Self-sensing for electromagnetic actuators. Part I: A coupled reluctance network model approach

Abstract

A self-sensing arrangement in active magnetic bearings (AMBs) comprises a single electromagnetic transducer to realize the actuation and sensing functions concurrently. Minimizing the number of sensing devices and associated interfacing directly reduces possible failure points, system costs, and system complexity. Currently, self-sensing performance is degraded due to problems such as magnetic cross-coupling, eddy currents, saturation, and high losses. This first paper in a two part series presents an integrated model for self-sensing of an 8-pole heteropolar magnetic bearing. The proposed self-sensing approach addresses mechanisms that contribute to modelling error and uncertainty by using several techniques in an integrated structure. A coupled reluctance network model (RNM) is developed which models the coil impedance at the switching frequency. The accuracy of the model is improved by incorporating terms for air gap fringing, complex permeability, and magnetic material nonlinearity. The RNM is verified and refined through a process of iteration using finite element method (FEM) results and experimental AMB measurements. The results demonstrate that a RNM with only 40 nodes can achieve high levels of accuracy when compared to an 80 000 node FEM analysis. In Part II of the series, the refined RNM is incorporated into a multiple input multiple output (MIMO) parameter estimation self-sensing scheme.