The Study Of Suppliers Evaluation And Selection For Company A Based On BP Neural Network

Automotive is a special collection of product.In addition to the body and assembly was done by OEMs themselves,All kinds of accessories need to supply by the auto spare parts companies.The research object company A mainly engaged in production,research and development,application and sale of auto parts and automotive electronic equipment system,occupying a leading position in the development and production of accessories.Since various types of accessories,most of the parts have taken outsourcing approach.It determined suppliers by a rigorous screening.Due to the company’s existing supplier evaluation methods have more qualitative components,and lack of scientific evaluation system and selection mechanisms,and hinder the company to maximize its own cost optimization and cooperation interests.In allusion to company A’s the status quo supplier evaluation system,we use qualitative and quantitative research methods combining with BP neural network to establish a proper,reliable and improvable supplier evaluation model.First,we try our best to understand the existing supplier evaluation system and find the problems.Secondly,through literature studies and questionnaires and combined with the existing index,we divide the A’s suppliers into four types and establish evaluation system for them: Strategic suppliers have 37 secondary indexes 14 primary indexes;influential suppliers have nine first grade indexes and 23 secondary indexes;Competitive/technical suppliers have 5 primary indexes,19 secondary indexes;ordinary suppliers have four primary indexes,16 secondary indicators.And each index is analyzed and quantified.Again,by analyzing and comparing vendor evaluation method,it found that BP neural network has a strong adaptive ability,generalization ability,parallel and distributed processing capabilities,learning ability,etc.,so we select BP neural network as A company’s supplier evaluation model,due to limited space,we choose the starter components-front drive cover’s strategic suppliers as the example,collecting the existing supplier data to training the model.Finally,we simulate the process to predicte the best vendor partners.This method will facilitate the A to evaluate the best supplier.