Dynamic decoupling control for radial position of bearingless induction motor based on neural networks inverse system

Abstract

The operation principle of bearingless induction motor was introduced, and then the mathematical equation of the radial force model for the bearingless induction motor with conventional 4-pole stator windings and additional 2-pole windings was transformed. A decoupling control method named as neural network inverse system (NNIS) was presented for the radial suspending model of bearingless induction motor. Aimed at the nonlinear and strongly coupled characteristics, the model is analyzed with reversibility and proved to be reversible. The multi variable, strongly coupled, nonlinear system was dynamic decoupled into two linear displacement subsystems by connecting a NNIS before the bearingless induction motor. Then the two decoupled linear subsystems were synthesized under the help of lineal closed-loop controllers. The simulation test results show that independent control on two degrees of freedom of radial position for the bearingless induction motor can be realized through NNIS method and the dynamic and static performance of the closed loop system designed is satisfactory.