Summary
A composite cage rotor bearingless induction motor (CCR‐BIM) is a complicated system with multi‐variable, nonlinear, and strong coupling. The decoupling control among speed and flux linkage and radial displacement is the base of the stable operation based on the structure, the suspension force and the mathematical model of the CCR‐BIM. A decoupling control strategy based on simulated annealing particle swarm optimization (SA‐PSO) support vector machine (SVM) inverse is proposed for the CCR‐BIM. First, the state equation and reversibility of the system are analyzed using a CCR‐BIM mathematical model and a compounded pseudo‐linear system consisting of the inverse system constructed is decoupled into two displacement subsystems by nonlinear SVM in series with the original system, a speed subsystem, and a flux linkage subsystem. Second, the SA‐PSO is introduced to improve the precision of the prediction model. Finally, results of simulation and experimental verify that the proposed strategy not only achieves the decoupling control between the speed, radial displacement and flux linkage, but also achieves stable suspension and high performance of the CCR‐BIM.