基于希尔伯特-黄变换(HHT)的机组同调研究

Research of Units Coherency with Hilbert-Huang Transform

概述了利用功角曲线进行同调识别所面临的困难,阐述了希尔伯特-黄变换(HHT)在非线性、非平稳功角信号处理中的优势,提出了一种基于HHT的同调识别新方法。该方法利用经验模式分解方法将功角信号进行分解,得到剩余分量和各固有模态信号,逐次对这些分量进行比较得到机组的同调特性。该方法不受系统模型、故障场景限制,能随着不同的故障自适应调整数据时间窗的长度,并能克服小波分析、prony分析等算法难以处理非平稳信号的不足。测试结果表明了该算法的有效性。

The drawbacks of coherency recognition based on generators angles are summarized. The advantage of Hilbert-Huang Transform in the nonlinear and nonstationary signal processing is briefed. Based on HHT, a novel coherency recognition approach is proposed. In the proposed approach, generators angles are decomposed into residual signal and intrinsic mode functions by empirical mode decomposition, from which generator coherency is attained. The approach is not restricted by system models and contingency scenarioes. It can adjust the length of original data adaptively for different disturbances and overcome the drawbacks of wavelet analysis and prony analysis. The effectiveness of the method is validated by test results.