本文已被:浏览 710次 下载 1020次
投稿时间:2019-06-19 修订日期:2019-08-09
投稿时间:2019-06-19 修订日期:2019-08-09
中文摘要: 针对当前智能电表状态评估存在精确度低、泛化性差和实时困难等问题,本文采用泛在电力物联网构建状态实时评估方法解决该问题。首先,采用决策树算法实现智能电表的分类,整体增强方法的匹配度和适应性;随后针对不同类别的智能电表,采用Apriori算法对样本集数据的特征集进行识别和提取,从而降低特征维度并增强关联性;接着,基于决策引擎实现对智能电表状态实时评估,并以度量学习实现新增物联网采集数据的有效性评估,反馈优化传感设备部署,从而根据评估结果实现对新增部署传感器及其位置的调整,进而根据应用场景不断优化智能电表状态实时评估应用模式。实验结果表明,本方法可实现智能电表运行状况的实时、普适、精准运维评估,进一步解决泛在电力物联网设备现场部署经验不足、校验无目标等问题。
Abstract:With the continuous development of the Internet of Things technology, the current power system has accumulated a large amount of heterogeneous data. The smart meter status evaluation method has problems such as poor combination, low generalization, and poor real-time performance. The method for real-time evaluation of universal smart meter status based on ubiquitous power internet of things can effectively solve the above problems. Based on the decision engine, the paper evaluates the state of the smart meter, and The acquisition data feedback is used to optimize the deployment of sensing devices in the ubiquitous power Internet of Things, so as to realize real-time, universal and accurate operation and maintenance evaluation of smart meter operation status. Further solve the problems of insufficient deployment experience of field equipment and verification without targets, and improve the scientific and effective operation of the power grid.
文章编号: 中图分类号: 文献标志码:
基金项目:
作者 | 单位 | |
薛斌* | 国网重庆市电力公司 | 597671431@qq.com |
张向东 | 国网重庆市电力公司 | |
段立 | 国网重庆市电力公司 | |
徐鸿宇 | 国网重庆市电力公司 | |
王刚 | 国网重庆市电力公司 | |
赵莉 | 国网重庆市电力公司 |
引用文本: