By the empirical mode decomposition (EMD) analysis results of the transient voltage signals, a new method of faulty phase identification was proposed by combing singular value decomposition and Shannon entropy theory. This method uses EMD-based singular value entropy of transient high frequency components to recognize the faulty phase. With calculation and comparison of the EMD-based singular value entropy of the sampled fault voltage signals, the fault type and the faulty phase can be identified. All possible faults of a typical 500 kV line were simulated in Matlab by using the proposed method. The result shows that the proposed method is free from noises,transitional resistance, fault locations and initial fault angle, and give the correct identifications of various faults quickly.%在对暂态电压信号进行经验模态分解(EMD)基础上,结合奇异值分解(SVD)及信息熵理论提出了利用高频暂态分量的奇异值熵实现故障选相.此方法对采集到的故障后电压信号求取EMD奇异值熵,并比较三相间熵值的大小来识别故障类型和判别故障相.基于Matlab环境,对一典型500 kV线路进行故障类型选相的仿真,结果表明该方法不受过渡电阻、故障位置、故障初始角和噪声强度等因素影响,能够快速准确识别各类故障.
展开▼
