Improved Model-Free Predictive Current Control for SPMSM Drives With Adaptive Prediction Horizon Strategy
Zonghao Su, Xiaodong Sun, Gang Lei, Ming Yao
Abstract
This article proposes an improved model-free predictive current control (MFPCC) for surfaced-mounted permanent magnet synchronous motor drives to improve their dynamic performance and robust performance under extreme operating conditions. First, a novel ultra-local model is constructed, and the designed adaptive nonlinear observer is introduced to enhance tracking and estimation accuracy under speed perturbations, particularly in high-speed operation scenarios. This observer incorporates an adaptive law that balances noise suppression and interference mitigation. Second, an extended rotating coordinate system is established to implement an adaptive predictive horizon strategy (APHS) under the MFPCC method, covering the entire dynamic process. Specific techniques are employed to reduce the order of the multistep MFPCC equations. Third, the Newton–Raphson algorithm is utilized to solve the nonlinear equation system constructed via the Lagrange multiplier method, iteratively updating the dynamic rise time. This approach achieves real-time optimization of the voltage vector to improve dynamic performance. Finally, experimental results are provided to validate the effectiveness of the proposed method.