【摘要】 汽车供应链具有节点企业多、链条长、节点企业间关联度高、技术和资金密集等特点,在竞争白热化、需求多样化的市场背景下,更易受到不确定因素影响,且风险引发后往往损失巨大。依托上海汽车集团及其合作企业进行问卷调查,创造性地将机器学习领域的反向传播神经网络运用于潜在风险因素的重要程度评价,结果显示高风险指标集中于供应商和制造商两段,其中制造商的生产风险、意外风险和财务风险以及供应商的战略风险对汽车供应链整体风险的影响最为显著。不仅从理论上丰富了供应链风险管理的技术方法,同时以中国最大汽车集团——上汽集团的风险评估状况为典型代表,为中国汽车产业供应链的风险防范和安全运转提供具有实践意义的操作建议和决策支持。更多还原

【Abstract】 The competitiveness of a supply chain greatly depends on the risk control ability in supply chain management.Due to such features as many partners involved,a long chain,high correlation between company nodes,and capital and technology intensiveness,the automotive supply chain is more susceptible to uncertain factors,which may lead to huge losses.Focusing on the risk assessment of automotive supply chain,first-hand data are obtained by questionnaire survey from the Shanghai Automotive Group(SAC)and its cooperative enterprises.Then the back propagation(BP)neural network is creatively applied in assessing and ranking the potential risk factors.The results show that the high risk indicators are mainly concentrated in the supplier and the manufacturer.Specifically,the manufacturer′s production risk,accidental risk and financial risk as well as the supplier′s strategic risk have the most significant impact on the overall risk of the automotive supply chain.The main contributions consist of two aspects:one is that the BP neural network is applied to risk assessment of automotive supply chain for the first time,which theoretically enriches and expands the methodologies for supply chain risk management;the other is that the risk status evaluated from the Shanghai Automotive Group,which is the largest one in China,could provide management implications and decision support for risk prevention and safe operation of China′s automotive supply chain.更多还原