基于知识图谱的智能电网多维数据关联挖掘方法

Multidimensional data association mining method for smart grid based on knowledge atlas

为提高智能电网多维数据关联挖掘方法的准确性,提出了基于知识图谱的智能电网多维数据关联挖掘方法。按照动态图谱的框架结构,抽取关键电网数据,再根据数据的清洗与转换原理,完成基于知识图谱的电网数据关联性预测。在此基础上,划分关联区间,根据知识图谱结构,建立多维关联规则,得到挖掘置信度的具体计算数值,完成智能电网多维数据关联挖掘方法的设计与应用。实验结果表明,在知识图谱作用下,电网数据关联规则挖掘准确度在低维度、中维度、高维度三种传输环境下,实验组的关联规则挖掘准确度在0.910以上,而对照组的关联规则挖掘精度均低于0.900,该文具有较高的准确性,更符合实际应用需求。

In order to improve the accuracy of smart grid multidimensional data association mining method,a multidimensional data association mining method for smart grid based on knowledge atlas is proposed.According to the frame structure of dynamic atlas,the key power grid data are extracted,and then according to the data cleaning and conversion principle,the power grid data correlation prediction based on knowledge atlas is completed.On this basis,the association interval is divided,the multi⁃dimensional association rules are established according to the knowledge map structure,the specific calculation value of mining confidence is obtained,and the design and application of multi⁃dimensional data association mining method in smart grid are completed.The experimental results show that under the action of knowledge atlas,the accuracy of power grid data association rule mining is more than 0.910 in the three transmission environments of low dimension,medium dimension and high dimension,while the accuracy of association rule mining in the control group is less than 0.900.This paper has high accuracy and is more in line with the needs of practical application.