APP下载

一种电网规划方案决策的灵敏度分析算法

2016-07-09司怀伟王清心丁家满

现代电子技术 2016年7期

司怀伟 王清心 丁家满

摘 要: 方案决策中的不确定因素越来越多,对决策结果在不确定环境下的灵敏度分析显得尤为重要。传统灵敏度分析方法常常忽略不确定性或只是简单平均将导致规划方案决策的偏差甚至错误。针对上述问题,提出一种基于概率盒理论的削减分析算法,首先对不确定性变量进行概率盒建模,其次将各指标概率盒输入决策方程计算并得到决策基数,然后依据决策基数对方案中各指标逐一消减,并计算得到其灵敏度。最后将该方法应用于电网规划方案决策中,实验结果表明该方法具有较强的实用性。

关键词: 灵敏度分析; 概率盒; 不确定性分析; 削减法

中图分类号: TN911?34; TP301.6 文献标识码: A 文章编号: 1004?373X(2016)07?0149?05

Abstract: The more uncertainty factors in the scheme decision are particularly important to the sensitivity analysis of the decision results in the uncertain environment. The traditional sensitivity analysis methods often ignore the uncertainty or simply average the deviation error, which results in the deviation of the planning decision or even mistake. For the above problems, a cutting analysis algorithm based on the probability box theory is proposed. The probability box of the uncertainty variables is modeled, and the input decision?making equation of each indicator probability box is calculated to obtain the decision cardinal number, then each indicator in the scheme is orderly cut according to the decision cardinal number to obtain its sensitivity. The method was applied to the decision?making of the power grid planning scheme. The experimental results show this method has strong practicability

Keywords: sensitivity analysis; probability box; uncertainty analysis; cutting method

0 引 言

决策方案中不确定性因素出现的越来越多,不确定性指标的处理与分析显得尤为重要[1]。灵敏度分析是不确定因素环境中多属性决策的必要补充[2],通过灵敏度分析可以较好地识别出关键指标[3]。灵敏度分析是一种根据参数的输入研究输出结果变化的不确定性分析[4]。如果输入参数变化很小,引起输出结果的变化很大,则认为该参数是灵敏的。由此可见,决策方案中某项指标发生变化却忽略该变化,则会造成一定的损失,所以有必要对规划方案进行灵敏度分析来规避方案中可能出现的风险。

许多研究者逐渐发现灵敏度分析的重要性,并且出现许多经典分析方法,如非参数方法[5]、方差分析法[6]以及矩独立分析方法[7]。……

登录APP查看全文