Prediction-based Adaptive Robust Compensation for Motion Control of Time-delayed Uncertain Linear Motors

The simultaneous presence of parameter variations, time-varying disturbances / uncertain nonlinearities, and the delayed signal in the control loop make the motion control of an iron core linear motor system difficult. Existing model-based control methodologies can handle only a subset of the above imprecisions and hence become insufficient to accurately track the desired motion. In this paper, a prediction-based adaptive robust control (PARC) design is developed for high performance motion control of linear motors subjected to above complexities. The proposed design, suited for a class of $n^{th}$ – order delayed uncertain dynamical system, comprises: i) prediction-based model compensation to attenuate the effect of input delay; ii) prediction-based projection type learning mechanisms to reduce the parameter uncertainties; iii) an improved robust prediction scheme that factors in both parameter and disturbance uncertainties; iv) and prediction-based nonlinear robust feedback to attenuate the effect of model approximation errors and disturbances. Performing all these simultaneously, the designed controller can successfully track a desired motion and achieve a guaranteed robust performance. The simulation results demonstrate improved performance of the proposed PARC scheme compared to conventional model reference adaptive robust control designs.