Research of the Application of Cargo Volume Forecasting and Simulated Annealing Scheduling Optimization Based on Multi-model Integration in Logistics Centers

SARIMA model and the XGBoost model are both classical time series forecasting tools, while the simulated annealing algorithm (SA) is an excellent multi-objective planning method. By combining these algorithms, this study aims to forecast the cargo volume of a logistics center and arrange the scheduling of the logistics center accordingly. First, the SARIMA forecasting, and XGBoost hybrid correction model can more accurately predict the cargo volume of this logistics center for the next 30 days. Subsequently, the study will analyze and determine the decision variables and constraints in scheduling planning. Finally, the planning problem will be solved by SA, which enables the minimum number of personnel to meet cargo volume requirements, scheduling constraints, and other constraints. The results of this study have significant practical value, providing a replicable forecasting and planning framework for the dynamically changing market environment, which effectively helps enterprises to anticipate market changes and optimize resource allocation to enhance market competitiveness.