考虑模糊和随机变量的混合算法的程序实现.docVIP

考虑模糊和随机变量的混合算法的程序实现 张土乔，邵煜* (浙江大学 土木工程学系，浙江 杭州 310027) 摘 要：针对模糊参数和随机参数同时存在的情况，运用蒙特卡罗随机模拟产生随机样本，结合模糊参数的可能性分布，采用修正进化策略的-截集优化算法进行模糊分析，获得模糊结果的集合，最终通过证据理论对模糊结果集合进行后处理，获得边界概率累积分布。通过VBA自定义函数，有机结合基于Excel的蒙特卡罗模拟工具Crystal Ball（CB）和Fortran模糊分析库，从而实现模型中同时含模糊和随机参数的不确定分析。计算结果表明该方法能够充分利用已有成熟软件具有高效性和稳定性的特点，有效结合随机分析和模糊分析，实现混合计算，从参数到结果的不确定性传递过程中人工误差较少。 关键词：模糊；随机；-截集优化；混合算法；程序实现； 中图分类号：TU435 文献标识码：A 文章编号： Program Realization for Hybrid Algorithm under Joint Fuzzy and Random Uncertainty Zhang Tu-qiao, Shao Yu* (Department of Civil Engineering, Zhejiang University, Hangzhou 310027, China) Abstract: In the case of both fuzzy and random parameters occur together, combining Monte Carlo random sampling of probability distribution functions with fuzzy possibility distribution of fuzzy parameter, the fuzzy results set was obtained, by using -level optimization method with modified evolution strategy(MES). The upper and lower probability distributions were derived from post-processing the fuzzy results set based on evidence theory. To solve the uncertainty analysis under joint fuzzy and random uncertainty, we combined the Monte Carlo simulation software Crystal Ball(CB) which based on the Microsoft Excel with the Fuzzy Analysis Library which was written in Fortran by developing the user-defined function in VBA. The results show that: This method makes use of the high efficiency and stability of the mature software which were independently used in fuzzy analysis and random simulation, and combines them to implement the hybrid algorithm; there are lower artificial errors when propagating uncertainty involving imprecision and randomness. Key words: Fuzzy; Random; -level optimization; Hybrid Algorithm；Program Realization； 客观世界现象总是存在各种不确定性，主要表现为两个方面：随机性和模糊性cit_bf[1]cit_af ref_bf(M?ller, 2004 ref_num1602)ref_af。随机性可由概率统计规律来描述，模糊性通常由边界不确定性引起，用隶属函数表示。当模糊和随机参数同时存在时，混合方法cit_bf[2]cit_af ref_bf(Guyonnet, 2003 ref_num1569)ref_af可同时考虑参数的模糊性和随机性，但是文献[2]对实现该混合方法的具体细节如蒙特卡罗模拟和模糊运算并没有详细交代，而蒙