| Most magnetic bearing control schemes use a bias current with a superimposed control current to linearize the relationship between the control current and the force it delivers. For most operating conditions the existence of the bias current requires more power than alternative methods that do not use conventional bias. With the existence of the bias current, even in no load conditions, there is always some power consumption. In earthbound applications of magnetic bearings this constant power loss may not be of critical importance but in aerospace applications it becomes an important concern. This research examines alternative methods which diminish or eliminate bias current.; The alternative methods presented here consequently reduce electrical power loss, but do not necessarily provide equal or better performance. Comparisons of the performance of these methods are made against the industry standard bias method. Some of the characteristics that are examined are slew rate, force output, response to saturation effects, control amplitude and frequency dependency, and steady-state operation. In evaluation of a couple of methods an optimization strategy is used to choose control parameters resulting in the best trade-off between power consumption and control performance. Subsequently, all methods will be compared and ranked according to their power loss and performance trade-off results.; This analysis is performed using computer simulation after which the proposed methods are used in a physical laboratory experiment. The computer simulation involves using Matlab/Simulinkt{dollar}sb{lcub}rm tm{rcub}{dollar} simulation software on a microcomputer. The experiment involves a magnetic bearing apparatus and supporting electronics. The magnetic bearing apparatus includes a flexible shaft, two magnetic bearings, and various disks which are used for rotational inertias, and a motor. Only one magnetic bearing will be used for control purposes. |
