Sensorless Rotor Position detection of Synchronous Machine using Direct Flux Control – Comparative evaluation of rotor position estimation methods

Direct Flux Control (DFC) is a sensorless method that replaces position sensors in permanent magnet synchronous machines (PMSM). Direct Flux Control extracts the flux linkage signals of machines which have the neutral point accessible and afterwards uses those signals to estimate the rotor position angle. In DFC, the flux linkage signals are obtained measuring the voltage between the neutral point of the machine and an artificial neutral point at particular states during the inverter pulse pattern. This paper presents a comparative evaluation of the different estimation methods that exist within the DFC framework to calculate the electrical rotor position using the flux linkage signals. The comparative evaluation is carried out using a high fidelity software simulation. A finite element model of a Wound Round Separated Excited Synchronous Machine was created, simulated and validated in ANSYS Maxwell. Power electronics, measurement circuits, control and modulation algorithms were modelled and simulated in the software Simplorer. The simulation of DFC is made possible through a co-simulation between the ANSYS softwares' Maxwell and Simplorer. In the simulation developed, DFC is implement using space vector modulation as modulation method to command the synchronous machine with an improved measurement sequence that obtains the flux linkage signals with the same offset level. Using the simulation results, the comparative evaluation shows that among the four existing methods to estimate the rotor position from flux linkage signals, the highest flux linkage signal produces the most accurate rotor position estimation.