Optimization Research On Inventory Management For Preventive Maintenance Spare Parts Of EMU

Under the background of technical innovation and development of China’s EMU,with more and more service vehicles,such as Hexie to Fuxing,the requirement for the matching maintenance mechanism of EMU becomes higher and higher.The reasonable supply of spare parts for EMU is the precondition for the efficient realization of the operation unit’s maintenance plan,but the cost of holding spare parts,Storage and order costs will occupy a large amount of liquidity for EMU operating units.Railway administration expect to achieve scheduled overhaul operations for EMUs at the lowest spare parts inventory management cost.The inventory management of current EMU overhaul process,process overhaul system and overhaul materials in China is not clear enough for the reasons of classification of spare parts,unclear demand and too general inventory strategy.At the same time,there are problems of insufficient spare parts in stock and redundancy of spare parts,which can not fully meet the realization of EMU overhaul plan.Therefore,aiming at the above problems,this paper carries out research on the classification method of spare parts,the forecasting method of spare parts demand and the inventory control strategy of EMU,the main contents are as follows::A spare parts forecasting model based on GA-LSTM is established.Taking historical spare parts consumption data of Road Bureau as an example,the historical data is divided into training data and test data in proportion.At the same time,the validity of the model is proved by comparing the test data with the predicted data of GA-LSTM model,which provides support for spare parts classification and inventory management.According to the characteristics of EMU spare parts and RAMS standard,the classification index system of EMU spare parts is established from five dimensions,such as reliability,economy,safety,availability and maintainability.According to IGAHP method,the index weight coefficient is obtained,and 11 classification indexes for quantifiable evaluation are designed.Considering the classification principle of DEA model indexes,11 indexes are divided into input and output indexes.Then use DEA dual model to calculate the relative efficiency of spare parts,i.e.relative importance.Finally,use ABC classification principle to classify spare parts into three categories: core spare parts,important spare parts and common spare parts.Design IGAHP-DEA-ABC classification method to facilitate spare parts inventory of EMU.According to the characteristics of spare parts of different types of EMU,due to the lack of historical data,the spare parts which can not reach the expectation with GA-LSTM prediction method are determined according to their life distribution type,and based on(S-1,S),(s,Q)spare parts ordering strategy,the ordering lead-time,order batch and inventory period of classified spare parts are determined.Then corresponding inventory control methods are established for various spare parts.This paper applies forecasting model of long and short memory neural network to spare parts demand forecasting of EMU,improves ABC classification method,and formulates corresponding inventory management strategies for different kinds of spare parts according to the characteristics of EMU overhaul process and overhaul schedule of China Railway Administration,which provides certain theoretical reference for material management of EMU operating units..