A Study On The Evaluation Methods Of The Supply Chain Risk Based On BP Neural Network

As a system of integrated management model, supply chain management hasbecome a hot topic between academia and enterprises. However, the environment ofthe node enterprises on the supply chain is unstable, once one part of the supply chainoccurs any risks, it will affect the normal operation of the entire supply chain.Therefore, it is very important to do the researches on all aspects of the supply chainrisk management. The study on evaluation methods of the supply chain risk, whichcan implement supply chain risk management for the enterprise, promotecollaboration between node enterprises and ensure the normal development of thesupply chain, has a very important significance.The study, learning from the existing research results, takes evaluation of thesupply chain risk as the research object, designs evaluation system of the supply chainrisk and builds the evaluation model based on the BP neutral network. Thus the studybased on the basis compares with other evaluation methods and comes to thesuperiority of the proposed method. In the paper, study background on the supplychain risk is firstly introduced, reviewed the relevant research of home and abroad andalso discussed the necessity and importance of the supply chain risk evaluation.Secondly, the study introduces the definition and methods of the supply chainevaluation, theory and characteristics of BP neutral network, analyses the feasibilityof evaluating the supply chain risk based on BP neutral network. By consulting theexcellent core researches in this field, understand the impact of supply chain riskassessment factors and build evaluation system of the supply chain risk. Then,implement BP neutral network to build the supply chain risk evaluation model and useMatlab to realize the model. Finally, compare the method proposed in the study withother evaluation methods and draw the effectiveness and superiority of the method inthe study.The artificial neutral network evaluation model built in the paper for supply chainevaluating has high accuracy, the relative error of the result, in terms of the supplychain risk sample, is less than10-3. Artificial neural network, solving such highlynon-linear problem of supply chain risk assessment, can still be considered to haveunique advantages and provide effective decision support for enterprise supply chain risk management.