Novel control strategy for non-minimum-phase unstable second order systems: generalised predictor based approach
Abstract A control structure based on generalized predictor is proposed to control non-minimum phase unstable second order processes with time delay. The scheme contains a predictor structure and a direct synthesis method based primary controller for servo tracking. The predictor structure consists of two filters acting on input and current output which are designed to provide noise attenuation and disturbance rejection. A set-point filter minimises the overshoot caused by the introduction of additional zeros of the controller in the overall closed loop transfer function so as to smooth the tracking performance. Different second order unstable time delay systems are considered and Integral Absolute Error (IAE) and Total Variation (TV) measures are used for comparing the performances quantitatively. The method is implemented experimentally on an inverted pendulum. The proposed predictive strategy is found to provide enhanced control performances in comparison to the existing literature methods.