汛限水位实时动态控制的概率多目标优选方法

方案优选对于汛限水位实时动态控制至关重要。可行方案集中的指标权重及各方案目标特征值将直接影响各方案的最终排序,在汛限水位实时动态控制中,二者通常受各种不确定性因素的影响,取值处于一个变化的范围内,而以往的优选方法对此考虑较少。为此,在HydeK.M.等人研究工作的基础上,建立方案概率优选方法,该方法不但能够优选出概率意义下的满意方案,而且能够提供方案的排序可靠度、相应于各排序的概率及方案优选目标值的累积概率分布值,从而为决策者提供更加全面的决策信息支持。实例计算验证了该方法的有效性和可行性。

Probabilistic Multi-objects Optimal Selection for Real Time Dynamic Control of Flood Control Water Level

The alternative optimal selection plays a key role in real-time dynamic control of flood control water level. Ranking alternatives is determined by the criteria weights (CWs) and criteria performance values (PVs). It is possible that, CWs and PVs are influenced by the uncertainty and they often vary in a range, but less contribution has been put on it in the past. Based on the Hyde K.M. research, probabilistic optimal selection method was proposed in this paper, which can select the satisfied alternative based on probability, the probability of alternative obtains each rank and the cumulative frequency distribution for results of alternatives can also be obtained so as to provide more support for decision makers. An example application is used to illustrate the methodology.