Study On Nonlinear Static Decoupling Control For Bearingless Induction Motor

The bearingless induction motor is a new type of high speed motor which hybridizes the motors and the magnetic bearings. It has been the focus of research work on high speed motors since 1990's. As the bearingless motor is constructed with contact-free bearings, it works no abrasion, no lubrication and without particle. Furthermore, compared with conventional magnetic bearing motors, the axial length of the rotor is decreased and the magnetic efficiency is increased. With these advantages and other features, the bearingless machine motor demonstrates the potential applications in space technology, machine tools, high vacuum technology, biologic medicinal engineering, and ultra high-speed drives and so on. And the bearingless induction motor is one of the earliest motors on which are researched, because of its easy field weaken, simple structure, high reliability.The bearingless induction motor is a multi-variable, strong-coupled and nonlinear complicated system. This dissertation focuses on fundamental theory and decoupling control of the bearingless induction motor based on rotor flux orientation including mathcmatic models, the reversibility of system, the use ofα-th inverse system and neural network inverse system method in decoupling of system. The detail contents are as follows:Firstly, the principle of radial suspension force is expounded. The mathmatic models of radial suspension forces are deduced, and the equations of torque force and radial suspension forces are inferred. Then the state equations of bearingless induction motor based on rotor flux orientation is set up.Secondly, the reversibility of system based on the state equations is discussed. As the bearingless induction motor is a multi-variable, strong-coupled and nonlinear system, methods based onα-th inverse system and neural network inverse system theory are used for realizing dynamic decoupling of the system. Cascading the inverse system with the motor system, a pseudo-linear system is completed. So it is decoupled into four kinds of independent linear subsystems—radial displacement subsystems, speed subsystem and torque subsystem, and then the robust servo controller is designed to control each of the subsystems. Then, the control syetem is simulated with Matlab/Simulink toolbox, the simulation results have showed that theα-th inverse system and neural network inverse system method can realize decoupling of thesystem effectively.Finally, the control structure based on digital signal processor (DSP) of bearingless induction motor is discussed; the main hardware system is introduced, and the main software function structure diagrams of the system are given, too. They are the theoretical foundation of the experiment.