Integrated Model and Algorithm Research on Cargo Volume Forecasting and Personnel Scheduling Based on SARIMA and Genetic Algorithm (GA)

Abstract

In the context of the swift growth of the logistics industry, the accurate prediction of cargo volumes and the rational arrangement of personnel scheduling have become essential for logistics companies to enhance operational efficiency and curtail costs. This study introduces a comprehensive model to address these challenges effectively. For cargo volume forecasting, the study developed an ARIMA-based time series model and further proposed a SARIMA model to account for seasonal variations. The SARIMA model, which incorporates seasonal differencing and seasonal autoregressive moving average terms, demonstrated significant improvements in prediction accuracy, especially for data with distinct seasonal patterns. In terms of personnel scheduling, the study formulated an optimization model utilizing the genetic algorithm (GA). This model aims to minimize the total number of man-days and balance the actual hourly man-efficiency per day by simulating natural selection and genetic mechanisms. The results from the study indicated that the GA-based model could substantially reduce scheduling costs while meeting service requirements. The integration of the SARIMA model for forecasting and the GA model for scheduling provides a robust solution for cargo volume prediction and staff scheduling in logistics sorting centers, offering a significant contribution to operational optimization.