Discrimination of environmental contaminants using machine learning

Machine learning algorithms; It is used for many different purposes, including the most efficient computational remediation of environmental pollutants, planning of effluent synthesis, and modeling of pollutant removal processes. In this study, it is aimed to determine how the chemical ratios used to create clean water can be used at a minimum level before filtering processes in drinking water treatment plants, by using machine learning methods, in the most accurate way. Before the filtering process, the color and turbidity values of the liquid are measured and it is decided how much of the chemicals to be used in the filtering process of the liquids will be used, if needed. The turbidity sensor and the color sensor used in the method measure simultaneously. In order for the values taken from the sensors to be correct, the measurement was made with pure water as the first reference value. While taking the reference value, the contaminated liquid; fuzzy and color values are read by sensors. Machine learning of fuzzy and color values read from sensors; It is created with linear regression, mean and standard deviation models, and the computer interface provides the user with the best indicator of live tracking and decision-making situations, and the value is produced and displayed. In this way, the pollutant particles in the contaminated liquid are detected and the chemical dosing to be used in the process of removing the contaminants from the liquid is decided by looking at the turbid and color values. The main aim of the research is to solve the problems related to the removal of liquid contaminants by machine learning models; To meet the needs of producing solutions in the most practical way, with a fast, low cost, and to provide new information on water quality assessment, risk warning and reduction of chemical dosing in the operation of water treatment plants.