A new approach to robust filtering with specified error variance constraints

Considers the problem of robust filtering with specified estimation error variance constraints. The system under consideration is subjected to time varying norm-bounded parameter uncertainties in both the state and measurement matrices. In traditional methods, an H₂ robust filter satisfying error variance constraints can not be obtained by solving only two Riccati equations once. In the paper, the generalized inverse technique of matrices is introduced, and a new algorithm is developed. After two Riccati equations are solved, the filter can be obtained directly, and the variance of estimation error of each state is guaranteed to be not more than the individual specified upper bound. Both the continuous-time case and the discrete-time case are considered in the paper