Aeration Volume Prediction in Wastewater Treatment Based on Multi-Model Ensemble and Auto-Weighted Fusion

Accurate aeration volume prediction is critical for enhancing energy efficiency in wastewater treatment systems, yet traditional models struggle with complex nonlinear dynamics. This paper proposes a heterogeneous multi-model ensemble framework combining LSTM, MLP, SVR, Random Forest, and others, fused via enhanced stacking and an automatic weighted blending mechanism. Experimental results on real-world data show the proposed AutoBlend model outperforms all individual models, achieving the lowest MAE (0.0100), MSE (0.0002), and highest R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> (0.9957). These results validate the robustness and accuracy of the approach in handling dynamic and nonstationary data in wastewater treatment.