Biogas production through anaerobic digestion (AD) presents a sustainable energy alternative with significant potential to reduce global warming. However, AD is a complex, nonlinear, and dynamic process influenced by time-varying parameters and non-stationary disturbances. These challenges, together with the limited availability of reliable online measurements for key concentration variables, hinder effective real-time monitoring. To address these limitations, this study proposes a joint state and parameter estimation approach based on a modified Advanced Monitoring and Control (AMOCO) model with a Particle Swarm Optimization (PSO) - tuned Extended Kalman Filter (EKF), a combination not previously applied to anaerobic digestion processes. The modified AMOCO model, originally developed for control applications, is adapted to better align with both simulated and experimental data. Sensitivity analysis identifies three key parameters whose estimation significantly improves system reconstruction. To further enhance estimation performance, PSO is employed to tune the noise covariance matrices of a discrete EKF. Validation using the Anaerobic Digestion Model No. 1 (ADM1) as a benchmark plant confirms reliable state and parameter estimation and accurate output predictions. Robustness is assessed by applying the EKF tuned for nominal noise to different measurement-noise levels, demonstrating stable performance under moderate noise mismatch and limited degradation under severe mismatch. Results show that the proposed PSO-EKF approach achieves a 70%–80% reduction in augmented state-estimation RMSE compared with a conventionally tuned EKF. The methodology provides a foundation for monitoring and control in AD and has potential for adaptation to other complex, non-linear bioprocesses, thus supporting more sustainable and efficient waste-to-energy systems. • Introduces a novel PSO–tuned Kalman filtering for anaerobic digestion. • Sensitivity analysis identifies critical parameters for improved estimation accuracy. • Demonstrates robustness under varying measurement noise conditions. • Validates the proposed approach using Anaerobic Digestion Model No. 1 as benchmark. • Real-time monitoring supports efficient and sustainable bioenergy generation.
Abera et al. (Sat,) studied this question.