Implementation and validation of a transient-optimized sensorless control algorithm for PMSMs

OData support
Dr. Stumpf Péter Pál
Department of Automation and Applied Informatics

In the automotive industry the current trend is that hardware components are replaced by software solutions to reduce costs without compromising reliability. In case of electric motors, the preferred method to acquire the rotor position information is to estimate it with a software algorithm instead of using position sensors. The goal of this project is to implement a transient optimized sensorless control algorithm for permanent magnet synchronous motors (PMSMs), and validate it with simulations.

Nowadays, field oriented control (FOC) is one of the preferred methods for the control of e-motors because of its high performance, smooth operation over the full speed range and full torque generation at zero speed. FOC requires the rotor position and speed information, which can be acquired very precisely by various types of mechanical sensors, for example encoders or resolvers. The drawbacks of these sensors are that they increase the cost of the system, they require mounting space, and they can be potential points of failure. To avoid these problems, the preferred method to acquire the rotor position and speed information is to estimate them with software algorithms.

There are several methods for the sensorless control of PMSMs that work well in the higher speed regions, but problems arise at low speed regions and standstill. Many methods rely on the calculation of the back electromotive force (BEMF) to estimate the rotor position (angle of the rotor) and speed (angular speed of the rotor), but at low speed the magnitude of the BEMF is insufficient because of the high signal-to-noise ratio. There is no BEMF at standstill.

Many sensorless control algorithms are proposed in the literature that are suitable for low speed and standstill operations as well. In most cases the proposed algorithms combine at least two estimation methods, one for low speed and one for medium and high speed. In this project four position and speed estimation methods are implemented in Simulink and tested in various operating conditions. Based on the test results, an optimal combination of the methods is determined and a switching logic is implemented in order to achieve decent behavior in the whole speed range and during transients.


Please sign in to download the files of this thesis.