Model-Based Levitation Control of A 100 kW Bearingless Electric Motor

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Sähkötekniikan korkeakoulu | Master's thesis
Control, Robotics and Autonomous Systems
Degree programme
AEE - Master’s Programme in Automation and Electrical Engineering (TS2013)
12 + 62
The use of magnetically levitated rotors for various applications, especially in pumps and compressors, has seen an unprecedented rise in the last few years. Bearingless motors combine levitation and torque production capabilities. They offer more compact footprint and require less power electronics compared to more traditional active magnetic bearing supported motors. A lot of significance has been attached to reducing cost, complexity and broadening applicability of the magnetically levitated rotors. Hence, the levitation control of rotors in such bearingless machines has become quite an interesting topic of research. Digital control strategies need to be adopted for proper levitation control of rotors. Furthermore, it has to be kept in mind that these rotors cannot afford to have too many oscillations under different environmental conditions because oscillations can eventually lead to instability and heavy losses. This thesis presents a state-of-the-art model-based digital control of the levitation of a 100 kW bearingless electric motor where the point-mass of the rotor is considered. This motor has a rated speed of 22000 rpm. The entire bearingless motor system is converted into state-space models by taking into account the bearingless machine's nominal operating points and conditions. Then, a model-based controller with Pincer's conditions, coupled with an estimator with Kalman filtering, integral action and state-command path, is implemented and tested for the levitation control. FEM derived Simulink model of the bearingless motor is tested to verify the proposed control strategies. The closed-loop poles and zeroes, step responses of the closed-loop system and the frequency responses are also recorded from the simulations. In the end, the control of the rotor is investigated with five different combinations involving controller, estimator, integrator and state-command path. Comparisons are conducted on the the proposed control strategies and conclusions are drawn based on the findings.
Zenger, Kai
Thesis advisor
Jastrzebski, Rafal
bearingless machine, levitation control, model-based control, DLQR controller, kalman filtering, error integral
Other note