Nonlinear model predictive control of anaerobic digestion process based on reduced ADM1

Abstract

In this paper, Nonlinear Model Predictive Control (NMPC) algorithm is developed to optimally control the anaerobic digestion process in biogas plants. The control algorithm relies on a detailed biogas plant model named the Anaerobic Digestion Model No.1 (ADM1). Since ADM1 has a large number of parameters and states that hinder its use as a predictive model, a reduced model is considered as a reasonable alternative. Meanwhile, to solve the problem that many state variables are unmeasurable, a Unscented Kalman Filter (UKF) is adopted to estimate the system parameters. The NMPC algorithm is developed to find optimal and constant substrate mixtures for long-term optimal steady-state operation while achieving a high production of biogas. The simulation results show that the proposed control scheme is able to reduce the effect of inhibition to maintain the anaerobic digestion system working efficiently and to make effluents of biogas plants satisfied.