The permanent magnet linear motor control based on data-driven control theory

Data-driven control methods have no relations with any structural information, and it is designed only by the I/O data of the controlled system. The model-free direct adaptive nonlinear predictive control (MFDANPC) algorithm of linearization of tight format of a class of SISO of a generalized predictive control (GPC) based on data-driven control theory is applied to permanent magnet linear motor speed and position control. The design of controller is based directly on estimate and prediction of pseudo-partial-derivatives (PPD) derived on-line from the input and output information of the motor motion model using a novel parameter estimation algorithm, predicted by approach for multi-degree prediction. Stability, validity and robustness against exogenous disturbance are proved for nonlinear systems with vaguely known dynamics by the simulation examples and real experiments.