Preview control for LPV/LFR networked control systems with packet dropouts
TL;DR
This paper addresses the challenge of observer-based preview control for linear parameter-varying networked control systems subject to simultaneous dual-channel packet dropouts in both the measurement output and control input channels, and develops a novel observer-based preview control scheme that effectively minimizes steady-state errors and enhances overall system performance.
Abstract
This paper addresses the challenge of observer-based preview control for linear parameter-varying (LPV) networked control systems subject to simultaneous dual-channel packet dropouts in both the measurement output and control input channels. The primary objective is to achieve high-precision reference tracking and robust stability despite random data losses. To this end, two mutually independent Bernoulli-distributed sequences are employed to characterize the dropout phenomena, and an augmented error model is constructed by embedding preview information of the reference signal into the system dynamics. A novel observer-based preview control scheme is then developed. Methodologically, by combining the Lyapunov function approach with a one-step design procedure, and leveraging linear fractional representation (LFR) techniques augmented with slack variables, less conservative sufficient conditions for stochastic stability are derived in terms of linear matrix inequalities (LMIs). Quantitatively, numerical simulations demonstrate that, under specific dropout rates, the proposed method achieves approximately a 61.30% reduction in performance metrics, demonstrating that the preview action effectively minimizes steady-state errors and enhances overall system performance.