摘要
针对当前数据匹配方法不能满足电力系统数据通用性的问题,考虑电气量的属性相似度建立了基于反馈机制的多指标相似度模型。引入基于余弦算法的文本相似度计算方法,对其根据电力系统特征词的语义模型进行改进。采用分层逐级匹配和基于拓扑的数据匹配策略,给出了计算文本相似度与属性相似度权重的动态计算方法。提出的方法已经成功应用于某省级电网公司整定计算专家系统,以及能量管理系统和故障信息系统与整定计算系统的数据匹配,验证了该方法的通用性与准确性。
Due to the discordance of the current data matching method with power system data generality,multi-index similarity model based on feedback mechanism has been built considering similarity of electrical quantity attribute.Semantic model based on power system key word has been improved by text similarity calculation method based on cosine algorithm.Dynamic calculation method of text similarity weight and attribute similarity weight is provided by using lay-dividing matching method and data matching strategy on topology.This method has been applied in setting relay protection expert system,and data matching between EMS,fault information system and calculation system in a provincial grid company.The verified generality and accuracy of this method prove it effective in data integration of heterogeneous information system.
出处
《电力与能源》
2015年第5期612-617,共6页
Power & Energy
