Robust time-varying sensor bias estimation for bounded-error systems: Application to the wind turbine benchmark

The objective of this paper is to propose a novel approach for simultaneous state and time-varying sensor bias estimation of dynamic systems. The main advantage of the presented strategy is its simplicity along with robustness to exogenous bounded disturbances. The proposed approach belongs to the wide class of the so-called bounded-error approaches. It constitutes an attractive alternative to the stochastic Kalman-filer-based framework as no knowledge about disturbances/noise distribution is required. The only preliminary requirement is the knowledge about their upper and lower bounds. In this paper, it is assumed that their belong the the ellipsoidal set. Under such an assumption, a novel estimator is proposed along with a comprehensive convergence analysis. It is worth to note that, the convergence analysis is realised with the so-called quadratic boundedness approach. The final part of the paper concerns application of the proposed approach to the wind turbine benchmark.