摘要
电力营销中传统大数据采集系统无法对输入信息实现全面的过滤,导致采集效率不高,数据吞吐量较小。为此,设计基于机器学习的电力营销大数据采集系统。通过建立电力营销数据采集架构,以组件技术灵活控制数据处理;配置数据互通互联多元化接口,消除电力信息孤岛,实现数据共享;设计驱动放大器反转电路,抵消共模电压变化。通过机器学习算法定义大数据分类基本任务,利用卷积运算过滤数据特征偏差,数据流算法给定顺序采集电力营销数据,完成大数据采集系统设计。实验测试结果表明,所提系统的数据采集效率更高,数据吞吐量更大,说明所提系统更加有效。
Traditional big data acquisition systems in power marketing cannot completely filter the input information,resulting in low acquisition efficiency and low data throughput.Therefore,a power marketing big data collection system based on machine learning is designed.By establishing the data acquisition system structure of power marketing,this paper can flexibly control data processing by component technology,configure diversified interfaces for data interconnection,eliminate isolated islands of power information and realize data sharing.The inverter circuit of the driver is designed to counteract the change of common mode voltage.The machine learning algorithm is used to define the basic task of big data classification,convolution operation is used to filter the characteristic deviation of data,and data flow algorithm is used to collect power marketing data in a given order,thus completing the design of big data acquisition system.The experimental results show that the data acquisition efficiency of this system is higher,and the data throughput is larger,which indicates that this system is more effective.
作者
左进
ZUO Jin(Yinchuan Power Supply Company of State Grid,Yinchuan 750011,China)
出处
《微型电脑应用》
2024年第9期107-110,共4页
Microcomputer Applications
基金
国网宁夏电力有限公司科技项目(20220120X)。
关键词
机器学习
电力营销
大数据
采集系统
卷积运算
machine learning
power marketing
big data
acquisition system
convolution operation
