Rotor Fault Research Articles

Multivariate time-varying complex signal processing framework and its application in rotating machinery rotor-bearing system

Affected by multi-field coupling factors, the vibration response of rotating machinery similar to the hydro-generator unit often exhibits strong time-varying frequency components, which makes rotor fault detection more challenging. The fusion analysis of the vibration signals of multiple bearing sections of the rotor has been proved to be a very effective method for rotor vibration fault diagnosis. However, how to more accurately and synchronously extract the instantaneous features of rotor non-stationary vibration signals associated with multiple sections has been unresolved. To this end, a framework for multivariate time-varying complex signal decomposition of the rotor-bearing system (RBS) is proposed, namely multivariate complex nonlinear chirp mode decomposition. First, the decomposition of multivariate time-varying complex signals is realized by two-stage processing. Second, instantaneous orbit features (IOFs) are obtained through the proposed framework. Finally, a three-dimensional instantaneous orbit map reflecting the time-varying process is constructed through the IOFs. The framework not only realizes the decomposition of the multi-channel time-varying complex signals of the rotor but also simultaneously extracts the instantaneous features of the multi-channel signals. In addition, it also realizes the description of the instantaneous vibration state of the RBS in the non-stationary process (such as startup and shutdown). Simulation experiments show that the framework is superior to other multi-channel signal processing methods in processing time-varying complex signals. The results based on field-measured signals show that the framework can guide the real-time analysis of the signals generated by rotating machinery, which improves the intuition of condition monitoring.

A novel diagnosis indicator for rub-impact of rotor system via energy method

Rub-impact between rotor and stator is a common mechanical fault for rotating machinery, which will bring risks into the normal operation. Detecting the severity of the rub-impact fault accurately is helpful to prevent serious mechanical accidents. However, traditional methods based on the displacements or spectrum analysis may occur some differences toughly distinguished with the naked eyes, and eventually cause inaccurate judgment in some cases. To overcome such shortcomings, this paper proposes a novel diagnosis indicator for rotor systems with rub-impact fault. The dynamic model of the rotor system is established by the finite element method. Based on the vibration power flow theory, the novel indicator is developed by considering the force, velocity, and phase information in a time interval (Time-averaged input power flow). Through theoretical analysis of the rub-impact faults, the energy responses and the law between input power and rub-impact of the rotor system are obtained. Meanwhile, the severity detection of faults by using the proposed indicator has been achieved. The corresponding tests are carried out to validate the accuracy of the theoretical results. The experimental results reveal the proposed indicator can help judge the different severity of rub-impact faults, which means the lower time-averaged input power flow indicates a serious fault. To evaluate the application effect of the proposed indicator in engineering, the rub-impact fault of the GE-T700 power turbine rotor is investigated through theoretical analysis. And the diagnosis results are consistent with the above consequences. The conclusions achieved in this paper can provide a reference for fault diagnosis of rotor systems in engineering practice.

Unbalanced, cracked, and misaligned rotating machines: a comparison between classification procedures throughout the steady-state operation

This article compares two intelligent methods for automatic detection of unbalancing, cracks, and parallel misalignment in rotary machines. The finite element method is used to model the faults in a rotating system. The modeled system then operates virtually under different conditions in the steady-state operation; the vibrational responses are calculated numerically. To compare the accuracy of different manners in the classification of defective systems, firstly, four distinct types of features, i.e., statistical, frequency, time–frequency, and uncertainty are exploited. The T test process is utilized to test the extracted characteristics; the unreliable features are removed from feature vectors, then the remained ones are used in four supervised machine learning classifiers, i.e., support vector machine, k-nearest neighbors, Naive Bayes, and decision trees. In the following, as the convolution neural networks (CNNs) approach, the persistence spectrums of raw signals are plotted, and these graphs are introduced as input data. Comparing results of the different classification methods, it has been observed that although CNNs based on persistence spectrum graphs are computationally heavy and time-consuming, they provide more accurate results than the other classifiers. The results show that the proposed approach for rotor fault detection is effective, accurate, and robust and that it has promise for real engineering applications.

Rotor Winding Short-Circuit-Fault Protection Method for VSPSGM Combining the Stator and Rotor Currents

Rotor winding short circuit faults are common faults for variable-speed pumped-storage generator-motors (VSPSGM). At present, the exciting rotor fault protection of VSPSGM is simple and has low sensitivity. It can only act when the instantaneous value of the rotor phase current reaches three times the rated current. Therefore, it is difficult to cover some rotor winding short-circuit faults with weak fault characteristics. It is urgent to study a novel rotor winding short-circuit-fault protection method for VSPSGM. In this paper, a protection method that combines the stator and rotor currents with different frequencies is proposed. The characteristics of the stator and rotor currents before and after the fault is analyzed by using Clark transformation. On this basis, a specific protection criterion is constructed based on the discrete integral operation, which is easy to implement and not affected by the change of rotor speed. Then, the calculation method of the protection setting is proposed, considering the effect of unbalanced voltage and sensor measurement error. Simulation results show that the proposed method can reliably realize the protection of rotor winding faults. It has faster protection action speed than other methods in the same field. The protection coverage rate is over 90%.

Intelligent detection of stator and rotor faults of induction motor based on improved backstepping sliding mode observer

Intelligent detection of stator and rotor faults of induction motor based on improved backstepping sliding mode observer

Rotor Fault Diagnosis Using Domain-Adversarial Neural Network with Time-Frequency Analysis

Intelligent fault diagnosis of rotors always requires a large amount of labeled samples, but insufficient vibration signals can be obtained in operational rotor systems for detecting the fault modes. To solve this problem, a domain-adaptive transfer learning model based on a small number of samples is proposed. Time-domain vibration signals are collected by overlapping sampling and converted into time-frequency diagrams by using short-time Fourier transform (STFT) and characteristics in the time domain and frequency domain of vibration signals are reserved. The features of source domain and target domain are projected into the same feature space through a domain-adversarial neural network (DANN). This method is verified by a simulated gas generator rotor and experimental rig of rotor. Both the transfer in the identical machine (TIM) and transfer across different machines (TDM) are realized. The results show that this method has high diagnosis accuracy and good robustness for different types of faults. By training a large number of simulation samples and a small number of experimental samples in TDM, high fault diagnosis accuracy is achieved, avoiding collecting a large amount of experimental data as the source domain to train the fault diagnosis model. Then, the problem of insufficient rotor fault samples can be solved.

Model-based Fault Detection and Identification of a Quadrotor with Rotor Fault

Model-based Fault Detection and Identification of a Quadrotor with Rotor Fault

Airgap Search Coil Based Identification of PM Synchronous Motor Defects

Online detection of rotor and load faults in permanent magnet synchronous motors (PMSM) based on spectrum analysis of current or vibration is not capable of identifying the root cause of the fault, since all faults produce identical fault signatures. The sensitivity of fault detection also depends on the motor and controller design making fault detection unreliable. In addition, operation under variable frequency and load limits the effectiveness of spectrum analysis based methods. In this article, airgap flux monitoring is investigated as an alternative for providing reliable identification of rotor and load faults in PMSMs. Based on the analysis of airgap flux under partial and uniform demagnetization, dynamic eccentricity, and load unbalance, an airgap search coil voltage based method for detection and classification of the faults is proposed. The claims made in the article are verified through experimental testing on an IPMSM under emulated fault conditions along with a comparison to vibration and current spectra-based detection. It is shown that the proposed method provides reliable online identification of the faults for cases where conventional spectrum analysis based methods fail.

Prediction of broken rotor bar in induction motor using spectral entropy features and TLBO optimized SVM

The information of the fault frequency characteristics is of great importance for all associated fault diag nostics. This requires a high-resolution spectrum analysis to achieve efficient monitoring of machinery faults, especially while diagnosing rotor bar breakage under light load conditions, because the fault frequencies almost overlap with the fundamental. In this context, rather than looking for frequencies associated with rotor faults, several frequency bands are observed separately in terms of the entropy contained within these bands. First, the motor current signal has been divided into several frequency bands using the continuous wavelet transform (CWT), and the spectral entropy is calculated from each band as the features to describe the rotor condition. Principal component analysis (PCA) has been used as a feature reduction tool, and the features projected onto the first two principal components have been fed into the SVM for inference. SVM is a supervised learning method used for classification and regression analysis. To improve classification performance, a radial basis function (RBF) kernel has been employed, and to find the optimal value of the kernel parameters, a metaheuristic approach, namely teaching learning-based optimization (TLBO), is utilized. The ANSYS 2D workbench is used to simulate the finite element model (FEM) of an induction motor with broken rotor bars, and the efficacy of the proposed method is then tested using simulated data. To investigate the robustness of the proposed approach, white Gaussian noise has also been added to the simulated data, and the performance of the SVM is tested with these spectral features.

Induction Motor Speed Estimation Based on Airgap Flux Measurement Using Hilbert Transform and Fast Fourier Transform

This paper proposes the rotor speed estimation of a squirrel cage induction motor fed by a sinusoidal power supply or an inverter based on the use of a Hall effect sensor installed between two stator slots of the machine. As well known in literature, inverter-fed induction motors usually contains a series of current harmonics and in case of rotor faults, such as broken bars, this operational condition becomes even more severe. The speed estimation has been extensively investigated for its use in control schemes, to compute output power or machine efficiency and particularly for fault diagnosis using motor current signature analysis. In this method, it is necessary an accurate speed estimation to prevent false positives and false negative indications. Therefore, the present research has been carried out by applying the Fast Fourier Transform and Hilbert Transform in the airgap signal measured by the Hall sensor, in order to estimate the rotational speed of the induction motor in a healthy condition and with broken rotor bars. This approach has been compared to the same signal processing techniques using the stator current phase, usually used for speed estimation in the state of the art solutions. The efficiency of this approach was evaluated from simulation and experimental tests, for motor running at distinct load torques, at very low slip conditions and considering a damaged rotor cage. The simulation and experimental results have shown a better accuracy with the measured or simulated speeds, using the airgap estimations when compared to those obtained from stator current processing.

Maximal overlap discrete wavelet packet transforms-based bipolar neutrosophic cross entropy measure for identification of rotor defects

This study propounds a novel methodology for automatic identification of rotor defects severity, when the machine is operated at constant speed, through maximal overlap discrete wavelet packet transforms (MODWPT) and proposed cross entropy measures of bipolar neutrosophic sets, single valued neutrosophic sets and fuzzy sets respectively. After the 3rd level of decomposition of raw vibration signals into eight frequency bands (called as WPT3 bands) through MODWPT, the energy reading in each WPT3 band is monitored. Thereafter, the lower (here called as truth membership degrees) and upper energy bounds from the training and testing samples of various rotor defect conditions at each WPT3 band are extracted and then normalized to lie in [0,1]. Finally, the energy intervals are constructed and thereafter converted into the desired forms of fuzzy sets, bipolar neutrosophic sets and single valued neutrosophic sets consecutively. The minimum cross entropy measure values between the fuzzy sets (Method 1), single valued neutrosophic sets (Method 2) and bipolar neutrosophic sets (Method 3) of training and testing samples of rotor defect conditions are computed and utilized to identify defect conditions such as angular/parallel misalignment of 10 Mils, unbalance, rub and defect free respectively. The novelty of the proposed rotor defect identification methodology lies in the fact that all the proposed variants of bipolar neutrosophic cross entropy measure have been found skilled enough to identify the desired rotor defects under study. On another hand, the existing cosine similarity measure [1] does not possess the necessary capability of identifying all rotor defects, especially, the rub defect condition. Furthermore, like the existing cosine similarity measure, our proposed variants of bipolar neutrosophic cross entropy measures have been found equally compatible for fairly furnishing the consistent and feasible results under intuitive analysis.

Enhanced seeded region growing algorithm and its application in signal decomposition

The components in an engineering signal may overlap in the frequency domain, which causes great inconvenience and even errors in traditional one-dimensional signal processing methods. To identify and extract components in the signal from the time-frequency domain, this paper proposes a time-frequency domain signal decomposition method based on enhanced seeded region growing (ESRG). This paper extends the seeded region growing method to adaptive recognition of target regions in the time-frequency representation and divides the time-frequency representation into several time-frequency components. ESRG can reduce the influence of noise, expand the self-adaptability of time-frequency decomposition and avoid errors caused by manually setting the growth point and threshold. The results of the numerical verification and decomposition of bat echo signals show that this method can accurately decompose multi-component non-stationary signals. The decomposition results of rotor fault and rolling bearing fault signals show that this method can be successfully applied to fault diagnosis.

Robust Dynamic Balancing of Dual Rotor-Active Magnetic Bearing System Through Virtual Trial Unbalances as Low and High Frequency Magnetic Excitation

Abstract This work focuses on in situ residual unbalance estimation of the dual rotor system with the implementation of active magnetic bearing (AMB) as a controller and exciter. The excessive vibration generated due to the presence of residual unbalances limits the operating speed of the system. The compact structure of the dual rotor system provides constraints to the conventional balancing procedure that requires manual addition of the trial unbalances for balancing. In order to overcome the difficulty in balancing of the dual rotor system, an identification algorithm based on the modified influence coefficient method (MICM) is developed for the simultaneous estimation of residual unbalances in both inner and outer rotors with the generation of virtual trial unbalances as magnetic excitation through AMB. The controlling action of AMB attenuates the vibrational response of the system within the required limit and allows the safe operation of the system in the presence of rotor faults and additional excitations. The vibrational responses of the system at the limited locations and the magnitude and phase of the virtual trial unbalances are only required in the MICM for the estimation of unbalances. To numerically illustrate the present methodology, the displacement response is obtained from the developed finite element model of the dual rotor system with discrete disk unbalances and randomly distributed shaft. The robustness of the algorithm in the estimation of residual unbalances is verified with the addition of a different percentage of measurement noises. After balancing, the dual rotor system is found to traverse its critical speed with a less vibrational response.

Thermal Field and Stress Analysis of Induction Motor with Stator Inter-Turn Fault

Inter-turn fault (ITF), a typical motor fault, results in significant variations in the thermal characteristics of a motor. For fault, temperature rise (TR) experiments and thermal field-stress field simulations of an induction motor are carried out to reveal the fault characteristics related to ITF. First, based on the actual structure and the cooling type of the motor, a whole-domain simulation model of the fault thermal field was established. The reasonable equivalence of the motor and the calculation of the heat transfer boundaries were conducted during the modeling process. Then, the three-dimensional transient thermal field under a rated load before and after the fault was obtained, and the accuracy of the simulation could be validated through the comparison of the measured TR at several temperature-measuring points. The heat-transfer law and the notable thermal characteristics of the fault can be presented by analyzing the simulated and measured temperature data. In addition, a fault feature is proposed to provide a reference for diagnosis using the temperature difference of winding at different positions at different moments. Finally, the rotor thermal stress distribution of the normal and faulty motor is obtained by thermal-stress-coupled calculation, which can be used to evaluate the possibility of rotor fault caused by ITF.

Quantitative Analysis of Broken Rotor Bars in Cage Motor Based on Energy Characteristics of Vibration Signals.

The rotor, as the power output device of a cage motor, is subject to a type of invisible fault, BRB, during long-term use. The conventional motor vibration signal fault monitoring system only analyzes the rotor qualitatively for the fault of BRBs and cannot evaluate the fault degree of BRBs quantitatively. Moreover, the vibration signal used for monitoring has nonstationary and nonlinear characteristics. It is necessary to manually determine the time window and basis function when extracting the characteristics of the time-frequency domain. To address these problems, this paper proposes a method for quantitative analysis of BRBs based on CEEMD decomposition and weight transformation for feature extraction and then uses the AdaBoost to construct a classifier. The method applies CEEMD for adaptive decomposition while extracting IMFs' energy as the initial feature values, uses OOB for contribution evaluation of features to construct weight vectors, and performs a spatial transformation on the original feature values to expand the differences between the feature vectors. To verify the effectiveness and superiority of the method, vibration signals were collected from motors in four BRB states to produce rotor fault data sets in this paper. The experiment results show that the feature extraction method based on CEEMD decomposition and weight transformation can better extract the feature vectors from the vibration signals, and the constructed classifier can accurately perform quantitative analysis of BRB fault.

Fundamental Frequency Normalization for Reliable Detection of Rotor and Load Defects in VSD-Fed Induction Motors

Rotor fault is a condition that may occur in induction motors due to high forces and stresses on rotor bars or end rings. When this occurs, the machine may lead to a forced outage in production lines. Several methods have been proposed in the literature for the detection of rotor faults; however, they cannot work under variable frequency conditions, and they are very sensitive to mechanical load oscillations. This paper presents a novel frequency normalization methodology for reliable rotor defect diagnosis in variable-frequency drive-fed systems. The normalization is implemented by demodulating the stator current signal with a time-variant factor. As a result, the fundamental component of the stator current is normalized in a constant frequency value, whereas avoiding spectral leakage and frequency smearing. This enables the transient stator current signature analysis to accurately locate targeted fault signatures. The effectiveness of the proposed frequency normalization approach is demonstrated by analyzing the stator current of motors under non-stationary conditions. Experimental results show that the proposed methodology can detect and distinguish between load oscillations and rotor fault signatures for avoiding false rotor fault alarms.

Inverse thresholding to spectrogram for the detection of broken rotor bar in induction motor

This paper introduces a reliable approach based on start-up transient analysis using inverse thresholding to spectrogram for early detection of a broken rotor bar fault in an induction motor. The investigation has been done in the time–frequency plane since it is based on the non-stationarity assessment of fault frequencies. Typically, time–frequency analysis is carried out in the presence of the fundamental, resulting in the fault features appearing unnoticed. The proposed technique addresses this issue by weakening the fundamental, making fault features more clearly apparent. To achieve this, inverse thresholding to spectrogram has been utilized for motor start-up current. In addition, a quantitative analysis of machines’ health has also been offered. A health score has been calculated based on the spectral entropy in the start-up current to quantify the condition of the motor. To examine the rotor fault, a Finite Element Model (FEM) of the machine is simulated in the ANSYS 2D workbench. The suggested method has been validated with simulated data in multiple simulation scenarios with varying load torques to demonstrate the method’s efficacy. Finally, the diagnosis method is tested against a publicly available data set related to the rotor bar fault.

Maximal overlap discrete wavelet packet transforms and variants of neutrosophic cubic cross-entropy-based identification of rotor defects

The present study proposes a novel procedure for automated testimony of rotor defects through maximal overlap discrete wavelet packet transform (MODWPT) and the proposed neutrosophic cubic cross-entropy, fuzzy cross-entropy and single-valued neutrosophic cross-entropy (NCE) measures consecutively. Discrete wavelet transform is an efficient data decomposition technique, but the technical barrier with this technique is that it can decompose only low-frequency (approximate) signals and also it does not possess the shift invariance property. MODWPT possesses the shift invariance property and is highly capable of decomposing both approximate and detailed signals in time–frequency analysis. After the decomposition of raw signals through MODWPT, the lower and upper bounds of normalized energy readings of various rotor defect conditions (familiar and unfamiliar) are extracted and converted into the forms of fuzzy sets, single-valued neutrosophic sets and neutrosophic cubic sets respectively. The minimum neutrosophic cubic cross-entropy measure as well as fuzzy and single-valued NCE measure values between each rotor defect condition (familiar and unfamiliar) are computed and then utilized to identify various rotor defect conditions, such as parallel and angular misalignment of 7 Mils, rub, unbalance and defect-free respectively. The proposed variants of cross-entropy measure are intelligent in identifying the rotor defect conditions in comparison with the existing cosine similarity measure which exhibits unjustified and meaningless results during mathematical treatments.

Fault diagnosis of rotor based on Semi-supervised Multi-Graph Joint Embedding

Traditional graph embedding methods only consider the pairwise relationship between fault data. But in practical applications, the relationship of high-dimensional fault data usually is multiple classes corresponding to multiple samples. Therefore, the hypergraph structure is introduced to fully portray the complex structural relationship of high-dimensional fault data. However, during the construction of the hypergraph, the hyperedge weight is usually set as the sum of the similarities between every two vertices contained within the hyperedge, and this “averaging effect” causes the relationship between data sample points with high similarity to be weakened, while the relationship between data sample points with low similarity to be strengthened. This phenomenon also leads to the hypergraph cannot accurately portray the relationship of high-dimensional data, which reduces the fault classification accuracy. To address this issue, a novel dimensionality reduction method named Semi-supervised Multi-Graph Joint Embedding (SMGJE) is proposed and applied to rotor fault diagnosis. SMGJE constructs simple graphs and hypergraphs with the same sample points and characterizes the structure of high-dimensional data in a multi-graph joint embedding. The edges of the simple graph are the direct description of the similarity between sample points so that SMGJE can overcome this “averaging effect” of the hypergraph. The effectiveness of the proposed method is verified by two different fault datasets.

A rotor fault diagnosis method based on BP-Adaboost weighted by non-fuzzy solution coefficients

To ensure the long-term safe and reliable operation of mechanical equipment, it is necessary to carry out fault diagnosis of key components in the mechanical system. A rotor fault diagnosis method based on non-fuzzy solution weighted BP-AdaBoost (NFSW-BP-AdaBoost) is proposed in this paper. First, the rotor vibration signal is transformed into a high-dimensional feature set. Then, the VC value is used to obtain the high-dimensional feature sequence. Finally, the NFSW-BP-AdaBoost method is used to diagnose the type and degree of rotor fault. The core of this article is to analyze the fuzzy solution of the combined binary classifier applied to multi-classification problems in decision-making, and put forward the non-fuzzy solution coefficient to weigh the conclusion of the combined multi-classifier, highlighting the sensitivity to the non-fuzzy solutions of each multi-classifier. The experimental analysis results show that the recognition rate of the classification method for fault types can reach more than 97%.