Endogenous model state and parameter estimation from an extensive batch experiment

Area of Science:

• Environmental microbiology
• Biochemical engineering
• Process modeling and control

Background:

• Understanding endogenous respiration is crucial for optimizing aerobic biodegradation processes.
• Accurate modeling of microbial dynamics, including substrate and biomass changes, is essential for effective wastewater treatment.
• Existing models may require refinement to capture complex behaviors observed in real-world conditions.

Purpose of the Study:

• To investigate endogenous process behavior in aerobic biodegradation through extensive batch experiments.
• To analytically reconstruct states and unknown parameters of a four-compartmental model using experimental data.
• To apply a recursive state estimation procedure, specifically an Extended Kalman filter, for noise handling and model refinement.

Main Methods:

• Conducted extensive batch experiments measuring mixed liquor volatile suspended solids (MLVSS) and respiration rate.
• Employed analytical reconstruction to determine states and parameters within a four-compartmental model.
• Utilized an Extended Kalman filter-based observer for recursive state estimation and noise filtering.

Main Results:

• Successfully reconstructed states and unknown parameters from experimental data.
• The Extended Kalman filter effectively handled data noise properties during state estimation.
• Identified the necessity to modify the initial model structure, particularly the hydrolysis term.

Conclusions:

• The study highlights the importance of model structure validation against experimental data.
• Refinement of the hydrolysis term is critical for accurately representing endogenous process behavior.
• The applied methodology provides a robust approach for analyzing and improving biodegradation models.