Cargo quantity prediction method and system for Spring Festival period

Abstract

The invention discloses a cargo quantity prediction method and system for a spring festival, and the method comprises the steps: selecting historical cargo quantity data belonging to the same day as aprediction day in the last week, inputting the historical cargo quantity data into a trained LSTM model, and obtaining an initial cargo quantity prediction value of the prediction day; obtaining theactual cargo quantity of the day ahead of the prediction day and the daily average actual cargo quantity of multiple consecutive weeks adjacent to the day ahead of the prediction day, and comparing the cargo quantity change ratio of the two; judging whether the cargo quantity change ratio is greater than a prediction threshold; and if so, correcting the initial cargo quantity predicted value according to the cargo quantity proportional change coefficient in the Spring Festival period to obtain a final cargo quantity predicted value. When the cargo quantity change ratio between the actual cargoquantity of the day-ahead day and the daily average actual cargo quantity of the adjacent continuous multiple weeks exceeds the prediction threshold value, the initial cargo quantity prediction valueobtained by the LSTM model is corrected according to the cargo quantity proportion change coefficient in the spring festival, so that the prediction accuracy is improved.

Description

technical field [0001] The present invention relates to the technical field of cargo volume forecasting, in particular to a cargo volume forecasting method and system during the Spring Festival. Background technique [0002] In the logistics industry, in order to carry out cargo loading and transportation efficiently, it is often necessary to arrange the deployment of vehicles in advance. Therefore, it is necessary to carry out cargo volume forecasting. In order to ensure the reasonable deployment of logistics vehicles, the accuracy of cargo volume forecasting is very important. appear more and more important. Most of the existing logistics companies use the combination of experienced staff and general forecasting models to forecast the volume of goods, and can obtain a better forecasting result on normal working days. However, when it comes to holidays, the volume of goods before and after the holidays will change due to the appearance of the festival, especially during th...

AI Technical Summary

Problems solved by technology

[0003] The invention provides a method and system for forecasting cargo volume during the Spring Festival to solve t...

Images