A novel method for induction motors stator inter-turn short circuit fault diagnosis based on wavelet energy and neural network

Fault Tolerant Control Strategy Based on Model Predictive Control and Unscented Kalman Filter for Permanent Magnet Synchronous Motor

“…The time update equations include the calculation of the covariance matrix P based on (14), and are weighting factors and they are equal to 1 2( + ) ⁄ . Finally, the measurement update equations will include the Kalman gain for the correction of the next state estimation ̂ and covariance matrix P .…”

“…However, these techniques take a long time in analysing the measured data which may not be suitable for all types of faults. Also, Artificial Intelligence (AI) techniques are being used such as knowledge-based fault diagnosis techniques for PMSM faults, for example, Neural Networks [14], fuzzy logic [15], and particle swarm optimization [16]. However, the AI techniques require a set of data logging measurements for the machine in case of fault which sometimes not available in all machine cases.…”

On-Line Detection of Coil Inter-Turn Short Circuit Faults in Dual-Redundancy Permanent Magnet Synchronous Motors

“…Assuming that the mechanical angle of the small teeth is κ, the cross-sectional view of DRPMSM, the star graph of the fundamental EMF and the diagram of phase separation, as well as the stator windings outspread diagram of DRPMSM are shown in Figures 1-3 spectrum of voltage used to detect ISCFs could be unstable, so it is important to find a way to diagnose the ISCFs of DRPMSMs under non-stationary conditions. References [27][28][29] presented some methods for diagnosing the ISCFs of motors under non-stationary operation conditions. In order to detect the ISCFs of permanent magnet machines under varying speed and load conditions, an adaptive algorithm based on extracting non-stationary fault sinusoids using current signals was proposed in [27].…”

“…In order to detect the ISCFs of permanent magnet machines under varying speed and load conditions, an adaptive algorithm based on extracting non-stationary fault sinusoids using current signals was proposed in [27]. In [28], a wavelet neural network technique was adopted to detect and locate ISCFs in induction motors under non-stationary operation conditions. The discrete wavelet energy related to the fault was generated by the DWT of the stator current and used as the input for neural networks.…”

Wavelet transform to induction motor analysis: review

“…In some cases the energy information associated with the faults is used as input to a neural network [39][40][41][42][43][44] which detects and locates it immediately. In other cases, in order to improve the results, the information obtained with WT has been combined with that extracted one using the Hilbert Transform [41], [45][46].…”