Regular articleEstimation of kinetic parameters of an anaerobic digestion model using particle swarm optimization

- Particle swarm optimization is improved by including multiple loops.
- A glucose digestion model is refined by applying more parameters.
- Every parameters are estimated by the improved particle swarm optimization.
- The refined model fits measured data better.

Calibrating parameters of an anaerobic digestion model is often difficult and time consuming. In order to reduce the complexity of tuning a complex anaerobic digestion model, a particle swarm optimization-based smart algorithm was developed to estimate all parameters of an anaerobic digestion model. A glucose anaerobic digestion model was refined and applied to test the feasibility of the smart algorithm. A reactor was continuously fed with glucose until a steady state was achieved. The steady state and a transient state of the reactor were simultaneously included in the smart algorithm. Results shows that the algorithm acceptably estimated activated sludge concentrations and 14 sensitive parameters, though the glucose anaerobic digestion model was complex. The values of most estimated parameters were close to those reported data, while the values of four sensitive parameters deviated a little from reported data. By applying the estimated parameters, the glucose anaerobic digestion mode matched experimental data well. This verifies the applicability of the algorithm as well as the validity of the model structure.