مهندسی منابع آب (Apr 2015)
Effectiveness of the Correlation Inputs in Flooding Control System by Uncertainty Analysis
Abstract
Uncertainty analysis is an essential part of every project dealing with uncertain phenomena, therefore the policy –makers might face unexpected happenings that may delay, or at the extremes, cause complete failure of any projects. Uncertainty analysis is usually preformed in the framework of risk management. The Monte Carlo simulation method is used for integrated and simultaneous analysis of different combinations of uncertainties. This method is a powerful tool to study the various states of the outcomes of a system due to the uncertainties that have significant advantages such as considering all the uncertainties and the ability to show different aspects of the utility function. Existence of dependent random variables is quite common and likely in natural and engineered systems. In order to evaluate the need for correlation actions uncertainty, the design of flood control systems in terms of hydrological, hydraulic and economic uncertainties is discussed. In particular, design of levees that are one of the most common structural flood control, has been studied. In the developed nonlinear optimization model, uncertainty once without correlation and again by considering correlation was achieved. Models have been solved by LINGO-13 software. In the model without correlation, the height and other dimensions of levee is less than what the model with correlation predicts. The results indicated that ignoring the type and level of correlations between each uncertainty, combinations of the outputs of the system will provide results that are impossible in the real world and are somewhat far from reality.
