An Improved One-Cycle Control Algorithm for a Five-Phase Six-Leg Switching Power Amplifier in Active Magnetic Bearings

To achieve levitation in active magnetic bearing (AMB) systems, levitation coils must be driven by a switch power amplifier (SPA). Therefore, SPA performance is important to stable levitation. In previous studies, five-phase six-leg (5P6L) SPA topology was proposed to reduce hardware costs. However, current tracking performance degrades due to the coupling problem of the five-phase current and freewheeling loss. Then, the rotor may deviate markedly from the balance position, affecting rotor stability. The keys to solving these problems lie in the duty cycle control of the common leg and the compensation of freewheeling loss. This article proposes an improved one-cycle control (OCC) algorithm in which the duty cycle of the common leg is optimized by finite control set model predictive control (FCS-MPC), and the freewheeling loss is compensated for by a high-precision mathematical model. In addition, the parameter uncertainty of the FCS-MPC model is analyzed, and the convergence of the optimization algorithm is investigated. Simulations and experiments show that the proposed algorithm is effective, and the high dynamic response and low steady-state error performance of SPA are achieved.