This study presents a novel stochastic simulation–optimization approach for optimum designing of flood control dam through incorporation of various sources of uncertainties. The optimization problem is formulated based on two objective functions, namely, annual cost of dam implementation and dam overtopping probability, as those are the two major concerns in designing flood control dams. The nondominated solutions are obtained through a multi‐objective particle swarm optimization (MOPSO) approach. Results indicate that stochastic sources have a significant impact on Pareto front solutions. The distance index (DI) reveals the rainfall depth (DI = 0.41) as the most significant factor affecting the Pareto front and the hydraulic parameters (DI = 0.02) as the least. The dam overtopping probability is found to have a higher sensitivity to the variability of stochastic sources compared to annual cost of dam implementation. The values of interquartile range (IQR) indicate that the dam overtopping probability is least uncertain when all stochastic sources are considered (IQR = 0.25%). The minimum annual cost of dam implementation (2.79 M$) is also achieved when all stochastic sources are considered in optimization process. The results indicate the potential of the proposed method to be used for better designing of flood control dam through incorporation of all sources of uncertainty.

中文翻译：

---

防洪坝随机设计的新型仿真优化策略-以贾米山大坝为例

这项研究提出了一种新颖的随机模拟-通过结合各种不确定性因素来优化防洪大坝设计的方法。最优化问题是基于两个目标函数制定的，即大坝实施的年度成本和大坝的超越概率，这是设计防洪大坝的两个主要问题。非支配解是通过多目标粒子群优化（MOPSO）方法获得的。结果表明，随机来源对Pareto前沿解决方案有重大影响。距离指数（DI）显示降雨深度（DI = 0.41）是影响帕累托锋和水力参数（DI）的最重要因素。= 0.02）。与大坝实施的年度成本相比，发现大坝超标概率对随机源的变异性具有更高的敏感性。四分位间距（IQR）的值表明，当考虑所有随机源时（IQR = 0.25％），大坝超车概率的不确定性最小。当在优化过程中考虑所有随机来源时，还可以实现最低的大坝实施年度成本（279万美元）。结果表明，通过结合所有不确定性因素，该方法可用于更好地设计防洪大坝。