【摘要】 当前电力设备状态参量规模逐渐增加,电力设备状态数据来自多个不同系统,较为复杂。传统诊断方法不能有效处理大规模数据,导致电力设备局部放电诊断结果不可靠。为此,提出一种新的基于大数据分析挖掘技术的电力设备局部放电诊断方法。给出谱图生成过程,对电力设备局部放电特征进行提取。对电力设备状态参量进行大数据分析挖掘,完成对电力设备各种状态参量的组合、特征合并处理。通过皮尔逊相关性理论计算相关系数。依据相关系数,利用优化迭代将电力设备样本分成若干类,获取对应聚类中心。将局部放电样本聚集在一起,依据局部放电特征实现诊断。实验结果表明,所提方法诊断可靠性与实用性强。更多还原

【Abstract】 At present,the state parameters of power equipment gradually increase in size,and the power equipment status data comes from many different systems,which is more complicated. Traditional diagnostic methods can not effectively deal with large-scale data,resulting in partial discharge diagnosis of electrical equipment is not reliable. Therefore,a new partial discharge diagnosis method for power equipment based on big data analysis and mining technology is proposed. The spectrum generation process is given,and the partial discharge characteristics of power equipment are extracted. The state parameters of power equipment are analyzed by big data mining,the combination of various state parameters of power equipment and the feature merging are completed,and the correlation coefficient is calculated by Pearson correlation theory. According to the correlation coefficient,the power equipment samples are divided into several classes by using the optimized iteration,the corresponding cluster centers are obtained,the partial discharge samples are gathered together,and the diagnosis is realized according to the partial discharge characteristics. The experimental results show that the proposed method has high reliability and practicability.更多还原