Robust Loss Minimization Control for Three-Level Inverter-Fed Linear Induction Machine Drive System Adapted to Linear Metro

Existing loss minimization control (LMC) methods for linear induction machine still suffer from poor parameter robustness and limited effect on system loss suppression. To solve these issues, this article proposes a robust LMC method for three-level inverter-fed LIM drive system under low switching frequency of linear metro. First, a hybrid optimal primary flux search method is proposed to select the optimal flux reference, where the loss model and the gradient descent method are combined for system loss minimization and robustness enhancement. Second, an ultralocal model-based predictive flux control with two-step prediction optimization is developed to flexibly manipulate optimal flux under low switching frequency. By observing the lumped disturbance, the reference voltage vector (VV) independent of parameters can be obtained for optimal VV determination. Moreover, after incorporating with the magnetizing inductance identification-based flux observer, the parameter robustness of the core parts in existing LMC methods, i.e., flux observation, optimal flux selection and manipulation, has been significantly enhanced. Finally, comprehensive simulation and experiment have been conducted, showing that compared to the existing LMC method, the system loss is further reduced by 1.9% and 5.3% under accurate parameters and mismatched parameters, respectively.