered transportation costs and a single distribution center to optimize the cold chain logistics route, overcome the insufficient of traditional genetic algorithm for solving vehicle routing problem, and proposed a function of time and quality factors based satisfaction function to meet the reality of business demand, which has an important theoretical and practical significance.Firstly, the paper have a review of the research at home and abroad in cold chain logistics route optimization, and pointed out the shortcomings of current research, and expounds the basic theory of cold chain logistics issues related to VRP as a solid theoretical foundation for the following research papers.Secondly, starting from the basic VRP model, combined with the characteristics of cold chain logistics, in consideration of cost optimization including damage costs and energy costs, the cold chain VRP model with satisfaction constraint and the cold chain VRP model with multi depots were constructed respectively.Thirdly, in order to avoid damage to chromosomes and "premature convergence" problem when solving vehicle routing problem with traditional genetic algorithms, the paper proposede a partial elite selection strategy, designed three kinds of single parent chromosome gene mutation operator and crossover operator, which include a merge operation to optimize number of vehicles. Adaptive genetic algorithm was designed as a comparison, and both were realized by MATLAB.Finally, taking the S food company data as an example, the validity and superiority of the two models and the single parent genetic algorithm are verified.