Robust state estimation for uncertain discrete-time stochastic systems with missing measurements

In this paper, results of robust estimation of are extended to state estimation with missing measurements. A new procedure is derived under the estimation framework suggested in, which is based on penalizing the sensitivity of estimation errors with respect to modelling uncertainties. Compared with the existing estimators, a distinguished feature of the new estimator is that it takes completely the same structure as that of the conventional Kalman filter and can be recursively realized without verifying any LMI existence conditions. Another attractive property of this estimator is that modelling uncertainties can take an arbitrary structure as long as plant parameters are differentiable with respect to them. Numerical simulation results indicate that the proposed estimator even has an estimation accuracy slightly better than the estimator of.