Sensitivity-driven control strategy and analysis of operating parameter MLSS in the stacking total nitrogen...

Area of Science:

• Environmental Engineering
• Water Quality Management
• Predictive Modeling

Background:

• Wastewater treatment plant (WWTP) operations are complex due to variable influent and nonlinear processes.
• Accurate wastewater quality modeling is crucial for effective WWTP management and decision-making.

Purpose of the Study:

• To develop a precise predictive model for effluent total nitrogen (TN) in WWTPs.
• To identify key parameters influencing effluent TN prediction.
• To explore MLSS adjustments for optimizing effluent TN.

Main Methods:

• Utilized a stacking ensemble model combining five foundational models.
• Employed SHAP analysis to determine influential prediction parameters.
• Conducted simulations by adjusting MLSS to forecast effluent TN.

Main Results:

• The stacking model achieved a high coefficient of determination (R² = 0.90).
• Key predictors for effluent TN included electricity, Inf_BOD5, Inf_TN, and MLSS.
• Increased MLSS concentration correlated with reduced effluent TN.

Conclusions:

• The developed stacking model enhances effluent TN prediction accuracy.
• MLSS adjustment presents a viable strategy for reducing effluent TN.
• The study provides valuable engineering insights for WWTP optimization.