MULTIVARIABLE ITERATIVE EXTENDED KALMAN FILTER BASED ADAPTIVE-CONTROL - CASE-STUDY OF SOLID SUBSTRATE FERMENTATION

In solid substrate fermentation (SSF), the on-line measurements of the states of the fermentation like biomass content, dry matter content, and moisture content are not possible. Also, the control of the temperature and the moisture content is critical for optimization of the process. A multivariable adaptive control structure along with state estimation using an iterative extended Kalman filter (IEKF) is proposed for the control. The IEKF uses the measurements of total wet weight and CO2 evolution rate to estimate the states. An autoregressive with exogenous inputs (ARX) model is used as a model of the process relating controlled variables with manipulated variables (dry air flow rate and the water replenishment rate) and forms the basis for the determination of future control inputs. The simulation results show that a better control of the moisture content can be achieved when compared to the single input-single output (SISO) control strategy.