Artificial Neural Networks Model Used for Clear Water Treatment Plants

Artificial neural networks, ANNs, have been widely used especially in the last two decades. ANNs have a wide range of scientific applications, they are used for pattern recognition and forecasting of natural mechanisms. In water treatment, ANNs have enormous potential, especially to support workers in plant operation. In clear water plants are taking large volumes of data, especially information about water quality parameters. Analysis and classification of such information is quite difficult without any software to support the analysis. The relationship between water quality parameters and pattern recognition is also difficult even using statistical methods. The use of ANNs make pattern knowledge easier and pattern recognition of the hidden relationship between water quality parameters. With the last point in perspective, ANN models could be used as a support for workers in operating clear water treatment plants. Plant managers, based on ANN modelling, can take action to improve water treatment operations. ANNs can be used for water quality predictions. With use of ANNs in real time, systems will get more efficient, so reducing operation costs and improving the quality of water produced. In the Puerto Mallarino clear water plant, located in Cali, a series of numerical experiments which used optimal doses of coagulants for water treatment were analyzed. The experiments, in real time, demonstrated the closeness of fit that can be achieved with data sets collected using Artificial Neural Networks (ANNs).