Routing Optimization Of Cold Chain Logistics With Fuzzy Time Window

With the development of economy, the demand for the fresh products become more and more, people care more about the quality than quantity. So the storage and preservation technology of fresh products has been developed. Thus also lead the rapid development of the cold chain logistics. Because the fresh products are often some perishable goods, so the requirement of timeliness is strong. In the distribution process, the requirement for both temperature and time is high. Therefore, for the fresh food cold chain distribution, considering the time window of customer requirements and good advance planning of distribution route, not only can reduce the enterprise distribution costs can also ensure the quality of fresh goods, and also improve the customer satisfaction.In the study of the cold chain distribution problem, the research on the customer time window is few, and most of the research is about the hard time window and soft time window, the study about the fuzzy time window is more scarce. But in reality, the fuzzy time window is the most reasonable. Therefore this paper studies the vehicle routing optimization problem of cold chain logistics with fuzzy time window, which to minimize the logistics cost and maximize customer satisfaction, to constitute the double objective function of optimization model.In this paper, the fuzzy membership function is used to deal with the fuzzy time window. Fuzzy process the customers’ expectation time window and maximum tolerable time window. Convert it to customer satisfaction level of goods delivery time. With the customer to buy the amount of goods accounted for the proportion of the distribution plan of individual customer satisfaction level of weighted average level of the overall satisfaction of the customers of the service time, established the fuzzy multi-objective optimization model of time window of cold chain logistics vehicle routing. In order to handle the multi-objective function, the overall customer satisfaction level is set for a fixed value, will become as constraint conditions. Through a two-stage approach to solve the single objective function and indirectly to achieve the purpose of double target so as to avoid the second goal has a direct impact on the primary target. Then, aiming at the characteristics of the model and the practical problems in this paper, a genetic algorithm which is more suitable for this model is designed, which is called the improved genetic algorithm. Finally, through the Python language programming algorithm and analyze the example of the data to solve the optimal distribution path for the S enterprise.