Combining ARIMA Model And Bp Neural Network To Forecast Coer Volume

The main purpose of creating SPDB Management System of production and operation is to fully reflect the condition of production operation.The core volume is an important index in the system.Predicting core volume can help responsible person to deploy tasks and reduce risk.Predicting core volume is difficult,because the time series of core volume is influenced by many elements.So it has complex characters,such as linear characters and non-linear characters. In the first place,this paper studies the ARIMA model which is adept at fitting linear sequence,researches the method of the model identification and improves the extended ability of it, and uses Eviews to carry on simulation. Then this paper continues to study the BP neural network which is adept at fitting nonlinear sequence and explores the most suitable network structure and chooses the better learning function and training function. Then this paper uses the MATLAB to realize the prediction of the time series.Although single model has perfect ability to predict,it has some imperfection.Due to overcome the defects,this paper puts forward the forecasting concept which ARIMA combined with BP neural network, then deeply studies on the prediction principle and improves the combination model to achieve the goal of improving prediction performance.Finally, this paper makes a comparison of all models to choose the best model for predicting the core volume.