In this paper, we consider the problem of detecting sensor faults in the aeration system of an activated sludge process. The purpose is to detect possible faults in the dissolved oxygen sensors. The dissolved oxygen concentration in each aerated zone is assumed to be automatically controlled. As the basis for a fault detection algorithm we propose to use the ratio of air flow rates into different zones. The method is evaluated by using the Benchmark Simulation Model no1 (BSM1) via Monte Carlo simulations. Results show that this method gives a good performance in terms of correct an early fault detection and isolation.
Available as PDF (736 kB, no cover)
Download BibTeX entry.