A Machine Learning Approach for Wastewater Treatment using Feedforward Neural Network and Batch Normalization

In present scenario, the freshwater is a contest as the world’s population raises. Most of Earth’s water comes from brackish water and seawater those are the largest sources of water. Water purification is crucial during this water crisis, and can be achieved through desalination and various water treatment techniques, making water safe for human consumption and other uses. Wastewater treatment downsides including with high energy consumption and chemical usage in mixture of water. Limited removal of certain pollutants sludge generation release are also concerns which are included with various pollutants with concern of overfitting leads to the overreliance on a specific treatment. To overcome these issues, the method of (Feedforward Neural Network (FFNN) and Batch Normalization (BN) (FFNN-BN)) is proposed by collecting water pollutants. This process involves with the collection of input from the dataset of Wastewater Treatment Plant (WWTP) and then features are extracted by proposed method FF-BN and classifying extracted features by the SVR handles high-dimensional data with a large number of features. The proposed FFNN-BN produce more results than traditional methods of Advanced Oxidation Processes (AOP) for modeling extreme values such as R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> Mean Absolute Error (MAE), Mean Squared Error and Root Mean Squared Error (RMSE) respectively.