OPTIMIZING URBAN WATER SUPPLY HEADWORKS USING PROBABILISTIC SEARCH METHODS

A realistic optimisation of the operation and planning of urban water supply headwork's system with the Monte Carlo method. A simple case study system consisting of a reservoir which has 3 decision variables to be optimised was studied. Uses of an objective function based an reservoir cost and economic penalties for water shortage. The Generic Algorithm (GA) and the Shuffled Complex Evolution (SCE) methods were investigated. The SCE method did not need tuning, had comparable robustness and required fewer evaluation than GA. Nonetheless, it was concluded that the GA was the preferred method.