一种基于FPN的变电站故障推理机制

针对变电站故障推理过程中存在的不确定性问题,提出一种基于模糊Petri网的变电站故障推理机制。对于告警信号漏报、误报问题,设计模糊诊断规则,采取模糊推理策略;对于适用多条规则的争议性信号,在信号单一及信号重复时分别给出分配方案;引入岭型隶属函数,计算在时间上信号对规则的隶属度。通过故障诊断实例的仿真,验证了该推理机制的有效性,并对实验数据进行比较分析,阐述不同的信号接收情况及时间分布对推理结果产生的影响。实验表明,该FPN推理机制可在不确定的信息基础上进行较为准确的推理。

Fault reasoning mechanism for substation based on FPN

For the uncertain problems existing in fault reasoning process for substation, this paper presents a substation’s fault reasoning mechanism based on fuzzy Petri nets. For the problems of missing alarm signals and false signals, we develop fuzzy diagnosis rules, and adopt the fuzzy reasoning strategy. For controversial signals which apply to several rules, when the signal is single or repeated, we give the distribution plans respectively. We calculate the signals’ membership degree in time by introducing ridge type membership function. Through the simulation of fault diagnosis example, it verifies the effectiveness of this reasoning mechanism, and comparative analysis of the experimental data is conducted to describe the influence on reasoning results by different signal reception and time distribution. Experiments show that this FPN reasoning mechanism can do more accurate reasoning based on uncertain information.