Reserch On Suppliers Section And Evalution Based On BP Neural Network

The rapid development of economy and information technology has brought a fierce competition pressure to the company round and round. Since the 21st century, the change of the market and procurement condition has put forward higher requirements to the procurement officers. In order to adapt the market supply which changes rapidly, the companyhas to strengthen their competitiveness in the whole supply chain by coordination and resource integration. As the head of the supply chain, suppliers have a decisive position in the core competence of the company, it’s an important link in the process of optimizing. Based on that background, through qualitative-quantitative analysis, this paper starts a deep discourseon selecting and evaluating the suppliers.First of all, it introduced the background and significance of this research,summarized the status quo of supplier-selecting-evaluating index systemhome and abroad. Secondly, based on the theory of supplier-selecting-evaluating, other theories like supplier-management、BP neural network, etc were defined. With the related Articles synthesized, this paper generalized the construction principles for supplier-selecting-evaluating index such as comprehensive conciseness、objective comparability、operability、extendability、 combination of qualitative and quantitative analysis, integrating the industry characteristics and supplying characteristics of Sichuan Tianyun Engineering Technology Limited, a supplier-selecting-evaluatingindex system for the military enterprises was built, this is one of the innovations in this paper. Thesystem includes 9 first class index:product price、product quality、delivery time、production capacity and flexibility、R&D level、financial system、after-sales service ability、 enterprise honor、information level; 23 two grade index purchased price、 variable cost、quality certification system、product percent of pass、on-site control capability, rate ofproducts returned to the factory、delivery lead time、fill rate、annual production capacity, time flexibility, variety flexibility, quantity flexibility, R&d spending、new product sales satio、ratio of liabilities to assets、liquidity ratio、ROA、 service commitment to fulfill rate、customer complaint satisfaction、industry reputation、enterprise discipline bad record、information system hardware-software level、information system used and maintenance facility levels.After that, different evaluating methods home and abroad were contrasted with, based on the nonlinear characteristics of supplier-selecting-evaluating and the status quo that Sichuan Tianyun Engineering Technology Limited has a lot of reliable supplier-evaluating historical data, we finally chose BP neural network mathematical methodsto researchsupplier-selecting-evaluating, this is another innovation of this paper.And then, based on the supplier-selecting-evaluating system, the BP neural networktheory was used to build a supplier-selecting-evaluating model. First, through studying the sample data, we build the BP neural network, then we start a training simulation on the network, calculate the error between the predicted and measured value, so that we can choose the satisfactory supplier. After that, by useing the data module、network module and output module in the working box of MATLAB, we execute the supplier-selecting-evaluating model inBP neural network. At last, with Sichuan Tianyun Engineering Technology Limited detailly introducing the process step on how MATLAB execute the model in BP neural network, they verified and analyzed the validity and feasibility of this supplier-selecting-evaluating method in BP neural network. I hope this researching method and the conclusion of this paper can provide some reference on supplier-selecting-evaluating research, in theory or in fact.