Research On Parts Purchasing Plan Based On GRU-BP Neural Network

With the deep integration of advanced technologies such as information technology,network technology,communication technology and artificial intelligence,the supply chain shows the ability of rapid reflection,accurate matching,efficient coordination and powerful service in the process of economic development,helping enterprises to enter the benign development process of cost reduction and efficiency enhancement.Operation emphasizes the integration of supply chain management,organic combination of each node enterprise,thus there are many specific problems need to be solved,such as manufacturers as the core node in the supply chain,with the standardization of production parts and components to the upstream suppliers,migration and upstream suppliers requires that the core node in precision matching parts in production,but also closely integrated with downstream sales order required,as a manufacturer of core nodes in which both guarantee supply and supply chain management functions.Product demand plays a guiding role in the manufacturers' reasonable arrangement of purchase quantity of parts.If the product production volume is not properly arranged,the phenomena such as parts shortage,parts inventory backlog and product delivery delay often occur,which will affect the operation efficiency of the entire supply chain.Therefore,it is of great practical significance to study how to effectively predict the product demand within a certain period of time and then further derive the purchase quantity of various parts,so as to scientifically and rationally arrange the purchase of parts,guarantee the production of products and meet the order demand.This thesis takes a manufacturing enterprise with the characteristics of supply chain as the background,which automatically forms the characteristics of transnational supply chain with foreign suppliers,and has high requirements for the cooperation between core enterprises,suppliers and sellers.In terms of product production,there are many types of products in the same series.Due to the limitation of production capacity and market capacity,the demand for multiple types of products in the same series will have mutual influence and limitation.Based on the special needs of manufacturing enterprises,this thesis firstly analyzed and screened the influencing factors of product demand through correlation analysis arid grey relational degree method,and selected 9 influencing factors from the supply chain environment,product attributes and external environment,including market share,load-bearing coefficient and seasonal influencing factors.Secondly,considering the historical demand data on the time series of the adjacent relevance,through data analysis and research in the short term,cycle,three different time series for a long time limit under the influence of each type of product produced by different forms of demand,the factors affecting product demand,GRU helped-BP neural network combination forecast model is set up,to predict future demand over a period of a certain type of product.Adam optimization algorithm is used to solve the problem of local optimization in basic stochastic gradient descent algorithm.After solving the model,the feasibility of the prediction result is proved by the comparative evaluation of the model.Finally,on the basis of predicting the product demand in the next cycle,the author analyzes the parts purchase process of the manufacturer,determines the net demand of the product through the inventory of the product,and analyzes the transformation from.product production to parts demand based on the BOM list of the product.Combined with the inventory classification of various.parts,further develop the parts procurement plan from the upstream suppliers.By A manufacturer to actual enterprise Lily series A?D in five models of seat product demand forecast for the next cycle,to validate the proposed prediction model and procurement planning method can effectively realize the enterprise and the precise matching between suppliers in terms of parts and components production,can effectively guarantee enterprise according to the order production and timely supply,have very strong practical value.,the obtained results have reference to other supply chain.