Using the bees algorithm for artificial neural networks training and the contol application of wastewater treatment plant

The framework of this study may be drawn as, finding a solution for the ever increasing environmental problem of wastewater treatment by using artificial intelligence approach. After a general introduction in the first section, the description and aspects of wastewaters, environmental protection and laws regulating the treatment of wastewaters summarized. Karaman wastewater treatment plant have been introduced. In the advancing sections a general overview of Artificial Neural Networks, Genetic Algorithms and bees Algorithm have been done. The main purpose of starting this study is requirement of a system that predicts output values according to input parameters due to some reasons like pollution measurement process? long duration time, cost and sometimes urgency of evaluating the process outcomes. Furthermore one of the main problems in wastewater treatment plants is instant excessive pollution which reachs up in short period and can?t be seen in measurements. Another subject of this study is instant extra pollution determination and possiblity of taking precaution in extreme bias situation by matching the result of plant control tests and prediction datas based on past datas In the application section, wastewater databases of University of California Irvine and ADASU Karaman wastewater treatment plant of Adapazarı Great Municipality have been used to train neural network back propogation model. They are also used in genetic algorithm and bees algorithm and the test results were obtained. Finaly, the succesful application of the bees algorithm in the training of the artificial neural network have been done. Also, it has also been shown that, the acquired model can succesfuly be used in the control of wastewater treatmen plant. The possible, further search which may be done by employing the bees algorithm have been evaluated within this context