State estimation for Markovian Jump Linear System using quantized measurements

This paper investigates the state estimation problem of Markovian Jump Linear Systems (MJLSs) with quantized measurements. A moving horizon Monte Carlo (MHMC) sampling method is proposed in this paper to solve the state estimation problem. Both the state and the possibility of the mode are estimated at each instant. The proposed method makes full use of the statistical knowledge of the mode jumping to handle the problem of unknown mode. In addition, it considers the probability distribution of the measurement in a quantized interval, therefore it can get better performance than using the existing state estimation methods for MJLSs, in which the quantized measurements are regarded as the inputs of the estimators directly and the statistical knowledge of the quantized interval is not considered. Simulation example is presented to show the effectiveness of the proposed method and its advantage over one existing method for state estimation of MJLSs.