Research On Judgment Model And Demand Forecasting Of Rarely Used EMU Spare Parts

In recent years,China's high-speed railway development by leaps and bounds.In particular,the implementation of China's 2016 version of the "long-term railway network planning" will further promote the development of China's high-speed railway.However,EMU spare parts have many different kinds and all kinds of spare parts have different specifications.Unreasonable classification and rugged supply management of EMU spare parts have seriously affected the economic benefits of the high-speed railway in China.Facing a series of serious challenges,how to rationally carry out classification management of EMU spare parts,and to make reasonable demand forecasting of spare parts to ensure timely supply of spare parts and to effectively reduce the inventory have become urgent problems.This paper will mainly study the method of judgment model and demand forecasting model of rarely used EMU spare parts.First of all,we analyze the current situation of the classification of EMU spare parts.Combining with spare parts management approaches of the EMU overhaul and locomotives,we find some shortcomings of the current EMU spare parts classification.In this paper,a variety of classification methods are compared and the advantages and disadvantages of each method are compared before selecting the final classification method of EMU spare parts.Finally,this paper divides the spare parts by multi-factor comprehensive clustering analysis method,which will lay the foundation for the forecasting of the demand of the spare parts.Second,the commonly used exponential smoothing method,Croston method,etc.,which are used in the forecast of spare parts demand,cannot achieve satisfactory prediction accuracy.In this paper,we first analyze the characteristics of non-commonly used spare parts and the existing forecasting methods.We divide the demand series of the rarely used EMU spare parts into two separate sequences,which are the demand time interval sequence and the spare parts demand sequence.Then,the GA-BP neural network algorithm is introduced to forecast the demand of the spare parts.Finally,it summarizes the significance of the classification and the demand forecast for the management of rarely use EMU spare parts in China.At the same time,we look forward to some aspects of the future development of spare parts management,such as the data management of spare parts,the "Internet+" mode used in the supervision of the EMU spare parts,in order to further improve the efficiency of maintenance operations.