Genetic Algorithms for Reliability-Based Optimization of Water Distribution Systems

The proposed approach links a genetic algorithm as the optimisation tool with the first-order reliability method (FORM) for estimating network capacity reliability with particular reference to the probability of meeting minimum allowable pressure constraints across the network under uncertain critical nodal demands and uncertain pipe roughness conditions. FORM approximates Monte Carlo simulation reliabilities accurately and efficiently and identifies critical node location in the network and the corresponding capacity reliability.