摘要
提出了一种基于连续小波变换(continuous walelet transform,CWT)和奇异值分解(singular value decomposition,SVD)相结合的提升小波系数SVD辨识信号振荡频率和模式信息提取及信号去噪的新方法。克服了噪声较大或者密集模态时,小波脊线不清晰甚至会出现混叠和交叉难以提取频率的情况,根据提升的小波系数奇异值分解频率向量识别各阶振荡模式的频率。同时选用小波能量系数来识别主导振荡模式,用小波软阈值去噪和SVD分解后矩阵重构来进行信号去噪。CWT可以处理含时变振荡模式的低频振荡信号,且对模式参数具有较高的辨识精度。仿真算例验证了算法的有效性和适用性。
Based on the combination of continuous wavelet transform （CWT） with singular value decomposition （SVD）, a new algorithm to identify oscillation frequency of signal, extract mode information and denoise signal by raising SVD of wavelet coefficient is proposed. The condition that under high noise level or closely spaced mode of noise, the wavelet ridges are unsharp and even the frequency is hard to extract due to the aliasing and intersection of wavelet ridges can be overcome by the proposed method, and the frequencies of oscillation modes in different orders can be identified according to frequency vectors of the raised SVD of wavelet coefficients. Meanwhile the wavelet energy coefficient is chosen to identify the dominant oscillation mode, and signal denoising is performed by use of wavelet soft-thresholding denoising and restructured matrix after the SVD of wavelet coefficient. CWT can be used to deal with time-varying low-frequency oscillation signals containing time-varying oscillation mode, and the identification accuracy of mode parameters is high. Both effectiveness and applicability of the proposed algorithm are verified by simulation results.
出处
《电网技术》
EI
CSCD
北大核心
2012年第6期141-147,共7页
Power System Technology
基金
国家自然科学基金项目(51077103)~~
关键词
连续小波变换(CWT)
奇异值分解(SVD)
时变振荡
小波能量系数
主导模式
小波软阈值去噪
continuous wavelet transform （CWT）
singularvalue decomposition （SVD）
time-varying oscillation
waveletenergy coefficient
dominant modes
wavelet soft-thresholdingdenoising
