Related Experiment Video
Updated: May 2, 2026
09:44
Methods to Test Visual Attention Online
Published on: February 19, 2015
11.8K
Online soft measurement method for chemical oxygen demand based on CNN-BiLSTM-Attention algorithm
Libo Liu1, Xueyong Tian1, Yongguang Ma1
1School of Environmental and Chemical Engineering, Shenyang University of Technology, Shenyang, China.
Plos One
|June 28, 2024
Summary
This study introduces an advanced CNN-BiLSTM-Attention model for real-time chemical oxygen demand (COD) monitoring in wastewater treatment. The method offers accurate COD prediction, reducing costs and labor associated with traditional measurements.
Area of Science:
- Environmental Engineering
- Water Treatment Technologies
- Artificial Intelligence in Environmental Monitoring
Background:
- Accurate chemical oxygen demand (COD) measurement is crucial for effective sewage treatment, but traditional methods are costly and labor-intensive.
- Online monitoring of COD is needed to track treatment efficiency and trends in real-time.
- Existing methods lack the precision and efficiency required for continuous wastewater management.
Purpose of the Study:
- To develop an accurate and efficient online soft measurement method for COD prediction in wastewater treatment.
- To leverage deep learning techniques for improved COD monitoring in the aerobic stage of A2O processes.
- To reduce the economic and labor costs associated with traditional COD analysis.
Main Methods:
- A Convolutional Neural Network-Bidirectional Long Short-Term Memory Network-Attention Mechanism (CNN-BiLSTM-Attention) model was developed for COD soft measurement.
- Input variables including pH, dissolved oxygen (DO), electrical conductivity (EC), and water temperature (T) were selected based on correlation analysis.
- The model's performance was evaluated using RMSE, MAE, MAPE, and R2, and compared against other machine learning algorithms.
Main Results:
- The CNN-BiLSTM-Attention model demonstrated high accuracy in predicting COD values.
- The proposed model outperformed several benchmark algorithms, including CNN-LSTM-Attention, CNN-BiLSTM, and traditional methods like SVM and XGBoost.
- Wilcoxon signed-rank tests confirmed statistically significant improvements of the CNN-BiLSTM-Attention model over others.
Conclusions:
- The CNN-BiLSTM-Attention algorithm provides a superior and accurate approach for online soft measurement of COD in wastewater treatment.
- This AI-driven method offers a cost-effective and labor-saving alternative to conventional COD testing.
- The model's ability to capture complex temporal dependencies makes it highly effective for real-time environmental monitoring.
