The Supplier Evaluation And Selection Based On Evolutionary Neural Network

As the development of information networks, technological progress and demand diversification, the standardized internal management of enterprises makes the relationship between supply and demand more complex. In the integrated supply chain environment, the enterpriseschange the traditional supply management strateg and use new supplier selection and evaluation methods in order to gain competitive advantages. Agile supply chain combines the agile manufacturing and supply chain.It changes the supply and demand relationship from strategic competition into strategic partnership. It figures the quick change their product strategy according to market. It becomes particularly important of how to quantify and evaluate the supplier more precisely.Firstly classify the suppliers based on the added value and competitiveness of suppliers in the supply chain and design the principles of supplier evaluation. Design the new evaluation criterias based on agile management and market effects and comprehensive analysis of the new features in supplier selection.The criterias are made from the point of the whole supply chain benefit considering market response and emergency needs. It evaluates suppliers from five dimensions such as supplier qualification, credibility, service capabilities, supply capacity and long-term cooperation. The quantitative analysis of indicators quantitative method is used to increase the objectivity of the indicators.It does the supplier evaluation by intelligent algorithms which integrated adaptive genetic algorithm and neural network instead of subjective weight. Establish the evolutionary neural network model of strategic suppliers and fuzzy evaluation model of emergency suppliers. The models combine nonlinear and learning characteristicof neural networks, the evolutionary characteristics of the genetic algorithm, reasoning characteristics of fuzzy logic.Principal component analysis is used to reduce the dimension of the sample, remove redundant and abnormal information. It can speed up the convergence rate of neural networks and evolutionary algorithms and achieve the dynamic evaluation of suppliers. The results of the model anlysis show that it can precisely reflect the supplier level of both the existing supplierand the new ones.It’s able to meet the requirement of purchasing department in information technology environment and provide scientific solutions for supplier selection plans in the agile supply chain. Finally, establish corporate procurement management system based on the evaluation criteria and methods. Design the overall system architecture and the software functions and processes are finished by the use of industrial configuration software,.NET platform and database design. Then analyze the operation of the system. The system has been put into use in a high-tech manufacturing equipment company. The supplier evaluation and management is more scientific by the process-oriented and chain-oriented sysntem.