High speed identification and protection for commutation failure in HVDC systems based on wavelet entropy

The objective of this paper is to propose a new method of identification and protection for commutation failure in HVDC system. The technique of wavelet entropy is introduced by combining wavelet transform with entropy principles. It is used to decompose the transient signal of HVDC system into different resolution levels and extract the feature of the signal. According to the calculation of wavelet entropy, the criterion is defined, which is helpful to make a definite discrimination of commutation failure from dc line short circuit fault and normal operation. By moving sampling windows, the analysis and detection of fault can be realized online. The wavelet protection controller based on wavelet entropy is designed in MATLAB/Simulink. When the controller detects the commutation failure, it will improve the desired value of ignition delay angle and extend the margin of commutation. The control strategy can effectively prevent and mitigate commutation failures caused by different faults. The simulation results show that the controller is with well performance.