Fault Estimation for Discrete Time LPV Systems under Noisy Scheduling Measurements

Paper in proceeding, 2015

Developed in the paper, is a systematic design framework for parameter-dependent fault diagnosis, with noise corrupted measurements of the scheduling parameter. A stochastic fault estimation filter design method is addressed for a class of discrete-time Linear Parameter Varying (LPV) systems. First, the average valued filter stability is ensured by means of a quadratic dissipativity condition. Apart stability, within the same mean square framework a robust performance objective is defined. Conditions to satisfy robust stability requirements are translated to an optimization problem subject to Bilinear Matrix Inequality (BMI) constraints. A comparative numerical example is given to illustrate and validate the proposed methodology.