基于多源异构信息的藻类水华治理群决策方法  

Group decision-making method for algal bloom governance based on multisource heterogeneous information

作　　者：王小艺[1,2,3] 白玉廷 阳译[1,2,3] 赵峙尧 王立 金学波 WANG Xiao-yi;BAI Yu-ting;YANG Yi;ZHAO Zhi-yao;WANG Li;JIN Xue-bo(School of Artificial Intelligence,Beijing Technology and Business University,Beijing 100048,China;Beijing Laboratory for Intelligent Environmental Protection,Beijing Technology and Business University,Beijing 100048,China;State Environmental Protection Key Laboratory of Food Chain Pollution Control,Beijing Technology and Business University,Beijing 100048,China)

机构地区：[1]北京工商大学人工智能学院,北京100048 [2]智慧环保北京实验室,北京100048 [3]国家环境保护食品链污染防治重点实验室,北京100048

出　　处：《安全与环境学报》2022年第3期1575-1584,共10页Journal of Safety and Environment

基　　金：国家社会科学基金项目(19BGL184);北京优秀人才培养资助青年拔尖团队项目(2018000026833TD01);国家重点研发计划课题(2020YFC1606801);北京工商大学优青培育计划项目。

摘　　要：藻类水华污染治理属于一项复杂的环境工程,传统的藻类水华污染治理方案的选择决策通常依赖人为知识经验。为实现决策管理的理性化和自动化,需要充分利用监测信息和领域专家知识等多源信息,并建立自动决策计算机制。为此提出一种基于多源异构信息的群决策方法。首先提出信息采集与决策总体框架,实现对多源异构信息的有效组织,为决策计算提供规范化计算流程;其次提出基于层次聚类的群决策方法,运用聚类与TOPSIS(Technique for Order Preference by Similarity to an Ideal Solution)方法实现不同类型决策数据的综合;最后将群决策方法应用于湖泊藻类水华治理方案的决策问题,通过案例展示了所提方法在自动决策和多源信息利用方面的可行性。A group decision-making method is proposed based on multi-source heterogeneous information,which aims at the rational and automatic management of the algal bloom governance.Firstly,the collection and decision-making framework is built for the multi-source heterogeneous information,which obtains the effective organization and processing of the information.Specifically,in the framework,real-time monitoring data are acquired with the automatic water environment monitoring system and an indicator system of the water environment is established.The monitoring data are used to determine whether the water bloom is breaking out,and to calculate the weights of the water environment attributes in real-time.Meanwhile,experts’resumes are taken into account in the framework.The characteristics of experts are analyzed to select the appropriate experts who are invited to provide the decision-making opinions.Besides,the management objectives are collected from the administrators,which are used to set the attributes of the objectives.Secondly,a group decision-making method is proposed based on hierarchical clustering.The hierarchical clustering algorithm based on the two-tuple linguistic measurement is introduced to gather the multiple experts into different categories.Based on the clustering center,the experts’weights can be calculated.Then multiple decision matrices are integrated into a comprehensive matrix which can determine the positive and negative ideal attributes.The TOPSIS(Technique for Order Preference by Similarity to an Ideal Solution)method is applied to the comprehensive evaluation of decision information.The ranking of advantages and disadvantages of alternatives can be obtained.Finally,the method is applied to the decision-making process of algae bloom governance in the lake.The case demonstrates that the proposed method is feasible in automatic decision-making with multiple-source information.

关 键 词：环境工程学 藻类水华 群决策 治理 信息融合 

分 类 号：X524[环境科学与工程—环境工程]