Sensor fault detection in a biohydrogen production process based on an interval observer and adaptive thresholds

Abstract

In this paper, an interval observer-based fault detection strategy is proposed to deal with faults in the hydrogen flow rate sensor of a biohydrogen production process. A control Hinf-based robust interval observer is designed to (i) estimate the glucose and biomass concentrations from measurements of the hydrogen flow rate produced, (ii) attenuate the effect of the unknown inlet glucose concentration, and (iii) detect faults in the hydrogen flow rate sensor through adaptive thresholds. The gain is designed by solving a semi-definite optimization problem subject to LMIs. The feasibility of the proposed strategy is demonstrated by simulations.