Martha Bugsch

Extending the bandwidth of high-frequency test signal-based sensorless control methods for low- and zero-speed ranges can be achieved by increasing the test signal frequency. A square-wave-shaped voltage injection (SWSVI) enables this. Originally, a SWSVI is a pure alternating voltage injection with a constant voltage amplitude. However, it is recommended to use adaptive voltage levels for Synchronous Reluctance Machines (SynRMs), otherwise the non-linear machine behaviour will result in operating-point-dependent current responses and therefore cause additional audible noise and HF losses. This method has the disadvantage that the operating-point-dependent voltage levels have to be determined in advance. This paper proposes an alternative to pure adaptive SWSVI that actively controls the high-frequency (HF) current ripple caused by SWSVI using a HF current control with a static pre-control. Experimental results demonstrate the good performance of this sensorless control method.

For many industrial applications, sensorless control of synchronous reluctance machines (SynRMs) is beneficial. Additional sinusoidal high-frequency (HF) test signals are often injected for sensorless control at standstill and in the low speed range. However, they have a number of drawbacks for SynRMs combined with a standard industrial inverter and without current oversampling. Hence, this paper presents an approach tailored to SynRMs using square-wave-shaped voltage test signals. They simultaneously increase the bandwidth, because they enable a higher HF frequency. The paper takes into account the highly nonlinear machine behaviour by using adaptive HF voltage levels. It compensates for the electrical angle error caused by cross-coupling effects and for that caused by load steps or dynamic behaviour during demodulation of the HF current amplitude. Only a special regular-update current sampling method is sufficient. A two-degree-of-freedom (2DoF) current and speed control is implemented. The theoretical background of the method is explained, and the control concept is validated on the basis of experimental results.

An increasing number of sensorless control methods are being developed for a recently rediscovered machine type: synchronous reluctance machines (SynRMs). The advantage of sensorless control methods is that the position sensor can be omitted. Many alternative methods are conceivable, but in all of them, it is impossible to observe and estimate the electrical rotor position at standstill and at low speed without additional excitation via high-frequency (HF) test signals. To calculate the position information from the current or voltage response of these test signals, a detailed machine model is necessary. The challenge is to identify the parameter of the machine model that characterises the HF machine behaviour. Thus, there is a need to find a proper identification method and to analyse these HF parameters for different HF test signal frequencies and amplitudes. Thereby, one can determine which test signal frequencies and amplitudes are best suited for sensorless control of SynRMs. Therefore in this paper an HF parameter identification method is described and the corresponding measurement results for a SynRM are analysed. A test current signal injection method is used. Different test signal frequencies and amplitudes, as well as the position dependencies, are compared. It can be concluded that, first a appropriate test signal frequency and amplitude combination has to be selected that offers enough voltage reserve to control the fundamental behaviour of the machine. Moreover, the HF parameters should be measured exactly for the combination used because there is a dependency of the HF parameters on the test signal frequency and amplitude that is significant but not equally strong. Position averaged measured values ensure high-precision HF parameter results. SynRMs are suitable for saliency-based sensorless control methods based on test signal injection because they have an inherent machine parameter saliency due to their rotor design.

Increasing interest in synchronous reluctance machines (SynRMs) shows the need to examine sensorless control methods for this machine type. For different speed ranges, different sensorless methods are necessary. The basic control structure used here is a cascaded field-oriented control (FOC) with a Two-Degree-of-Freedom (2DoF) current control with adaptive tracking controllers. This paper compares two alternating high-frequency (HF) test signal injection-based sensorless control methods for standstill and low speed. The first one superimposes an alternating HF test current signal and the second one an alternating HF test voltage signal onto the fundamental quantities in the estimated synchronous reference frame. To control the whole speed range, a method for the medium and upper speed range is necessary. Therefore, a back-EMF-based method is used. The implementation- and measurement-based differences for both alternating HF signal injection methods are analysed and the existing advantages and drawbacks are compared. Experimental results are included. They prove a similar behaviour of both alternating HF test signal-based sensorless control methods.