A Soft Fault-concept Diagnosis Method of Analog Circuits Based on Cloud Model Theory

As the component parameters of analog circuits are influenced by tolerance, a soft fault-concept diagnosis model in this paper, which is based on the qualitative concepts of cloud model theory, is proposed to achieve the complete description of fault states and the fast diagnosis. The diagnosis model uses the cloud transform method to represent the conceptual soft classification of the multiple test-point voltage intervals. At first the model generates the atomic concepts of test-point voltages which are indicated by the characteristic values of cloud model, and promotes these atomic concepts to the qualitative concepts of test-point voltage which are easy to be understood. Then the model forms the fault category concept table based on a combination of qualitative concepts of multiple test-point voltages which are associated with the range of fault component parameters. Lastly according to the table, the soft fault-concept classification and diagnosis of analog circuits are achieved in this paper based on these formalization concepts. The simulation results show that the soft fault-concept method provides a good solution of analog circuits fault diagnosis avoiding the effect of component tolerance, at the same time the method has achieved a higher diagnosis accuracy rate.