Research on cargo volume prediction of logistics sorting center based on LSTM

With the rapid development of e-commerce industry, logistics express industry is facing increasing business pressure, logistics sorting center cargo volume prediction has become a key link to improve operational efficiency and optimize resource allocation. Based on the long short-term memory network (LSTM), this paper constructs a cargo volume prediction model. Based on the daily cargo volume data of a certain 10 sorting centers in the past four months, the daily cargo volume of each sorting center in the next 30 days is predicted by the triple standard deviation method after removing outliers. The ARIMA model is used as a comparison model to make the same prediction. The results show that the highest prediction accuracy of LSTM model is 91.31%, and the average prediction accuracy is 80.6%. The highest accuracy of ARIMA model is 85.45%, and the average prediction accuracy is 65.3%. LSTM model is superior to ARIMA model in all evaluation indexes, indicating that it has higher accuracy and adaptability in processing such complex data. The research results are of great significance for improving resource utilization and optimizing personnel arrangement decision in sorting center. Future work will explore a hybrid prediction model combining the advantages of the two models to further improve the prediction accuracy.