Modified robust finite-horizon filter for discrete-time systems with parameter uncertainties and missing measurements

A robust finite-horizon Kalman filter is designed for linear discrete-time systems subject to norm-bounded uncertainties in the modeling parameters and missing measurements. The missing measurements were described by a binary switching sequence satisfying a conditional probability distribution, the commonest cases in engineering, such that the expectation of the measurements could be utilized during the iteration process. To consider the uncertainties in the system model, an upper-bound for the estimation error covariance was obtained since its real value was unaccessible. Our filter scheme is on the basis of minimizing the obtained upper bound where we refer to the deduction of a classic Kalman filter thus calculation of the derivatives are avoided. Simulations are presented to illustrate the effectiveness of the proposed approach.