Aiming at the uncertainty of traditional fault diagnosis method in analog fault propagation and diagnosis reasoning,a fault diagnosis analysis method based on extended color fuzzy fault Petri net (CFFPN)was proposed. Firstly,based on the traditional Petri net model,we introduced the the color rule of token,place and transition. The forward and backward reasoning were carried by further combining with the fuzzy generative rule and matrix operation,which enhanced the fault tolerance ability. The model of forward reasoning achieved intelligent evaluation of fault state with changing ignition discrimination matrix and reflected the fault propagation characteristics;The backward inference had preferred diagnosis by using the minimal cut sets theory methods and provided diagnostic order to avoid the blindness of the diagnosis. Finally,the pneumatic system was taken as a typical application object,which verified the effectiveness of the above methods.%针对传统故障诊断方法在模拟故障传播及诊断推理过程中存在不确定性等问题，提出一种基于扩展染色模糊故障Petri网的故障诊断分析方法；首先，在采用传统的Petri网模型描述系统故障状态的基础上，引入令牌、库所染色规则，并进一步结合模糊产生式规则与矩阵运算进行故障正反推理，提高模型的容错性；该模型正向推理采用变迁点火判别矩阵实现故障状态的智能评价，反映故障传播特性；逆向推理采用最小割集理论方法进行优先诊断，提供诊断次序，避免诊断的盲目性；最后以气源系统为典型应用对象，验证上述方法的有效性。
展开▼
