A novel stochastic wastewater quality modeling based on fuzzy techniques

Area of Science:

• Environmental Science and Engineering
• Water Quality Monitoring and Management
• Computational Intelligence in Environmental Modeling

Background:

• Accurate measurement and prediction of wastewater quality parameters are essential for assessing risks to receiving water bodies.
• Existing non-linear modeling approaches for wastewater quality parameters like BOD, COD, and TSS have limitations in prediction accuracy.
• Effective data pre-processing techniques, including outlier identification and smoothing, are critical for robust wastewater quality modeling.

Purpose of the Study:

• To develop and evaluate novel data pre-processing methods for wastewater quality modeling.
• To compare the performance of a new hybrid model against existing linear and non-linear approaches.
• To enhance the prediction accuracy of key wastewater quality parameters: 5-day biochemical oxygen demand (BOD), chemical oxygen demand (COD), and total suspended solids (TSS).

Main Methods:

• Development of a hybrid data processing technique combining autoregressive integrated moving average (ARIMA) models with adaptive neuro fuzzy inference system with fuzzy C-means clustering (ANFIS-FCM).
• Application of the hybrid method for outlier identification and data smoothing prior to modeling.
• Comparison of the hybrid model's performance against selected linear models and previously employed non-linear methods for influent/effluent wastewater parameters.

Main Results:

• The hybrid ARIMA-ANFIS-FCM model demonstrated superior performance in predicting wastewater quality parameters, achieving a high coefficient of determination (R²) of 0.95 for influent BOD.
• Non-linear models showed acceptable predictions, with R² values ranging from 0.8-0.87 for influent and 0.83-0.89 for effluent BOD and TSS.
• The developed hybrid model significantly outperformed other recently developed models in the literature for wastewater quality prediction.

Conclusions:

• The integration of ARIMA and ANFIS-FCM provides an effective hybrid approach for wastewater quality parameter prediction.
• The proposed hybrid model enhances the performance and efficiency of wastewater quality modeling, offering improved accuracy over existing methods.
• This advanced modeling technique is crucial for better management and protection of receiving water bodies from wastewater pollution.