针对电力系统故障诊断中存在的告警信息错误或缺失,以及信息出现的时间不确定等问题,提出一种基于扩展模糊Petri网(EFPN)的故I牵诊断方法。该方法通过将模糊规则映射到扩展模糊Petri网,对故障诊断过程中的不确定性问题进行定量分析。首先,采用统计计算获得信号置信度,运用层次分析法设定权值,进而构建时间隶属函数定量分析信号的时间不确定性。由变电站故障诊断实例可知,该方法在处理不完备信号和误信号时,表现出较好的适应性和容错性,准确度高,且降低了直接使用专家经验的主观性。
Aiming at the problems of false or lacking alarm information, and the uncertain time of information occurrence in the fault diagnosis of power system, this paper presents a fault diagnosis method based on Extended Fuzzy Petri Net. This method does quantitative analysis of uncertainties which exist in the process of fault diagnosis, by mapping the fuzzy rule to Extended Fuzzy Petri Net. First, it uses statistical computing to get the confidence of signals, applies analytic hierarchy process to set weight, and then constructs the time membership function to quantitatively analyze the time uncertainty of signals. The example of substation fault diagnosis shows that, in dealing with incomplete signals and false signals, this method has better adaptability and fault tolerance, shows high accuracy, and reduces the subjectivity of using experts' experience directly. This work is supported by National Natural Science Foundation of China (No. 60773048 and No. 61170223) and State Key Laboratory Open Project Fund of China (No. RCS2009K003).