基于改进矩阵束滤波与检测的异步电动机故障诊断新方法

发布时间：2018-07-11 16:17

本文选题：异步电机 + 故障诊断　；参考：《电工技术学报》2015年12期

【摘要】：针对传统电机故障诊断技术,断条故障边频易被基频淹没、对短时数据分辨率低、复合故障引入平方交叉项等缺点,提出了一种改进矩阵束滤波与检测的异步电机转子故障诊断技术。首先,利用定子电流实测信号构造Hankle矩阵,并对其进行奇异值分解,形成不同的信号子空间。然后对奇异值的部分置零,剔除了基波和噪声的影响,从而准确地估算出故障信息。最后,通过对改进MP算法得到的特征故障频率幅值,定义了转子断条故障严重因子。试验对异步电机转子断条和偏心的故障进行验证,表明该算法可以有效抑制强大的基频和噪声对故障特征频率的影响,且具有远高于快速傅里叶变换(FFT)的高分辨率。
[Abstract]:In view of the traditional fault diagnosis technology of motor, the edge frequency of broken bar fault is easily submerged by the fundamental frequency, the short time data resolution is low, the square cross term is introduced into the composite fault, and so on. An improved matrix beam filter and detection technique for rotor fault diagnosis of asynchronous motor is proposed. Firstly, Hankle matrix is constructed by using the measured signal of stator current, and it is decomposed into different signal subspaces by singular value decomposition. Then the partial zero of the singular value is eliminated and the influence of fundamental wave and noise is eliminated and the fault information is estimated accurately. Finally, the fault severity factor of rotor broken bar is defined by the amplitude of characteristic fault frequency obtained by improved MP algorithm. The fault of rotor bar break and eccentricity of asynchronous motor is verified by experiments. It is shown that the proposed algorithm can effectively suppress the influence of strong fundamental frequency and noise on the fault characteristic frequency, and has a higher resolution than that of fast Fourier transform (FFT).
【作者单位】：华中科技大学强电磁工程与新技术国家重点实验室;华中科技大学电气与电子工程学院;山东电力工程咨询院有限公司;
【分类号】：TM343

【参考文献】