Research On Detection And Recognition Of Transient Power Quality Disturbances

In recent years, with the power electronic equipments widely used in power system, power quality has become numerous focus. The detection and recognition of power quality disturbances can provide important information to improve the quality of power. Power quality detection and recognition is the premise to improve power quality, and has become the research focus.This paper mainly studies the transient power quality disturbances detection and recognition method. Based on wavelet transform and support vector machines, we implement transient power quality disturbance detection and classification. Based on wavelet modulus maxima detection signal singularity theory, we used the high frequency wavelet decomposition coefficients of the disturbances signal to realize transient power quality disturbances detection and time orientation. After multi-resolution signal decomposition of PQ disturbances, multi-scale information in frequency domain and time domain of the signal can be extracted as the characteristic vectors. Then the feature vector is input to the SVM classifier. Support vector machines are used to classify these eigenvectors of different power quality disturbances. Through Matlab simulation, we realized the detection and recognition of disturbances. Taking into account noise in practical situation, add white noise to the simulation signal. For the wavelet transform's sensitivity to noise, signal de-noising is done before detection and recognition of disturbance signal. We adopt a soft and hard threshold compromised method to eliminate noise of the signal and achieved the good de-noising effects. Simulation results showed that the proposed method has strong generalization, simple model and high identification accuracy advantages and is a kind of effective power quality disturbances recognition analysis method.