A Demand Forecasting System for Automobile Spare Parts Based on Multi-model Optimization

Abstract

The embodiment of the invention discloses a demand forecasting system for automobile spare parts based on multi-model optimization, which relates to the technical field of data forecasting. Wherein, the system includes: a model training module; a threshold determination module, which is used to finally average the average error evaluation index value of all automobile spare parts under each threshold to be selected; determine the final threshold according to the variation of the average error evaluation index value; model The optimal module is used to determine the prediction model corresponding to the minimum error evaluation index value to predict the demand if the difference corresponding to the same kind of auto spare parts is greater than or equal to the final threshold value, and obtain the predicted value; if the corresponding same kind of auto spare parts If the difference is smaller than the final threshold, at least two forecasting models for predicting the demand for spare parts of the same type of automobile are determined to predict the demand, and an average forecast value is obtained. In the embodiments of the present invention, accurate prediction results are obtained by scientifically and rationally selecting the optimal model or combination among multiple prediction models.

Description

technical field [0001] The embodiments of the present invention relate to an electrical digital processing technology, and in particular, to a demand prediction system for automobile spare parts based on multi-model selection. Background technique [0002] Demand forecast for auto spare parts, that is, an estimate of the future demand for auto spare parts. By predicting customer demand for spare parts in a certain period of time in the future, car companies can purchase raw materials in advance, arrange production activities, and set reasonable inventory levels. Auto spare parts are frequently upgraded, and there are tens of thousands of spare parts units. The supply chain faces severe cost and inventory challenges. The accuracy of spare parts demand forecasts directly affects the supply chain costs and customer satisfaction of auto companies. [0003] The demand forecast of auto spare parts is generally based on the past historical data, using various mathematical formulas...

AI Technical Summary

Problems solved by technology

Existing methods generally choose a mature model for prediction, which makes the prediction result heavily dependent on the accuracy of the model, thus affecting the accuracy of the prediction

Images