Researches on the Fault Diagnosis of Circuit by Energy Based Information Entropy

The quantitative descriptions of fault responses´ characters are the chiefly presupposition of circuit fault analysis and diagnosis. Aimed to solving the deficiency of traditional information entropy theory in characters extraction of the information mutation, an energy based information entropy model is proposed to measure the complexity feature of signals. Firstly, the energy distribution uncertainty and asymmetry degree of signal sequence in different characteristic space are denoted by the compositional distribution probability of energy based information entropy, furthermore the system states´ mutations and fault responses are converted to the quantitative descriptions of energy based information entropy. Methods in this article are proved feasible to extract signal feature and fault characters by theoretical analysis and experimentations.