| 应用DBN深度学习算法的电能计量反窃电技术研究* |
投稿时间：2020-04-11 修订日期：2020-06-10  点此下载全文 |
| 引用本文： |
| 摘要点击次数: 187 |
| 全文下载次数: 0 |
|
| 基金项目:河北省国网电力公司科技项目 |
|
| 中文摘要:在电力技术发展的形式下,窃电是阻碍建立公平、合理的用户秩序的障碍。区别于常规技术,基于云计算的智能电网大数据处理平台 (smart power system big data processing platform in cloud environment,SP-DPP)提出了融合自适应加权融合算法和深度置信网络(Deep Belief Networks,DBN)学习算法的反窃电系统,实现了海量窃电信息的存储、管理和应用,并对海量的数据信息进行融合处理,实现多种传感器数据的集中处理,通过样本数据训练,采用DBN逐层贪婪训练算法,利用双层RBM结构,构建出DBN深度学习算法,该算法能够对获取的电能计量窃电信息进行归一化处理,将获取的宏观高纬度数据信息转换为容易识别和计算的低纬度数据,大大提高了数据计算的能力,有利于用户快速获取电能计量窃电信息。试验表明,本研究的算法识别率高,稳定性能好。 |
| 中文关键词:窃电 SP-DPP 自适应加权融合算法 深度置信网络 逐层贪婪训练算法 |
| |
| Realizing the analysis of the influence of measurement error of intelligent electric energy meter in harmonic environment |
|
|
| Abstract:In the form of power technology development, theft of electricity is an obstacle to the establishment of a fair and reasonable user order. Different from conventional technologies, the cloud-based smart grid big data processing platform SP-DPP (smart power system big data processing platform in cloud environment) proposes an anti-theft system that combines adaptive weighted fusion algorithms and DBN deep learning algorithms to achieve The storage, management and application of massive electricity stealing information, and the fusion of massive data information to achieve the centralized processing of multiple sensor data, through sample data training, using DBN layer-by-layer greedy training algorithm, using double-layer RBM structure Constructed a DBN deep learning algorithm, which can normalize the acquired electricity metering information and convert the acquired macro high-latitude data information into low-latitude data that is easy to identify and calculate, greatly improving the data calculation ability It is helpful for users to quickly obtain energy metering information. Experiments show that the algorithm of this study has high recognition rate and good stability. |
| keywords:Electricity theft SP-DPP adaptive weighted fusion algorithm deep confidence network layer-by-layer greedy training algorithm |
| 查看全文 查看/发表评论 下载pdf阅读器 |
|
|
|
位访问者,今日访问
