Data Based Modeling of a Wastewater Treatment Plant by using Machine Learning Methods

Abstract

It is observed through studies and recent data, energy consumption increases proportionally to the electricity costs. In order to reduce the costs, energy consumption needs to be analyzed and one way to reduce the energy consumption is by implementing the energy efficiency. In this study, the focus is on wastewater treatment plant where energy efficiency is implemented to reduce the costs so that organized industrial water supply and sewerage can be brought to a larger population in the developing countries. If clean water and wastewater treatment can be offered more affordably in less developed regions, the quality of living be generally improved and thus, can save lives. There are few ways to implement energy efficiency such as optimizing the machine and equipment that used lots of energy or predict the energy consumption by using data based modeling which will be done in this study. The main objective of this study is to obtain a suitable method that can predict energy consumption with least error and provides maximum amount of cost saving. By predicting the energy output, the energy consumption for the next month can be determined and in that way, the energy consumption can be reduced. This paper presents data based modeling by using machine learning methods to predict the energy consumption. From the analysis, it is found that the prediction by using artificial neural network is better than other methods as it produces least error with root mean square error is 52084. By using the energy prediction results that were obtained through neural network, we can save as much as RM156499 which is 2.23 percent from current usage.