Abstract: A direct current machine monitoring system includes a sensor configured to monitor a load current of the machine and provide a signal indicative of an alternating current component of the load current; and a computer configured for: transforming the signal to provide a time domain transformation of the alternating current component of the load current, identifying features in the time domain transformation corresponding to sparking, and assessing a condition of the direct current machine using the features. The transform may be performed using a wavelet transform. The features may be used to determine a sparking intensity and a sparking frequency in the direct current machine, which may be used to determine a sparking index for assessing the condition of the direct current machine. The sparking index may be compared to a limit value, and statistical analysis may be performed on the sparking index.

Abstract: A direct current machine monitoring system includes a sensor configured to monitor a load current of the machine and provide a signal indicative of an alternating current component of the load current; and a computer configured for: transforming the signal to provide a time domain transformation of the alternating current component of the load current, identifying features in the time domain transformation corresponding to sparking, and assessing a condition of the direct current machine using the features. The transform may be performed using a wavelet transform. The features may be used to determine a sparking intensity and a sparking frequency in the direct current machine, which may be used to determine a sparking index for assessing the condition of the direct current machine. The sparking index may be compared to a limit value, and statistical analysis may be performed on the sparking index.

Abstract: A direct current machine monitoring system includes a current sensor for monitoring load current of the machine; and a computer for obtaining a power spectrum in a range including a machine trait-passing frequency, determining a magnitude of a maximum peak in the power spectrum in a range including the trait-passing frequency plus or minus an uncertainty frequency, and evaluating the magnitude of the maximum peak to assess a condition of the machine. The computer may additionally or alternatively be used for obtaining a low frequency power spectrum of the load current, obtaining at least one magnitude of a component of the power spectrum at a respective predicted frequency, and evaluating the at least one magnitude of the component to assess the condition of the machine.

Abstract: One method for predicting incipient failure includes starting an engine of a locomotive; obtaining charge current sample values of the energy storage unit during energy storage unit charging at a predetermined rate for a predetermined period of time; calculating an average value of the charge current sample values; and using the average value of the charge current sample to predict incipient failure in the energy storage unit or the energy storage unit cabling. A related method includes obtaining crank voltage sample values of the energy storage unit during cranking; calculating an average value of the crank voltage sample values; and using the average value of the crank voltage sample to predict incipient failure in the energy storage unit or the energy storage unit cabling.

Abstract: A method for obtaining a power system phasor includes sampling a power system signal at instants determined by a nominal power system frequency to provide a plurality of data samples; using the data samples to estimate a phasor having an elliptical trajectory; generating a deviation formula representing a frequency deviation between actual and nominal power system frequencies; and using the deviation formula to perform a coordinate transform of the estimated phasor to provide a scaled phasor having a substantially circular trajectory. If the data samples are unevenly spaced, a method in another embodiment includes minimizing squares of errors between the unevenly spaced data samples and an approximating sinusoidal function of the data samples.

Abstract: A method for monitoring a rotor speed of an X-ray tube anode drive includes acquiring axial flux pickup data from the rotor, using the axial flux pickup data to provide a flux spectrum, and estimating the rotor speed by analyzing the flux spectrum. The step of acquiring axial flux pickup data can include situating an axial leakage flux pickup coil so as to pick up a flux signal from an X-ray tube anode drive rotor while the rotor rotates. An analog flux signal can be converted to a digital flux signal and a fast Fourier transform can be used to transform the digital flux signal into a flux spectrum. The flux spectrum can be used to estimate a drive frequency and a slip frequency and the speed can be estimated by subtracting the slip frequency from the drive frequency and dividing by the number of pole pairs of the X-ray tube anode drive.

Abstract: A method for detecting turn faults in an induction motor includes obtaining motor current and voltage waveforms and converting the motor current and voltage waveforms to digitized current and voltage waveforms. Fundamental phasors of the digitized current and voltage waveforms are extracted and, a symmetrical component transform is applied to the fundamental phasors to obtain symmetrical component current and voltage phasors including a negative sequence voltage phasor (V.sub.-) and negative sequence current phasor (I.sub.-). A fault injected negative sequence current (Ii.sub.-) is estimated according to the following equation:Ii.sub.- =I.sub.- -V.sub.- /Z.sub.- -Ir,wherein Z.sub.- is a characteristic negative sequence impedance and Ir is a residual injected negative sequence current. The existence of a turn fault is determined by comparing the estimated fault injected negative sequence current with a threshold fault injected negative sequence current.