Please enter the words you want to search for:

[Return to folder listing]

   Voltage Square-Wave-Injection-based HF Parameter Identification Method for Sensorless Control of a Synchronous Reluctance Machine   [View] 
 [Download] 
 Author(s)   Martha BUGSCH 
 Abstract   For sensorless control at zero or low speed, extending the bandwidth by higher HF signal-injection frequencies is desirable. Therefore, knowledge of the HF machine parameters for the electrical angle calculation is necessary. Conventional methods to determine the HF machine parameters using sinusoidalsignals reach their limits using a standard industrial converter. The restricted switching frequency limitsthe possible number of voltage steps for sinusoidal-shaped signals of higher frequencies. Moreover, thesampling frequency to reconstruct the current signal also is limited without current oversampling.An alternative and simultaneously simplified signal shape to avoid these limitations is a square-wave-shapedvoltage signal. This kind of signal, in contrast to approximately sinusoidal-shaped signals, containsa different amount of higher harmonics if it is realised via pulse-width modulation. Thus, it is notrecommended to use HF machine parameters that are measured with sinusoidal-signal injection even ifthe same frequency would be possible. Hence, a new method needs to be developed to measure the HFmachine parameters for this signal shape and HF frequency under the restrictions of a standard industrialconverter.This paper proposes a method to estimate the time-discrete transfer function between these square-wave-shaped voltage signal and the resulting current response with a least-square estimation method. A timesynchronous averaging algorithm is implemented to reduce the measurement noise of the sampled currentvalues. As a result, the HF admittances necessary for electrical angle estimation can be calculated.Measurement results are also presented. 
 Download 
Filename:0523-epe2017-full-12374820.pdf
Filesize:3.05 MB
 Type   Members Only 
 Date   Last modified 2018-04-17 by System