一种面向数控系统的递阶式风险评估方法

Hierarchical Risk Assessment Method for CNC System

针对数控系统中复杂特性对安全性能的需求,提出一种递阶式风险评估方法.给出递阶式层次结构模型将复杂系统逐级分解简化,构建评估序列集矩阵以融合各层级专家经验,设计序列集模糊映射方法将专家经验进行量化,从而获取各层级中因素权重,然后从顶层开始将各层级因素权重进行逐层传播扩散直至底层,从而定位关键风险因素;同时辅以水平分片法,与安全等级匹配规则集相结合,确定数控系统风险程度等级.最后采用实际案例对所提出方法和有向图矩阵法进行性能对比,结果证明该方法能够有效减少评估结果相似度,降低评估时间,适合于数控系统风险评估过程的应用.

To meet the requirements of safety of Computer Numerical Control （CNC） system, a hierarchical risk assessment method was presented. The hierarchical model of CNC system was proposed, and an assessment sequences set matrix was deeply researched with the corresponding fuzzy mapping algorithm to quantize the expert experience. A layered diffusion method of factor weights was proposed to locate key risk factors, and a horizontal fragmentation method was adopted to calculate the risk degree of CNC system. And then case studies were adopted to compare the performance between the digraph and matrix method and the new mechanism. The results proved that the new mechanism could decrease the similarity performance, reduce the execution time, and be more suitable for the application of risk assessment process in CNC systems.