基于免疫遗传算法水资源配置多阶段群决策优化模型研究  

Optimization of Multi-Stage Group Decision Making for Water Resource Allocation

作　　者：娄帅[1,2] 王慧敏[1,2] 牛文娟[1,2] 许叶军[1,2] 

机构地区：[1]河海大学水文水资源与水利工程科学国家重点实验室,南京210098 [2]河海大学管理科学研究所,南京210098

出　　处：《资源科学》2013年第3期569-577,共9页Resources Science

基　　金：国家社会科学重大基金(编号:12&ZD214);国家社会科学基金(编号:10AJY005;10CGL051);国家自然科学基金(编号:50979024);国家重点基础研究发展计划(973计划)(编号:2010CB951104);2011江苏省研究生创新基金(编号:CXLX11_0463)

摘　　要：研究水资源系统决策特点,构建多阶段水资源配置群决策模型。在水资源配置决策方案选择时,将水资源配置模式分为经济效益、社会效益、生态环境效益和资源利用效率等四种模式,并制定模式对应决策方案。分析水资源配置参与主体特殊性,通过主观权重和客观权重的划分确定专家群体动态权重。在群决策过程中,专家群体不断协商和信息共享,加深自身知识得出最优专家权重,并在协商过程中引入调整因子变量,通过专家评价矩阵调整因子的改变实现专家群体意见收敛。然后结合WAA(Weighted Arithmetic Average)算子和免疫遗传算法对专家决策矩阵进行拟合,得出最优配置方案,并对多阶段群决策过程进行仿真计算。最后通过漳河流域实例分析其结果的合理性,以及优化模型在综合意见收敛上的有效性。Here, we research decision making characteristics for a water resource system and build a multi-stage group decision making model. We divided the water resource allocation mode into four modes： economic efficiency, social benefits, ecological environmental benefits and resource utilization efficiency. Because different positions and attitudes impact group decision making we divided the objective weight into legal weight and expertise weight. To determine the dynamic weight of expert groups we divided subjective weights and objective weights. Convergence of expert opinion was by altering adjustment factors of the evaluation matrixes of experts. Using weighted arithmetic average operators and immune genetic algorithms we fit the expert group decision making matrix. Using the Zhang River basin we evaluated the results of our model. The expert group contained experts from the central government, upper reaches of Zhang River Bureau,water conservancy officers, and representatives of water users. Our goal in this case study was to verify the effectiveness of the optimization model for the convergence of comprehensive opinion.

关 键 词：水资源配置 免疫遗传算法 群决策 动态权重 模拟仿真 漳河流域 

分 类 号：TV213.4[水利工程—水文学及水资源]