摘要
针对传统证据理论在电能质量综合评估中存在的"全冲突悖论"、"1悖论"等固有弊端,提出一种基于支持概率距离的电能质量综合评估方法。首先将各项电能指标进行等级划分并对数据进行归一化处理;然后运用模糊数学的方法生成关于电能质量等级的基本概率分配矩阵,求取基本概率分配矩阵中的每项电能质量指标的支持概率距离,确定电能质量指标之间的相互支持程度,并以此作为权重将初始电能质量指标生成的基本概率分配矩阵进行二次修正;最后将重新修正后的基本概率分配矩阵采用经典证据理论合成规则进行融合得到电能质量的综合评估结果。实际算例表明,此方法能够有效克服传统证据理论运用在电能质量评估上当权重差别过大时导致评价结果违背实际的局限,同时能够得到更为准确、客观的综合评估结果。
In allusion to inherent defects of traditional evidence theories such as full conflict paradox,1 trust dilemma,zero trust dilemma,and so on in comprehensively evaluating power quality,this paper presents a kind of comprehensive evaluation method for power quality based on support probability distance. This method firstly grades each power quality index and processes data for normalization,then applies fuzzy mathematics in generating basic probability distribution matrix of power quality grade and gets support probability distance of each power quality index in the distribution matrix. Meanwhile,it determines mutual support degrees of power quality index which are taken as weights for secondary amendment for the basic probability distribution matrix of initial power quality index. Finally,it uses classical evidence theory to fuse the revised basic probability distribution matrix so as to get comprehensive evaluation result of power quality. Actual example indicates that this method is feasible to effectively overcome application limitation of traditional evidence theories in evaluating power quality as big differences of weights cause the evaluation result violate reality. In addition,it can get more correct and objective comprehensive evaluation results.
出处
《广东电力》
2017年第9期63-69,共7页
Guangdong Electric Power
基金
国家自然科学基金资助项目(51577073)
