基于粗糙集理论的往复压缩机规则提取方法研究

A Research for Rule Acquisition of Reciprocating Compressor Based on Theory of Rough Set

知识获取一直是困扰诊断机器智能发展的瓶颈问题。往复压缩机因其故障诊断知识尚没有达到深入的地步,使得往复压缩机智能诊断更多地依靠故障案例做指导,缺乏良好的解释能力。依据粗糙集理论能够解决模糊性和不确定性问题的能力,通过建立案例决策表、计算差别矩阵、属性约简和规则提取,将往复压缩机故障案例转化为有效的规则知识,为实现基于规则推理的智能诊断系统提供知识基础。

Knowledge acquisition has always been a bottleneck problems that plagued the intelligent development of diagnostic machines. Because the knowledge of reciprocating compressor fault diagnosis has not yet reached the point of in-depth, the intelligent diagnosis of reciprocating compressor most rely on fault cases which make the guidance and have the shortcoming of lack of explanatory power. Based on rough set theory which has the ability of resolving the ambiguity and uncertainty problems, this article takes the steps of establishing case decision table, calculating discernibility matrix, attribute reduction and rule extraction, then transforms failure cases of reciprocating compressor into effective knowledge rules. This method provides the knowledge base for the realization of the rule-based intelligent diagnostic system.