摘要
在总结了变电站巡视周期影响因素的基础上,提出了基于竞争神经网络的变电站巡视周期分类方法,根据变电站的电压等级、重要程度、历史故障/缺陷发生频次、设备运行情况等属性进行巡视周期聚类。使用某地区的变电站数据进行了仿真分析,结果表明该方法可以利用机器学习实现变电站巡视周期的科学合理分类。
On the basis of summing up the influencing factors of substation patrol cycle, a classification method of substation patrol cycle based on competitive neural network is proposed. The patrol cycle clustering is carried out according to the voltage level,importance, frequency of historical faults/defects, equipment operation situation and other attributes of the substation. A simulation analysis is carried out by using the real data of several substations in a certain area. The results show that the method can, through machine learning, realize the scientific and reasonable classification of substation patrol cycle.
出处
《科技创新与应用》
2020年第18期31-32,35,共3页
Technology Innovation and Application
关键词
变电站
巡视周期
竞争神经网络
聚类
substation
patrol cycle
competitive neural network
clustering