Using shuffled complex evolution to calibrate water distribution network model

Calibration of water distribution network model is of paramount importance for the optimal management of water delivery systems. This includes the determination of network parameters such as pipe roughness coefficients and nodal demands. The parameters are not often exactly known and very much sensitive to the age of the pipe. The calibration is usually accomplished by mimicking the model results to the field conditions. However, it becomes tedious if this is performed manually. In this paper, a population based meta-heuristic evolutionary algorithm, Shuffled Complex Evolution (SCE), is applied to determine the network parameters. Two example problems have been analyzed to demonstrate the robustness of the model. The model results show that SCE is capable in reaching the optimal solution in an effective manner. © 2004 Institution of Engineers, Bangladesh. All rights reserved.