A STUDY OF WAVELET ENTROPY APPLICATION TO ELECTRIC POWER SYSTEM TRANSMISSION LINES FAULT DETECTION

摘要

The wavelet analysis based new measure, wavelet entropy measure, which inherits the merit of wavelet in the analysis of changing resolution and the advantage of the information entropy in the description of signal uncertainty, is able to observe the complexity of the unsteady signals or transients in the time-frequency domain. This paper, stemming from the demands for electrical power transmission line fault detection and identification, taking the fault transient signals as object, study the wavelet entropy measure's application mechanism in the analysis and detection of fault transients far the first time. It also analyzes the fault detection ability of wavelet time entropy, wavelet energy spectrum entropy, wavelet singular entropy and wavelet time-frequency entropy with two theoretic signals by the use of simulation. The result shows that all these wavelet entropies can effectively detect fault time and this exhibits the application potential of wavelet entropy in the power transmission line fault detection and classification.