Self-sensing for electromagnetic actuators. Part II: Position estimation

Abstract The second paper in a two part series presents the position estimation scheme for an 8-pole heteropolar active magnetic bearing (AMB). The integrated self-sensing assembly comprises a nonlinear MIMO parameter estimator together with a coupled reluctance network model (Part I), by which problems associated with magnetic cross-coupling and saturation are collectively addressed. The parameter estimator utilizes the first harmonic components of the current and voltage waveforms to determine estimates for the x and y rotor positions. Magnetic saturation is accounted for using a saturation factor, which scales the demodulated coil currents to ensure that the actuator with the lowest flux density contributes the most to the position estimate. Basic functionality and feasibility of the proposed self-sensing scheme are demonstrated via an experimentally validated transient simulation model (TSM). The TSM incorporates magnetic effects such as eddy currents, cross-coupling, and hysteresis. In the second part of this work, the static and dynamic performance of the self-sensing sensor is evaluated. The influence of magnetic cross-coupling, saturation, and duty cycle variation on the position estimate is documented. The results demonstrate the importance of including a mutual coupling term in the position estimation model in order to stably suspend the rotor. Furthermore, stability margin analyses indicate that the robustness of the magnetic bearing control is satisfactory for unrestricted long-term operation.