Experimental Verification of Adaptive Control in Active Magnetic Bearings

The strong nonlinearity and uncertainty of the active magnetic bearing (AMB) makes the controller difficult to design. In many applications the design of the control is performed under the assumption that all system parameters are exactly measured in advance and do not change during operation. For these plants, it is sufficient to design a robust controller for a limited range of operation (Peter Wurmsdobler, 1997). In high performance machinery, however, a change in operation condition may easily occur. This event may lead to destabilizing finally. This paper presents a concept of adaptive control in combination with KALMAN filter, in order to guarantee the optimal control behavior in various operating condition and a desired performance for a wide range of operation. This paper extends this tropic from the former experiment to a test facility with strong nonlinearity / uncertainty and draws a brief comparison to a fixed control design.