A Mixed H_2/H_8 Approach to Fault Detection for Discrete-Time Systems

This paper deals with the problem of fault detection in linear discrete-time systems of gas turbine which are simultaneously affected by two classes of unknown inputs: unknown finite energy disturbances and white noise disturbances. This problem is formulated as a mixed H₂/H_infin inf filtering problem. The gain matrix of filter is then resolved by solving a convex optimization problem with a set of linear matrix inequality (LMI) constrains. A numerical example is given to illustrate the advantage of the mixed H₂/H_infin inf approach as compared to existing techniques which are based on optimization of H₂ and H_infin criteria.