Research On Evolutionary Algorithms For Clustering Orientation In Logistics Distribution Path Optimization

With the rapid development of economic globalization, the logistics industry’s status androle in the economic activity is increasingly obvious. This paper describes the theory of logisticssystem in detail, focusing on two hot issues, the decision model of the logics center location anddelivery routes. Based on the traditional logistical decision-making model, using the systemanalysis, the paper studied the combinatorial optimization problem of the distribution centerlocation problem and vehicle routing problem.In order to solve the Location-Routing Problem (LRP), various algorithms have beenproposed, evolutionary algorithm is a simple and effective intelligent algorithm. Aiming at theshortcomings of the traditional evolutionary algorithm for LRP problem (such as the too longbinary encoded string, algorithm premature convergence), the paper carried out a series ofimprovements, through simulation analysis of simulated data, its feasibility and effectiveness havebeen proved. The main job of paper is as follows:First, the traditional models of LRP ware studied in detail, giving the mathematical modeland its calculation method of the location and vehicle routing problem, and based on that, anintegration model was established, we designed an effective algorithm in two stages.Secondly, the shortcomings of the traditional evolutionary algorithm for LRP problem wareimproved. Using natural number coding scheme combined with self-adaptive crossover andmutation makes the search capacity of the solution space enhance, while bringing the punishfunction in constraint processing, and taking appropriate measures against illegal individual,reduce the likelihood of the premature convergence.In the end, the paper summarizes the results of the research, and points the shortcomings ofthe model algorithms, then combines with the complexities of LRP in real world, forecasts thedirection for research on this subject.