Study on Cargo Volume Forecasting and Binning Planning Based on Prophet Time Series and Taboo Search

Abstract

With the rapid development of artificial intelligence, domain-specific prediction tasks such as optimizing the storage of goods in warehouses are becoming increasingly important. In this paper, we investigate categorized warehouse planning that integrates multiple attributes of goods. To solve the daily sales forecasting problem, a Prophet model is used to predict the data for the next three months with respect to the temporal correlation and seasonal characteristics of inventory and sales, which improves the computational efficiency. For the optimization problem of inventory distribution scheme, a multi-objective planning model is established, based on the fuzzy evaluation method to evaluate the optimization cost, capacity utilization, capacity utilization and category relevance, and finally the dynamic stochastic model is used to find out the results, and a program is written by LINGO to solve the optimal solution when the objective function is maximum, which results in the “one product, one warehouse distribution scheme “. For the further optimization of the “one product, one warehouse” model, the individual priority taboo search of the optimal solution is used, and the model algorithm can flexibly deal with constraints such as warehouse capacity. The model can be extended to the prediction of energy consumption and other aspects, and the prediction accuracy can be improved by dynamically updating the data.