Research On Warning System Of Supply Chain Crisis

Along with the knowledge economy, the strategic fellowship becomes more and more important under the direction of dynamic and globalization competition, so the Supply Chain Management gets the advantage place in the enterprise and the academic field because of its capacity of response, flexibility and low costing. Currently, the differentiates of the same kind of product from different business become more and more minor, the business which has the technology advantage has been fewer, so the competition is not between the isolate business, but the supply chain. In order to be the priority in own field, the early-warning system of the supply chain in risk has been regarded as the "battle armor" to offer protecting driving escorting.This thesis researches on the risk of supply chain and inducement factors of crisis based on supply chain management theory, risk management theory and crisis early-warning theory. This thesis applies system engineering methods. Then, an index system of crisis early-warning performance and the early-warning system have been established, which achieve successful results.First of all, this thesis analyzes the development of supply chain management and crisis management comprehensively, including the development history of supply chain, current research focus, and development trend of both supply chain management and crisis management.Secondly, this thesis analyzes the inducement factors of supply chain crisis, and proposes the study hypotheses of inducement of crisis. Based on the hypotheses, survey questionnaire has been designed. Through making statistic analysis on the results of questionnaire, rationality of hypotheses of inducement of crisis has been estimated. After eliminating and categorizing the inducement factors, inducement factors system has been proposed.According to the index system design strategy, this thesis indexes the inducement factor of supply chain crisis on the basis of scientific analysis of inducement factors of supply chains. Besides, every index has been discussed and the early-warning index system of supply chain crisis has been established.Finally, early-warning model of supply chain crisis has been created based on support vector regression, which is a variant of support vector machine. Support vector machine is based on statistics learning theory and structure risk minimization. They represent special advantages when solving the pattern recognition problems with finite training set, non-linear and high dimension. In the end, this thesis choose the Dongfeng Electric Co., Ltd to verify the early-warning model.