N. Rama Devi

Bio: N. Rama Devi is an academic researcher from Bapatla Engineering College. The author has contributed to research in topics: Stator & Induction motor. The author has an hindex of 4, co-authored 5 publications receiving 48 citations.

Diagnosis and classification of stator winding insulation faults on a three-phase induction motor using wavelet and MNN

Abstract: The growing industrialization needs techniques to diagnose the incipient faults in induction motor at inception stage itself for avoiding the downtime of the production. In this regard detecting the faults in a 3-phase induction motor at an early stage is a vital component in process industries. The condition of the supply unbalance, under voltage and sudden load changes are other involuntary issues which may tend to exhibit current signature similar to the stator winding insulation faults. This paper proposes a robust technique to detect, classify various stator winding insulation faults and severity of stator inter-turn faults when an induction motor works under various operating conditions. In the present work, disturbance features are extracted from three phase residues which are obtained from wavelet multiresolution analysis. Three modular neural networks are implemented, in which one is used to classify various disturbances such as single phasing, supply unbalance, under voltage, stator inter-turn faults, sudden load change and phase faults, second one is used for classifying the stator winding phase faults and third one is used for identifying the faulty phase and severity level of stator inter-turn faults. Simulation and Experimental data demonstrate the validity of the proposed method and improvement in classification accuracy as compared to traditional method.

Wavelet ANN based stator internal faults protection scheme for 3-phase induction motor

Abstract: This paper proposes a protection scheme based on Wavelet Multi Resolution Analysis and Artificial neural network which detects and classifies various possible stator winding fault of a three-phase induction motor such as inter turn faults, line to ground faults and line to line faults. The wavelet decomposition of three-phase stator currents is carried out with Bior5.5. The maximum value of absolute peak d1 coefficients of three-phase currents is defined as fault index which is compared with a predefined threshold to detect the fault. The normalized peak d1 coefficients of these currents are fed to a Feedforward neural network to classify various faults. The algorithm has been tested for various incidence angles and proved to be simple, reliable and effective in detecting and classifying the various stator winding faults.

A transient current based bus zone protection scheme using wavelet transform

Abstract: This paper deals with the application of wavelet transform for the detection of busbar faults and providing backup for external faults In the event of internal faults the transient behavior of feeder and zone-2 currents is opposite to that of zone-1 source current In the event of feeder or zone-2 faults, the corresponding currents exhibit similar transient behavior as that of source current The transients of these current signals are analyzed using wavelet based multi resolution analysis to obtain detailed coefficients over a narrow moving window The fault indexes of each current signal, defined based on these detailed coefficients, are utilized to detect the internal faults as well as external faults The scheme is tested for different types of internal and external faults with various incidence angles and fault impedance The proposed scheme discriminates internal faults from external faults even in the presence of current transformer saturation

Wavelet — ANN based fault diagnosis in three phase induction motor

Abstract: This paper proposes a protection scheme based on Wavelet Multi Resolution Analysis and Artificial Neural Networks which detects and classifies various faults like Single phasing, Under voltage, Unbalanced supply, Stator Turn fault, Stator Line to Ground fault, Stator Line to Line fault, Broken bars and Locked rotor of a three-phase induction motor. The three phase Induction Motor is represented by a universal model which is valid for a wide range of frequencies. The same has been simulated using MATLAB/Simulink software and tested for various types of motor faults. The wavelet decomposition of three-phase stator currents is carried out with Bi-Orthogonal 5.5 (Bior5.5). The maximum value of the absolute peak value of the highest level (d1) coefficients of three-phase currents is defined as fault index which is compared with a predefined threshold to detect the fault. The normalized fourth level approximate (a4) coefficients of these currents are fed to a Feedforward neural network to classify various faults. The normalized peak d1 coefficients of three-phase currents are fed to another Feedforward neural network to identify the faulty phase of stator internal faults. The algorithm has been tested for various incidence angles and proved to be simple, reliable and effective in detecting and classifying the various faults and also in identifying the faulty phase of stator.

Detection and identification of stator inter-turn faults in three-phase induction motor in presence of supply unbalance condition

Abstract: Induction machines play a vital role in process industries due to their low cost, ruggedness and low maintenance. Even though the induction machine is very reliable, many failures can occur due to their non-ideal operating conditions. Literature survey reveals that stator faults occupy a prominent place among the reasons for such failures. In particular, an undetected stator inter-turn fault may progressively lead to a permanent damage of the machine. Hence, early detection of stator inter-turn faults is essential for preventing damage to the adjacent coils and the core of the stator. Also, detection of stator inter-turn faults in the presence of supply unbalance is another challenging task. This paper proposes a new adaptive approach based on wavelet multi-resolution analysis for detecting and identifying the stator inter-turn faults. A fault index is defined based on the slope of detail coefficients to compare with an adaptive threshold for setting the flags. A fault is detected when the flag count reaches 6 over a moving window of 10 samples. Severity of the fault is identified by defining a sensitivity index based on the three phase energies of 4th level approximate coefficients. The proposed method is verified by using experimental data considering supply unbalance as well. Results indicate the effectiveness of the proposed method.

Optimal Feature Selection via NSGA-II for Power Quality Disturbances Classification

Abstract: This paper presents an application of nondominated sorting genetic algorithm II (NSGA-II) for multiobjective feature selection in power quality disturbances classification. Classification error and number of features are collectively minimized to ensure good accuracy and feasible computation time. NSGA-II gives different Pareto-optimal solutions based on the combination of objectives. Considering equal priority for both the objectives, a fitness function is provided to retrieve the best solution set from the first Pareto-front. S-transform and time–time transform are employed for detection and feature extraction. Decision tree is used for classification. The proposed technique is tested on disturbances simulated as per IEEE-1159 standards and real disturbances acquired from an experimental setup. The results show quick convergence, admirable accuracy, and reduced computational time.

Incipient Fault Detection in Stator Windings of an Induction Motor Using Stockwell Transform and SVM

Abstract: In this article, Stockwell transform (ST) is used to analyze the stator current signals for diagnosis of various motor conditions such as healthy, stator winding interturn shorts, and phase to ground faults. ST decomposes the current signals into complex ST matrix whose magnitude has been utilized for the fault detection. The nature of the fault, that is, ground or interturn is identified using the zero sequence currents followed by postfault detection. Two separate frequency bands are defined to extract the features which are fed to two different support vector machine (SVM) models for faulty phase detection for both types of faults. Under both cases, a heuristic feature selection approach is utilized to find the optimal features for classification purposes. Average classification accuracy of 96% has been achieved for both types of faults.

Early detection and localization of stator inter-turn faults based on discrete wavelet energy ratio and neural networks in induction motor

Abstract: This paper proposes an improved diagnosis method for early detection and localization of Inter-Turn Short Circuit (ITSC) faults in the stator winding of the induction motor (IM). The main advantages of the method are the simplicity, low cost, and accurate diagnosis of these types of faults such that it can detect and localize even a low number of shorted turns faults in the stator winding of the motor. This is achieved by using a novel indicator that is based on the Discrete Wavelet Energy Ratio (DWER) of three stator currents, with Artificial Neural Network (ANN). Three different models of typical neural networks, namely, Multi-Layer perceptron (MLP), radial basis function (RBF), and Elman Neural Network (ENN) based on Bayesian Regularized (BR) training algorithm are proposed for ITSC classification based on fault feature extraction using discrete wavelet transform. To test the effectiveness of the proposed method, several experimental tests were carried out under different operating conditions of the IM, which contains the healthy and the ITSC faults cases that have experimented under various loads and different numbers of shorted turns in the three phases of the motor. The obtained results proved that the DWER is an accurate and robust indicator to diagnose the ITSC fault, this is confirmed by ANN results which gave the best results with the Bayesian regularized Elman network model that has the highest performance with minimal error rate is 10−9.Consequently, the combination DWER-ENN has assured its ability to accurately detect high and even low numbers of the shorted turns and localize the defective phase even within various loads in the IM.

Convolutional Neural Network-Based Inter-Turn Fault Diagnosis in LSPMSMs

Abstract: Stator inter-turn fault diagnosis system for electric motors is of a considerable concern due to its significant effect on industrial production. In this paper, a new method for detecting the inter-turn fault and quantifying its severity in the line start permanent magnet synchronous motor (LSMPSM) is proposed. The new method depends on monitoring the stator current during steady-state period to detect the fault. The convolutional neural network (CNN) method is proposed to correlate the motor steady-state current with the status of the motor winding conditions and detect any presence of inter-turn faults. The data used in this study is extracted from both an experimental setup of a one-horsepower LSPMSM and the corresponding verified mathematical model through several testing cases under various loading conditions. One of the main features of the proposed technique is that it does not require separate feature extraction phase. The results indicate that the proposed technique is able to detect the inter-turn fault under different loading conditions varies from 0NM to 4NM with accuracy of 97.75% for all defined fault levels. The use of steady-state current for fault detection regardless of motor load enables the proposed technique to detect the fault online without disturbing the system functionality and reliability as well as without adding any extra hardware to the system.

A New Demagnetization Fault Recognition and Classification Method for DPMSLM

Abstract: This paper investigates a new method for the demagnetization fault recognition and classification in double-sided permanent magnet synchronous linear motors that are used in linear motion applications This method is based on time–time-transform (TT) coupled with extreme learning machine (ELM), which are especially suitable for the industrial occasions such as motor batch demagnetization inspection before delivery and periodic maintenance First, a finite element analysis model with demagnetization faults is built to extract three lines (up-line, center-line, and down-line) magnetic flux density signals Second, TT is first applied to conduct magnetic signals waveform transformation, and digital picture processing technology is innovatively used to extract the pixel rate of its diagonal elements contour surfaces as the fault feature Then, machine learning algorithm called ELM is utilized as a classifier to obtain the unique fault labels that can represent the demagnetization occurred positions, sides, and severity types in detail The validity and superiority of ELM is verified through comparison with back propagation neural network, and probabilistic neural network Finally, prototype motor experimental platform is designed to confirm the correctness and effectiveness of this proposed method