基于SVD滤波技术与快速四阶累积量ESPRIT算法的异步电动机转子断条故障检测新方法

发布时间：2018-05-07 23:07

本文选题：异步电动机 + 转子断条　；参考：《电工技术学报》2015年10期

【摘要】：提出了一种基于奇异值分解(Singular Value Decomposition,SVD)滤波和快速四阶累积量(Speedy Fourth-Order Cumulants,SFOC)旋转不变信号参数估计技术(Estimation of Signal Parameters via Rotational Invariance Technique,ESPRIT)的异步电动机转子断条故障检测方法。SVD滤波方法可以理想地滤除电机定子电流信号的基频分量与背景噪声,从而凸显转子断条故障特征频率分量;四阶累积量ESPRIT方法可以有效减少噪声干扰、扩展信号阵元并以高频率分辨力提取定子电流信号中的转子断条故障特征频率分量;特别是,将二者结合即可在短时采样信号条件下以高频率分辨力提取转子断条故障特征频率分量。为了改善四阶累积量ESPRIT方法的快速性,提出了精简算法以消除均匀线阵的DOA(direction ofarrival)估计中的大量冗余数据,从而大幅减小计算量。转子断条故障检测实验表明:基于SVD和SFOC-ESPRIT的异步电动机转子断条故障检测方法效果良好。
[Abstract]:Based on singular Value decomposition (SVD) filter and fast fourth-order cumulant Speedy Fourth-Order Cumulants (SFOC) rotation invariant signal parameter estimation technique, this paper presents a novel fault detection method for rotor bar breakage of asynchronous motor based on estimation of Signal Parameters via Rotational Invariance technique and Esprit. The fundamental frequency component and background noise of the stator current signal of the motor are ideally filtered. The fourth order cumulant ESPRIT method can effectively reduce the noise interference, expand the signal matrix and extract the rotor broken strip fault characteristic frequency component from the stator current signal with high frequency resolution, especially, the fourth order cumulant ESPRIT method can effectively reduce the noise interference and extract the rotor broken bar fault characteristic frequency component from the stator current signal with high frequency resolution. The fault characteristic frequency component of rotor broken bar can be extracted by high frequency resolution under the condition of short time sampling signal. In order to improve the rapidity of the fourth-order cumulant ESPRIT method, a reduced algorithm is proposed to eliminate a large amount of redundant data in the DOA(direction of arrival estimation of uniform linear arrays, thus greatly reducing the computational complexity. The test results show that the rotor bar broken fault detection method based on SVD and SFOC-ESPRIT is effective.
【作者单位】：华北电力大学电气工程学院;
【基金】：国家自然科学基金资助项目(51177077)
【分类号】：TM343

【参考文献】