电能质量扰动识别

Poor quality power supplies could interfere with communication networks, increase power losses, shorten lifespans of electrical/electronic equipment, and result in various malfunctions of power generation, transmission, distribution, and end-users’ systems. One of the crucial tasks, therefore, is to ascertain what quality problems that the power grids are currently suffering and what are the patterns and the occurring frequencies of them. Electric utilities and regulators could then find countermeasures accordingly to mitigate the impacts. In the paper, the authors present a novel power quality (PQ) disturbance recognition system with multiclass classifiers exercising techniques of support vector machines and error correcting output codes. Furthermore, a Fourier transform based feature extraction is proposed by finding the connection between the PQ disturbances and the relevant Fourier magnitude and phase spectral components. Simulations have shown that the developed PQ disturbance system with simplified feature extraction and linear classifiers can achieve superior performance compared with other counterparts in terms of simplicity of structure, high predictive precision and robust performance.