For the operation of many drinking water treatment processes influences of raw water quality and operational settings on process performance are unknown. Therefore black box models such as neural networks are a promising way to model drinking water treatment processes. The combination of neural networks with genetic algorithms also enables fast process optimization. The application of neural networks and genetic algorithms in drinking water treatment will be shown for a ceramic membrane microfiltration plant. First, neural networks were applied for prediction of the course of transmembrane pressure (TMP) over several cycles with high precision. In a second step these models were applied for operational costs optimization by genetic algorithms. Based on Darwin's idea of the survival of the fittest, settings for filtration time, flux and aluminium dosage were optimized, leading to minimized operational costs with a costs reduction of about 15%. The study proved the effectiveness of genetic algorithms and the applicability for online optimization being planned for further studies.

You do not currently have access to this content.