基于CNN的轻量级神经网络单幅图像超分辨率研究摘要：筛选智慧小区评估参量时，针对经济成本相关属性挖掘效果差、小^供电负荷仍处于较高区间的问题，提出了基于多属性决策的智慧小区等级多参量评估方法。以电气智能化建设资源作为智慧小区重要组成要素，获取包含经济成本、舒适指数、自动管理、智慧指数等属性参量。采用层次分析法和因子分析法，结合层次权重和因子权重，计算参量权重，赋予分值给不同等级的参量，计算智慧小区最终评估值，划分评估等级。选取某以电气建设为重点改造项目的智慧小区，设置对比实验，对电网供电功率、光伏电源供电功率、小区日负荷的评估结果进行对比，结果表明，基于多属性决策的智慧小区等级多参量评估方法与实际数据的拟合程度最高，提高了参量评估结果的参考价值。关键词：属性参量;参量权重;智慧小区等级;电气建设;评估等级：G267文献标识码：AMultiparameterEvaluationMethodofSmartCommunityGradeBasedonMultiattributeDecisionMakingYANGLiu，XUWeichao，XIAOChang，LI激nsheng（GuangxiInstituteofBuildingResearchDesign，Nanning，Guangxi530005，China）Abstract：Whenselectingtheevaluationparametersofsmartcommunity，aimingattheproblemthattheminingeffectofeconomiccostrel第1页共2页atedattributesispoorandthepowersupplyloadofthecommunityisstillinahighrange，amultiattributedecisionmakingbasedmultiparameterevaluationmethodforsmartcommunitygradeisproposed.Takingelectricalintelligentconstructionresourcesasanimportantcomponentofsmartcommunity，theattributeparametersincludingeconomiccost，comfortindex，automaticmanagementandsmartindexareobtained.AHP（AnalyticHierarchyProcess）andfactoranalysismethodareusedtocalculatetheparameterweightcombinedwithhierarchicalweightandfactorweight，andthescoreisgiventodifferentlevelsofparameters.Thefinalevaluationvalueofsmartcommunityiscalculatedandtheevaluationlevelisdivided.Selectasmartcommunitywithelectricalconstructionasthekeytransformationproject，setupacomparativeexperiment，comparedwiththeevaluationresultsofgridpowersupply，photovoltaicpowersupplyanddailyloadofthecommunity，theresultsshowthat，theintelligentcommunitygrademultiparameterevaluationmethodbasedonmultiattributedecisionmakinghasthehighestdegreeoffittingwiththeactualdata，whichimprovesthereferencevalueoftheparameterevaluationresults.第2页共2页