Rangarajan M. Tallam

A Survey of Methods for Detection of Stator-Related Faults in Induction Machines

As evidenced by industrial surveys, stator-related failures account for a large percentage of faults in induction machines. The objective of this paper is to provide a survey of existing techniques for detection of stator-related faults, which include stator winding turn faults, stator core faults, temperature monitoring and thermal protection, and stator winding insulation testing. The root causes of fault inception, available techniques for detection, and recommendations for further research are presented. Although the primary focus is online and sensorless methods that use machine voltages and currents to extract fault signatures, offline techniques such as partial discharge detection are also examined. Condition monitoring, fault diagnostics, insulation testing, interlaminar core faults, partial discharge (PD), temperature monitoring, turn faults.

A robust, on-line turn-fault detection technique for induction machines based on monitoring the sequence component impedance matrix

A simple and robust sensorless technique for online stator winding turn fault detection based on monitoring an off-diagonal term of the sequence component impedance matrix is proposed in this paper. Due to the destructive and rapidly propagating nature of insulation failure, it is critical to detect turn faults in an early stage to prevent further damage to the motor. The motor nonidealities such as the variation in the supply voltage unbalance, slip-dependent influence of inherent motor asymmetry, and measurement errors must be taken into account in order to reliably detect a turn fault in its incipient stage. Simulation and experimental results on a 5 hp induction machine are given to confirm the validity of the proposed method. It is shown that the proposed turn fault detection scheme is simple and is capable of providing reliable fault detection that is immune to the motor nonidealities.

Stator winding turn-fault detection for closed-loop induction motor drives

Sensorless diagnostics for line-connected machines is based on extracting fault signatures from the spectrum of the line currents. However, for closed-loop drives, the power supply is a regulated current source and hence, the motor voltages must also be monitored for fault information. In this paper, a previously proposed neural network scheme for turn fault detection in line-connected induction machines is extended to inverter-fed machines, with special emphasis on closed-loop drives. Experimental results are provided to illustrate that the method is impervious to machine and instrumentation nonidealities, and that it requires lesser data memory and computation requirements than existing schemes, which are based on data look-up tables.

A carrier-based PWM scheme for neutral-point voltage balancing in three-level inverters

A significant problem with neutral-point-clamped three-level inverters is the fluctuation in the neutral-point voltage. In this paper, a capacitor voltage balancing technique for carrier-based three-level pulsewidth modulation (PWM) is developed, with improved voltage control capability at high modulation index. The method incorporates a novel scheme that requires measurements of only the motor currents and capacitor voltages to implement voltage balancing control, and does not need to determine the direction of power flow between the inverter and the motor.

Universal Induction Motor Model with Low-to-High Frequency Response Characteristics

A three-phase induction motor model that depicts the motor behavior over a wide range of frequencies from 10 Hz to 10 MHz is presented in this paper. The model is universal in the sense that common-mode, differential-mode, and bearing circuit models are combined into one three-phase equivalent circuit model. The proposed model is basically an extension of the low-frequency IEEE Standard 112 circuit model. The proposed model was experimentally simulated and verified with results presented.

A survey of methods for detection of stator related faults in induction machines

As evidenced by industrial surveys, stator related failures account for a large percentage of faults in induction machines. The objective of this paper is to provide a survey of existing techniques for detection of stator related faults, which include stator winding turn faults, stator core faults, temperature monitoring and thermal protection, and stator winding insulation testing. The root causes of fault inception, available techniques for detection and recommendations for further research are presented. Although the primary focus is on-line, sensorless methods that use machine voltages and currents to extract fault signatures, off-line techniques such as partial discharge detection are also examined.

Determination of Parameters in the Universal Induction Motor Model

A systematic procedure to determine the parameters of a previously proposed low-to-high frequency induction motor model is presented. Analysis of the high-frequency behavior with regard to the impact of magnetic core selection, parasitic inter-turn and winding-to-frame capacitors, and skin effects of windings is investigated in greater detail for the proposed universal model of an induction motor. The model is universal in the sense that it is derived by extending the low frequency standard-T-equivalent circuit (IEEE Std. 112) to include high- frequency effects under both common-mode and differential- mode domains, and can be used for transient reflected wave studies and EMI emissions in motor drive systems. A test-based method and an analytical approach, respectively useful in application and design stages, is presented to determine the frequency characteristics of the induction motor. Findings of the investigation were verified experimentally with results presented in this paper.

Common Mode and Differential Mode Analysis of Three Phase Cables for PWM AC Drives

This paper proposes a new 3-phase transmission line model with the first successful use of 2D finite element analysis (FEA) to extract and predict cable R-L-C parameters over a wide range of frequencies. The significance is that the FEA methodology to extract high frequency parameters were shown to be accurate and eliminates costly cable testing. The 3-phase FEA parameters can also be transformed for use with historical single-line transmission line models. The FEA approach allows unsymmetrical cables to be studied in a 3-phase cable-drive and motor simulation setting. FEA predicted vs. measured R-L-C parameters as function of frequency was within 5-10%. A 3-phase drive-cable-motor power electronic simulator program using predicted FEA parameters to predict CM current and DM motor overshoot (ringing) voltage was found to be within 7-15% of measured waveforms from 10 ft to 1000 ft cable length variation for two different cable type configurations. Analysis provided is also applicable to power system transient studies on AC line connected equipment

Sensorless speed measurement of AC machines using analytic wavelet transform

A new method for the sensorless speed estimation of AC machines, using the analytic wavelet transform of the stator current signal, is proposed for a direct torque control drive. A comparison with results obtained using the short-time Fourier transform is included. The proposed method can be implemented in real time on a digital signal processor. The time-frequency resolution obtained and the computation time required by the proposed algorithm are improved in comparison to existing techniques, and the method can be applied over the entire speed range.

Common-Mode Voltage Reduction PWM Algorithm for AC Drives

The impact of common-mode voltage (CMV) generated by pulsewidth modulation (PWM) ac drives on motor bearings is well known. Several algorithms for CMV reduction have been proposed in the literature. While a few algorithms assume ideal switching and fall apart when nonidealities like inverter dead time are considered, some others are effective only in a limited operating range of the drive. In this paper, a previously proposed algorithm is modified for practical implementation to include compensation for dead-time and reflected-wave motor overvoltage stress while still producing output voltage waveforms with reduced common-mode content. Experimental results are provided to show the reduction in CMV over the entire operating range, with other performance attributes such as reflected-wave motor overvoltage that are identical to conventional space-vector PWM. The advantages of applying the algorithm to a fully regenerative ac drive are also demonstrated.