Adaptive Reliable Filtering against Sensor Faults

The problem of H∞ filtering has been a topic of recurring interest for some decades. Comparing with H2 filtering, the advantages of H∞ filtering approach are twofold. First, the assumption of boundness of the noise variance is loosened. Second, the H∞ filter tends to be more robust when there exist additional uncertainties in systems, such as quantization errors, delays and unmodeled dynamics [132]. A great number of results on H∞ filter have been reported and different approaches have been proposed in the literature [41, 44, 88, 139, 138, 146].