The Research Of Solving Uncertainty In Fault Diagnosis Based-On Rough Set

Uncertainty Problem is a bottleneck in Fault Diagnosis Expert System application and development. The core work in the thesis is applying rough set theory to solve uncertainty knowledge representation and uncertainty reasoning.The thesis presents compressor's fault-sample model and obtains centrifugal compressor's vibration fault-sample original decision table based-on rough set theory and the accumulated data in global compressor diagnosis. The thesis obtains centrifugal compressor's final decision table applying clear matrix and data analysis method, and extracts fault-rule. According to vibration characteristic information a new approach for compressor fault diagnosis based on rough set theory is proposed. The basic thread is that we compartmentalize vibration spectrum information many spectrum-sections, which are regarded as condition attribute, to consider all kinds of probable malfunctions. The inherent redundancy in the spectrum information also is revealed. The approach presents a new idea to compressor fault diagnosis on the incomplete characteristics.