Deep convolutional neural network with sine cosine algorithm based wastewater treatment systems

Area of Science:

• Environmental Engineering
• Artificial Intelligence
• Water Resource Management

Background:

• Wastewater treatment systems (WWTS) face increasing environmental regulations and the need for better discharge management.
• Existing systems require advanced control and information management for residual fluid handling.
• Intelligent technologies are crucial for analyzing data and forecasting process behavior in WWTS.

Purpose of the Study:

• To develop and validate an intelligent system for predicting Chemical Oxygen Demand (COD) in wastewater treatment.
• To enhance the precision of wastewater treatment models through data analysis and forecasting.
• To optimize WWTS operations for improved environmental compliance and efficiency.

Main Methods:

• Incorporation of industrial data into a wastewater treatment model.
• Application of Deep Convolutional Neural Network (DCNN) and Since Cosine Algorithm (SCA) for prediction.
• Utilizing the DCNN-SCA model for process optimization and performance enhancement.

Main Results:

• The DCNN-SCA model accurately estimated Chemical Oxygen Demand (COD) in both influent and effluent.
• Experimental validation demonstrated superior predictive performance compared to existing techniques.
• The DCNN-SCA-WWTS model achieved high precision (97.63%), recall (96.37%), F-score (95.31%), and accuracy (96.27%) with low RMSE (27.55%) and MAPE (20.97%).

Conclusions:

• The DCNN-SCA model offers a robust and accurate solution for predicting wastewater properties like COD.
• Intelligent technologies, specifically DCNN and SCA, significantly improve the predictive capabilities of WWTS.
• This approach enhances operational efficiency and supports adherence to environmental regulations in water treatment facilities.