摘要
传统的分布式能源运行在线监测方法对于高维非线性数据,存在判别数据能力较差,采集时容易出现数据混淆现象。为此,设计多状态下分布式能源运行在线监测方法。利用Labview提供的选项卡控件采集分布式能源运行数据,通过构造的最优分类面分类处理运行数据样本,确定合适的核函数,使线性决策函数保持稳定,分类完成后,通过选项卡控制分布式能源运行参数监测界面完成监测。测试结果表明:对于高维非线性数据,与传统的监测方法相比,该监测方法不同的数据区域之间没有出现交集,未发生数据混淆现象,数据判别能力较强,适合应用在分布式能源运行在线监测中。
For high-dimensional nonlinear data,the traditional online monitoring method of distributed energy operation has poor discriminant data ability,and it is easy to be confused when collecting data.Therefore,the online monitoring method of distributed energy operation under multi-state is designed.The tabbed control provided by Labview is used to collect the distributed energy operation data,and the optimal classification surface is constructed to process the running data samples,and the appropriate kernel function is determined to keep the linear decision function stable.After the classification is completed,the distributed energy operation parameter monitoring interface is controlled by the tab to complete the monitoring.The test results show that for high dimensional nonlinear data,compared with the traditional monitoring,Compared with the method,there is no intersection between the different data regions of the monitoring method,no data confusion occurs,and the data discrimination ability is strong,so it is suitable for online monitoring of distributed energy operation.
作者
乔连留
白建林
张碧华
张春辉
Qiao Lianliu;Bai Jianlin;Zhang Bihua;Zhang Chunhui(Yuxi Power Supply Bureau of Yunnan Power Grid Co.,Ltd,Yuxi Yunnan 653100,China)
出处
《科技通报》
2021年第9期56-59,共4页
Bulletin of Science and Technology
关键词
多状态
分布式能源
聚类
在线监测
multimode
distributed energy
clustering
On-line monitoring
