Research And System Implementation Of Inventory Forecasting Method In Auto Parts Storage Management

Warehouse management is an important part of modern enterprise production management.How to make full use of historical orders,spare parts consumption,asset account documents and other relevant data,through the inventory demand forecast,reasonable control of inventory quantity,reduce inventory cost,improve the accuracy of procurement plan and improve the utilization rate of storage space,is always the concern of warehouse management.The purpose of this thesis is to solve the problems that the prediction model in the traditional inventory prediction is not targeted and the prediction accuracy is not high.Based on the background of the warehouse management of the auto parts industry,the thesis focuses on the inventory prediction method based on the demand characteristics,and on this basis,the function transformation of the traditional warehouse management system is realized.The inventory forecasting method based on demand characteristics is based on the precise classification of spare parts.According to the characteristics of auto parts and the actual requirements of manufacturing enterprises,this thesis selects the purchase amount,demand,unit price,spare parts correlation degree as the key demand characteristics.In view of the limitations of the traditional ABC classification method,the fuzzy theory and FSN classification method are combined to improve it.The goal is to reduce the influence of unit price fluctuation on the purchase amount,take the factors of spare parts consumption speed and item correlation into account,so as to make the classification of spare parts more detailed and make inventory prediction more targeted.The improved fuzzy ABC-FSN classification method subdivides the original 3 kinds of spare parts into 10 kinds,namely AF,AS,BF,BS,AN,CF,BN,CS,CN and NN.Inventory prediction models are established for different types of key spare parts.For spare parts with high capital occupation,high demand and strong correlation,time series method is used to forecast inventory demand.In view of the inconsistent consumption speed of spare parts and the possible sudden change of spare parts demand,the improved time series prediction model is designed,taking the influence of time interval on inventory prediction into account,and four kinds of spare parts are selected for example analysis.The prediction results show that the improved model is better than the traditional prediction method,and the average absolute error percentage is reduced by 6.24%.For the spare parts with high capital occupation,general demand and strong correlation,a model is established by Bayesian analysis method for demand prediction.Before the prediction,the distribution function of the failure rate of spare parts is analyzed first,the prior distribution function is determined,and then the Bayesian conjugate prior score is obtained.Finally,the safety stock that meets the market demand level is calculated.The actual example analysis shows that the prediction error of Bayesian prediction model is 8.794%.Both prediction models meet the requirements of enterprises.On the basis of the above research,this thesis extends and improves the function of the existing auto parts storage management system in the actual production enterprises.The system uses XAML,C# language and SQL Server database,and combines front-end framework WPF and back-end framework.NET design and development,increase the inventory warning,inventory quantity forecast,statistical analysis visualization,warehouse visualization and other functional modules,conducive to users to timely understand the inventory situation,improve management efficiency and level.