Regression based state space adaptive model of two-phase anaerobic reactor

• The discrete state space correlation among parameters in anaerobic reactor.
• The relation is updated at every time point, giving it an adaptive feature.
• The proposed algorithm to estimate methane generation from industrial waste.

In this paper, a linear state space model for the two-phase anaerobic reactor system was developed based on historical data. Subsequently, the model was used to predict its future behavior. The state space model developed involved correlation analysis and model development. The model would be updated at every time point when a new data set became available, giving it an “adaptive” feature. The model was then applied to monitor two-phase anaerobic co-digestion of a feed comprising 2 industrial secondary sludges and 2 industrial wastewaters. The case study showed the proposed model was able to provide good predictions of various process parameters. In addition, it also predicted impending process failure and this would have allowed the operator to take necessary measures to prevent or reduce impact of such failure during plant operation.