Application of Genetic Algorithm in Emergency Logistics Optimization in the 5G Network Environment

E-commerce logistics networks are affected by unexpected events, which can disrupt the delivery of packages along various routes. In order to predict the package quantities at each logistics site and route, and to develop reasonable transportation arrangements and sorting plans, this paper first extracts data from each route. Gibbs sampling simulation is then applied to fill in the missing data, and the approximate trends of package volumes for three routes are observed. An ARIMA model is established to forecast the volume of goods for these routes. Utilizing the previous step's model, the freight volume from DC5 is predicted as the originating and receiving station. By introducing constraints and defining the objective function, a single-objective optimization model is used to solve for the freight volume allocation of DC5 as the sending and receiving station, respectively. Next, a multi-objective optimization method is employed to construct a multi-objective dynamic decision-making model for route planning. The simulated annealing algorithm is used to solve the model, and it is determined that opening a new route from DC3 to DC1 is feasible, and the load capacities of all routes are normal. Finally, an evaluation of the importance of different logistics sites and routes is conducted. A comprehensive evaluation model based on entropy-weighted Topsis is constructed to obtain indicator weights. Based on positive and negative solutions, the importance rankings of routes and sites are calculated. In the route evaluation, DC25-DC10 ranks first, while in the site evaluation, DC10 ranks first. Finally, the robustness analysis of the model is conducted.