Artificial Neural Networks in Water Supply Engineering

Artificial Neural Network (ANN) technology is rarely used in water supply engineering. Those who have applied this technology for water supply engineering problems have reported findings that were beyond the capability of traditional statistical/mathematical modeling tools. The purpose of this task committee report is to compile the experiences of those who have successfully applied ANNs for water supply engineering problems and to reduce the perceived mystery of the ANN models. The availability of several diverse applications, along with the basics of neural network modeling, in a concise report are expected to encourage the use of this powerful technology. The report comprises several chapters summarizing experiences of individual groups of researchers around the world who demonstrated significant benefits of using neural network technology for diverse applications in water supply engineering. Some of the applications include: forecasting salinity levels in River Murray, South Australia; predicting gastroenteritis rates and waterborne outbreaks; modeling pH levels in a eutrophic Middle Loire River, France; and ANNs as function approximation tools replacing rigorous mathematical simulation models for analyzing water distribution networks.