Pezhman Allahbakhshian Farsani, Afshin Honarbakhsh and Seyyed Ali Moasherir
Abstract
The calculation of energy loss in the design of water networks is an important factor, so that the original foundation of almost all the, water network construction design, studies and estimation of loss along the pipe lines has been made by empirical and theoretical formulae such as Darcy-Weisbach, I Iazen-williams and so on. In this research comparison among the loss calculation empirical formulae including Darcy-Weisbach and I Iazen-Williams was made which shows that regression analysis between the actual loss and calculated loss in Darcy-Weisbach formula is having the higher regression coefficient almost (R'=0.80) and on the other hand this comparison was made between the output of artificial neural network and the actual loss , that shows by far has higher regression coefficient almost (IV=0.87). The artificial neural net-works technique was used for prediction of losses and the results showed that the neural network of multi-layers along with one hidden layer tan transfer function is having the significant capability to estimate and prediction of losses in the pipes.