基于粗糙集理论和Petri网的断路器故障诊断

Circuit breaker fault diagnoses based on rough set theory and Petri nets

将粗糙集理论和Petri网相结合,一同用于断路器故障诊断。为了简化断路器故障诊断决策表,采用粗糙集理论中区分矩阵算法和规则提取的矩阵算法对决策表实现条件属性约简与规则简化,找到决策表中隐藏的潜在规则,减少工作量,降低不确定信息的影响,提高故障诊断准确率。用粗糙集理论得到的规则建立Petri网络模型,Petri网可以将知识表示和诊断能力融为一体,利用Petri网实现并行推理,通过简单的矩阵运算对断路器故障诊断进行快速计算。实例分析表明所提出方法准确有效。

The rough set theory （RST） and Petri nets were integrated for fault diagnoses of circuit breakers. In order to gain a reduced decision table of circuit breaker fauh diagnoses, the RST based discernibility matrix algorithm and the rule extraction matrix algorithm were combined to simplify the condition attributes and the rules. The combined method can be used to find the potential rules hiding in the decision table, reduce the workload and the impact of uncertain information, and improve the accuracy of fault diagnoses. Petri nets which integrates the knowledge representation and diagnostic capabilities were built based on the RST extracted rules. By using the parallel reasoning ability of Petri nets, the simplified matrix was caculated to analyze the circuit breaker fault quickly. The correctness and effectiveness were validated by the analysis results of practical fault examples.