Moving Horizon Set‐Valued State Estimation for Discrete‐Time Switched Descriptor Linear Systems

ABSTRACT Set‐membership state estimators (SMSEs) are proposed in this work for discrete‐time switched descriptor linear systems (SDLS) subject to bounded uncertainties. The proposed set‐valued observers provide both point‐valued estimations of the system state vector and guaranteed bounds on the estimation error. First, the observer gains are computed based on an performance to mitigate the negative impact of uncertainties on the accuracy of the estimation. Second, new moving horizon algorithms are designed to compute rigorous bounds of the system state vector using symmetric boxes and zonotopic set computations. An illustrative example is presented to show the interest of the proposed algorithms and to support the theoretical results.