Characteristic Model Based All-Coefficient Adaptive Control on Flexible Rotor Supported by Active Magnetic Bearings

Active magnetic bearings (AMBs) have found a wide range of high-speed rotating machinery applications in energy storage, electric power generation, machine tool operation, heart pumps and hybrid vehicles. Compared with conventional mechanical bearings, AMBs need feedback control to support and constrain the rotor and demand additional electronic devices.

When AMBs are used to support flexible rotors, their nonlinear characteristics and complex dynamics of flexible rotors cause problems with stabilization and disturbance rejection. Proportional-integral-derivative (PID) controllers are the most popular control method for AMB applications. They are easy to implement and reasonable control performance can be achieved. However, it is difficult for PID controllers to deliver robust and near optimal performance for complex dynamic systems. In recent years, robust control techniques such as μ-synthesis have also been applied to AMB applications. The μ-synthesis is able to better handle the uncertainties in complex systems and to achieve reliable performance. However, it requires plant and uncertainty models to generate the most suitable solution, which makes this control design not as practical as PID controllers.

The characteristic model-based all-coefficient adaptive control (ACAC) method has been widely used in process control and the aerospace industry. Numerous applications have demonstrated its effectiveness. This method is able to provide robust control performance on multidimensional complex dynamical systems, while maintaining a simple structure and not requiring the actual plant model. It focuses on the characteristics of the plant and the control performance requirements instead of precise system dynamics. It compresses the corresponding information of the high order plant into several characteristic parameters so that no information is lost. The simplicity of the characteristic model based ACAC method makes it convenient for practical engineering applications.

This thesis explores the application of the characteristic model based all-coefficient adaptive control method to the stabilization of a flexible rotor AMB system. Simulation has shown performance improvements over the μ-synthesis in terms of minimizing the vibration and maintaining a small orbit. Experimental results have also demonstrated some features of ACAC method and shown it is comparable to the μ-synthesis in certain measures. These results show its strong potential to perform well in spite of its simplicity.