H∞ filtering for markovian jump linear systems with mode partial information

The goal of this paper is to present new results for H_∞ filtering for continuous-time Markov jump linear systems with partial information on the jumping parameter. The central hypothesis considered here is the existence of a suitable detector which provides measurements of the Markov chain. This detector-based approach allows us to treat the case with complete observations, no information and cluster observations of the jumping process. The main result comprises a method for designing a mean square stable linear H_∞ filter by using the information given by the detector. The proposed filter design is given in terms of linear matrix inequalities. Furthermore, the result is applied to the state estimation of an unmanned aerial vehicle model.