Online Detection of Broken Rotor Bar Fault in Induction Motors by Combining Estimation of Signal Parameters via Min-norm Algorithm and Least Square Method

Current research in broken rotor bar （BRB） fault detection in induction motors is primarily focused on a high-frequency resolution analysis of the stator current. Compared with a discrete Fourier transformation, the parametric spectrum estimation technique has a higher frequency accuracy and resolution. However, the existing detection methods based on parametric spectrum estima- tion cannot realize online detection, owing to the large computational cost. To improve the efficiency of BRB fault detection, a new detection method based on the min-norm algorithm and least square estimation is proposed in this paper. First, the stator current is filtered using a band-pass filter and divided into short overlapped data windows. The min-norm algorithm is then applied to determine the fre- quencies of the fundamental and fault characteristic com- ponents with each overlapped data window. Next, based on the frequency values obtained, a model of the fault current signal is constructed. Subsequently, a linear least squares problem solved through singular value decomposition is designed to estimate the amplitudes and phases of the related components. Finally, the proposed method is applied to a simulated current and an actual motor, the results of which indicate that, not only parametric spectrum estimation technique.

Broken Rotor Bar Fault Diagnosis of Induction Motors Using a Hybrid Bare-bones Particle Swarm Optimization Algorithm

在传统定子电流频谱分析中,感应电机转子断条故障特征经常被基波分量淹没而无法准确检测。针对该问题,提出一种基于混合骨干微粒群优化算法的转子断条故障诊断新方法。该方法首先根据电流信号与单位余弦基函数的内积最大准则,利用混合骨干微粒群算法强大的全局搜索能力,准确估计出基波波形参数;然后利用波形参数构造出基波表达式,并将其从原电流信号中剔除,达到突出故障特征的目的。针对微粒群算法在进化后期收敛缓慢的缺点,通过K–均值聚类方式,引入单纯形法对其进行改进,使整个算法的广度探索与深度开发能力得到了有效均衡。最后,对模拟数据和实测信号进行实验,结果验证了所提方法的有效性和优越性。

IBPSO-Based MUSIC Algorithm for Broken Rotor Bars Fault Detection of Induction Motors

In spectrum analysis of induction motor current, the characteristic components of broken rotor bars（BRB） fault are often submerged by the fundamental component. Although many detection methods have been proposed for this problem, the frequency resolution and accuracy are not high enough so that the reliability of BRB fault detection is a ected. Thus, a new multiple signal classification（MUSIC） algorithm based on particle swarm intelligence search is developed. Since spectrum peak search in MUSIC is a multimodal optimization problem, an improved bare?bones particle swarm optimization algorithm（IBPSO） is proposed first. In the IBPSO, a modified strategy of subpopulation determination is introduced into BPSO for realizing multimodal search. And then, the new MUSIC algorithm, called IBPSO?based MUSIC, is proposed by replacing the fixed?step traversal search with IBPSO. Meanwhile, a simulation signal is used to test the e ectiveness of the proposed algorithm. The simulation results show that its frequency precision reaches 10-5, and the computational cost is only comparable to that of traditional MUSIC with 0.1 search step. Finally, the IBPSO?based MUSIC is applied in BRB fault detection of an induction motor, and the e ectiveness and superiority are proved again. The proposed research provides a modified MUSIC algorithm which has su cient frequency precision to detect BRB fault in induction motors.

A Fault Diagnosis Approach for Broken Rotor Bars Based on EMD and Envelope Analysis

Empirical Mode Decomposition (EMD) used to deal with non-linear and non-stable signals,is a time-frequency analytical method that has been developed recently. In this paper the EMD method is used to filter the noise from the stator current signal that arises when rotor bars break. Then a Hilbert Transform is used to extract the envelope from the filtered signal. With the EMD method again,the frequency band containing the fault characteris-tic-frequency components,2sf,can be extracted from the signal's envelope. The last step is to use a Fast Fourier Trans-form (FFT) method to extract the fault characteristic frequency. This frequency can be detected in actual data from a faulty motor,as shown by example. Compared to the Extend Park Vector method this method is proved to be more sen-sitive under light motor load.

Broken Rotor Bar Fault Detection of Induction Motors Using a Joint Algorithm of Trust Region and Modified Bare-bones Particle Swarm Optimization

A precise detection of the fault feature parameter of motor current is a new research hotspot in the broken rotor bar(BRB) fault diagnosis of induction motors. Discrete Fourier transform(DFT) is the most popular technique in this field, owing to low computation and easy realization. However, its accuracy is often limited by the data window length, spectral leakage, fence e ect, etc. Therefore, a new detection method based on a global optimization algorithm is proposed. First, a BRB fault current model and a residual error function are designed to transform the fault parameter detection problem into a nonlinear least-square problem. Because this optimization problem has a great number of local optima and needs to be resolved rapidly and accurately, a joint algorithm(called TR-MBPSO) based on a modified bare-bones particle swarm optimization(BPSO) and trust region(TR) is subsequently proposed. In the TR-MBPSO, a reinitialization strategy of inactive particle is introduced to the BPSO to enhance the swarm diversity and global search ability. Meanwhile, the TR is combined with the modified BPSO to improve convergence speed and accuracy. It also includes a global convergence analysis, whose result proves that the TR-MBPSO can converge to the global optimum with the probability of 1. Both simulations and experiments are conducted, and the results indicate that the proposed detection method not only has high accuracy of parameter estimation with short-time data window, e.g., the magnitude and frequency precision of the fault-related components reaches 10^(-4), but also overcomes the impacts of spectral leakage and non-integer-period sampling. The proposed research provides a new BRB detection method, which has enough precision to extract the parameters of the fault feature components.

供电电压闪变对异步电动机转子故障在线检测的影响及应对

供电电压闪变可能对异步电动机转子故障在线检测产生不利影响,导致基于定子电流信号分析(motor current signature analysis,MCSA)的转子故障在线检测方法失效。通过理论分析,揭示供电电压闪变恶化转子故障在线检测性能的机制。提出免于供电电压闪变影响的异步电动机转子故障在线检测方法,首先,根据转子故障主特征频率分量预判转子健康或故障;继而,根据转子故障独有的辅助特征频率分量进一步确认转子健康或故障。仿真与实验结果证明了该方法的有效性。

基于小波包能量分析和信号融合的异步电机转子故障诊断

为提高异步电机转子故障诊断的可靠性,文中介绍了一种基于小波包能量分析和信号融合的异步电机转子故障诊断方法。采用定子电流信号和振动信号的频谱特征融合作为转子断条以及气隙偏心故障的诊断依据,首先对信号进行小波包分解,获得不同小波包频带节点下对应的能量分布,并与正常电机信号进行比较,进而对能量异常的信号频段进行小波包节点重构,最后通过快速傅里叶变换识别故障特征频率,诊断电机故障是否发生。通过仿真分析,验证了该方法的有效性和实用性,对于电机运行状态的准确监测具有重要意义。

基于Attention-CNN的振动信号电机转子断条识别

针对基于振动信号的转子断条诊断技术依赖于人工特征选择,泛化能力差,以及常规卷积神经网络(Convolution neural network,CNN)模型在对时序信号自动特征提取时忽略序列信息的问题,利用Attention机制对局部特征在整体表达时的度量能力,提出了一种Attention-CNN网络模型。首先通过Attention在原始信号上分配注意力,其次结合CNN进行特征提取构建网络,然后利用粒子群优化算法(Particle swarm optimization,PSO)进行网络超参数寻优,训练转子断条识别模型,最后从整体和局部两个方面进行模型评价。试验结果表明,所提出的识别模型能够达到传统诊断水平,且泛化能力高于现有方法,更适用于通过振动信号进行电机转子断条识别。

融合TKEO和Goerztel算法的感应电机转子断条故障诊断

针对电机电流信号特征分析进行转子断条故障诊断时,在低负载工况或早期,转子断条故障存在故障特征微弱以及难以分离等问题,同时需借助速度传感器获取转速信息,增加了硬件成本,该文提出了融合Teager-Kaiser能量算子(TKEO)和Goertzel算法的感应电机转子断条故障诊断方法。首先,采用Teager-Kaiser能量算子对定子电流进行解调处理,将基频分量转化为直流分量并去除,实现边带故障特征分量到转子断条故障特征分量以及固有偏心谐波分量到转子速度谐波分量的转变,抑制基频频谱泄露对故障诊断和转速估计造成的干扰;其次,利用Goertzel算法对归一化解调信号进行频谱分析,完成转子断条故障特征分量和转子速度谐波分量的快速提取,进而实现不同负载和不同严重程度工况下转子断条的故障诊断。此外,与归一化频域能量算子进行比较,分析其计算效率和稳定性。实验结果表明,所提方法在不同负载和不同严重程度工况下均能够有效地识别转子断条故障特征分量,并且计算量少、可靠性高、实时性强以及硬件成本低。

一起火电厂循环水泵电动机电流波动案例分析

高压异步电动机常见的故障有转子断条裂纹及偏心,故障时定子电流表现出持续波动的特征。本文对某600MW级火电厂2B循环水泵电动机电流进行时域及频域计算与分析,确认电动机转子存在断条裂纹缺陷,并利用广义解调法对电动机电流特征分析(MCSA)结果进行验证;利用转子存在缺陷或故障时定子电流波动的特征,建立基于分布式控制系统(DCS)运行数据的异步电动机电流波动预警算法,并利用该故障电动机的历史电流数据进行验证,结果表明电动机转子断条的持续劣化能够有效地触发预警。

基于RELAX频谱分析方法的鼠笼式异步电动机转子故障诊断

在基于电流信号分析的异步电机故障诊断方法中,故障特征成分(尤其是断条特征)往往容易被基波分量和噪声信号中所淹没。因此,有效地克服基波和噪声的影响是诊断过程的关键。RELAX是一种对加性噪声以及系统误差假设可松弛的算法,提出了一种基于该算法的鼠笼式异步电动机转子故障监测与诊断方法。文中从理论上推导了鼠笼式异步电动机转子故障时的电流信号模型,将RELAX算法应用于电源基波特征参数提取,并从噪声和杂波的连续谱中估计出特定故障的离散谱参数,从而达到消除电源和噪声影响、突出故障特征的目的。通过对样机实测信号进行了分析处理,实验结果验证了RELAX算法的有效性和优越性。

基于Hilbert模量频谱分析的异步电机转子断条故障研究

转子绕组断条是异步电机最常见故障之一。在分析当前主要诊断方法的基础上,提出Hilbert模量频谱分析方法对异步电机转子断条故障进行研究。将Hilbert模量定义为原始信号与其共轭信号的平方和,利用Hilbert模量将原相电流中的基波分量转变为直流分量,原频谱分析的转绕组断条的故障特征分量转变为频率为二倍转差频率的电流分量,决了相电流频谱分析方法中断条故障特征成分容易被基波湮没而难以突出故障特征的问题。文中研究了Hilbert模量频谱分析方法的原理和实现方法,并对偏心故障电机和转子断条电机分别进行仿真和实验研究。仿真分析和实验结果均表明,该方法对异步电机转子断条故障诊断具有较高的诊断精度和很好的频率分辨率,能克服传统的电流频谱法主频泄漏的影响,且工程实现方便。

基于双PQ变换的感应电机转子故障诊断

提出了基于双PQ变换的感应电机转子故障诊断新方法。该方法以三相基波正序电压及其希尔伯特变换构造PQ变换矩阵,用其对定子三相电流进行变换,得到电机有功功率P及无功功率Q。由于正常电机P、Q均恒定,在PQ坐标系下对应1个点;故障电机则由于P、Q中附加频率为2sfs(s为转差率,fs为电网频率)的波动分量而对应于1个椭圆。通过比较电机正常及转子断条故障运行时的P、Q在PQ坐标系下对应的模式,可以实现转子断条故障的可靠检测。选择转子断条电机在PQ坐标系下对应椭圆的长轴为故障特征,椭圆中心与空载运行点之间的距离为归一化基准值,得到了1个近似独立于电机惯量及负载等级的故障严重因子。实验结果证明了该方法的有效性。

基于改进多信号分类法的异步电机转子故障特征分量的提取

在基于定子电流信号进行异步电机故障诊断时,转子断条故障特征频率分量常常被电流的基频分量淹没。针对这一情况,该文提出一种新的改进的MUSIC方法来提取这一故障特征频率。MUSIC方法通过特征值分解把自相关矩阵中包含的信息空间分成信号子空间和噪声子空间两个正交的子空间,该文提出的改进方法是将信号子空间中对应最大主分量的特征向量移到噪声子空间,这样构成两个新的正交子空间I和Ⅱ。子空间I由信号中的最大主分量和噪声所对应的特征向量张成,子空间Ⅱ由其他分量的特征向量张成。把不同频率的信号投影到子空间I,基频信号在该空间的投影将远大于其他的频率分量,因此在投影的倒数谱中,基频分量被抑制,凸显出了故障频率分量。仿真和实验表明,该方法用于提取转子断条故障特征是可行并且是有效的。

利用相关性基波消去法诊断电机转子断条故障

针对感应电机转子故障时,定子电流中新增的故障特征分量被基波信息淹没而难以识别的问题,提出基于相关性基波消去法来检测转子断条故障。根据定子电压与电流基波频率相等的特点,构造出与定子电压同频的参考信号,利用改进的相关算法提取定子电流基波信号的幅值与相位,将基波分量准确滤除来突出转子断条故障信息,然后对滤波后的故障信号作频谱分析,可容易诊断出转子断条故障。该方法无需复杂的坐标变换,计算量小,实时性大大提高。实验结果表明:该方法具备较强的故障辨识能力,即使在轻载的情况下仍能有效对转子故障作出诊断。另外,系统只需采样单相电流信号和电压频率信号,工程开销小,易于实现。

笼型异步电动机转子断条故障检测新方法

计及实际电机本身所固有的非对称、气隙偏心、转子不对中及其它因素,进行高灵敏度、高可靠性的笼型异步电动机转子断条故障检测,是一个亟待解决并颇具工程意义的研究课题,该文即以此作为研究重心。首先完成了笼型异步电动机转子断条故障数字仿真与物理实验。通过分析仿真与实验结果,总结出在转子断条故障检测实践中保证其灵敏度与可靠性的原则及方法。在此基础上,首次将连续细化傅里叶变换、自适应滤波、转子齿槽谐波转差率估计、检测阈值自整定技术有机结合,形成了笼型异步电动机转子断条故障检测新方法。实验结果表明,该方法是切实可行的。

考虑负荷及惯量影响的感应电机转子故障诊断

该文主要目的是分析感应电机转子故障诊断中两个关键问题:①如何定义独立于负载等级及惯量大小的故障严重因子;②转子故障与负荷波动的区分。详细推导了计及速度波动时转子断条故障在定子电流中引入的左右边频分量(1±2s)fs(s为转差率,fs为电网频率)的表达式,并指出其相位之间存在一定的关系。根据负荷波动及转子断条电机特征电流在dcqc坐标系下对应椭圆方向的不同,实现了二者的区分。以仅由转子故障引入的电流分量幅值为故障特征给故障程度1个更精确的量化指标,即得到1个独立于电机负载等级及惯量大小的故障严重因子。对1台4kW电机进行实验,结果证明了该方法的正确性。

Zoom-Prony方法及其在转子断条故障严重程度定量估算中的应用

在基于定子电流信号分析的感应电动机故障诊断中,通过求基波两侧由转子断条故障引起的边频分量的幅值、相位和频率,可以进行故障严重程度的定量估算。由于Prony方法对噪声非常敏感,而实际电动机电流信号包含丰富的谐波和噪声,且故障特征分量的幅值很小,直接应用Prony方法很难估算出故障特征分量的参数。本文提出Zoom-Prony方法,将Prony方法与选带细化技术(Zoom技术)相结合,对转子断条故障特征频率附近的小频段进行局部放大分析,估算这两个边频分量的参数,提高了计算的准确性,缩短了计算时间。此外,由于故障特征频率与基频过于靠近且幅值相差太大,边频分量常常被基频分量所掩盖而无法得到正确的结果,因此本文通过两次应用该方法来消除基频分量的影响。仿真与实验验证了这一方法的有效性。

基于Park矢量的变频电源供电的笼型异步电机转子断条故障诊断方法

基于工频电源驱动时扩展Park矢量方法诊断笼型异步电机转子断条故障的有效性,研究了在变频电源供电情况下该方法的诊断有效性。对断条故障电机变频电流进行了实测、分析和对比,结果表明:当采用变频电源供电时,传统的单相电流频谱分析方法无法正确诊断转子断条故障,而扩展Park矢量方法具有良好的诊断准确性和适用性。

多断条笼型转子感应电动机导条电流分布

为研究转子断条感应电动机的导条电流分布问题,采用基于拓展的综合矢量法建立了易于求解的数学模型。其数学模型的暂态及稳态仿真运行揭示了导条电流分布特征及最大导条电流与断条数量、相对位置的关系。结果表明,导条电流随空间变化的周期为一对极;具有最大电流的导条位置与断条相邻或相距一对极距;在多断条情况下,最大导条电流数值随断条位置而改变;当断条相对位置为一对极距时,最大导条电流达到相应断条数目的最高值;在特殊断条位置情况下,所有正常导条电流数值相等。所得研究结果有助于对感应电机在线故障诊断作正确评估。