Fault diagnosis of high voltage circuit breaker based on multiple entropy strips method

In order to reflect the state of mechanical fault of high voltage circuit breaker better, this paper presents a new method based on multiple entropy strips with which the state characteristics of circuit breaker can be extracted. Firstly, the singular value spectral entropy, the power spectral entropy, the wavelet state space characteristic entropy and the wavelet energy spectral entropy of one signal are calculated, and the minimum and maximum of the above each spectral entropy of many groups of tested data in the same fault condition are extracted. Secondly, the minimum of the above four kinds of spectral entropy is fused to form the minimum eigenvector of the fault, and in a similar way, the maximum value is fused to form the maximum eigenvector of the fault. Finally, after calculating the four kinds of spectral entropy for tested signal, the type of equipment fault can be determined according to the range of entropy strip which the obtained value belongs to. A large number of experiments show that the extracted characteristic parameters with this method can accurately distinguish various ordinary fault types, which can be used as the basis of fault diagnosis for circuit breakers.