• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>IEEE Sustainable Power and Energy Conference >Power System Disturbance Identification With Missing PMU Data
【24h】

Power System Disturbance Identification With Missing PMU Data

机译：电力系统干扰识别缺失PMU数据

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

With the wide coverage of wide-area measurement systems, considerable real-time monitoring of the entire network has been realized, making it possible to quickly and accurately identify the type of power system disturbance. However, many factors such as loss of synchronization signal and communication protocol error may result in PMU data loss problem, which may affect PMU based disturbance identification. Aiming at this problem, a power system disturbance identification method based on stacked denoising autoencoder (SDAE) and random forest is proposed. The nonlinear mapping relationship between a corrupted sample and a complete sample is established by SDAE. This avoids the limitations of constructing features based on human experience and enables extract high level feature representation that are more robust to samples containing missing data. In addition, a classification method based on random forest with integrated learning ideas is proposed to improve recognition accuracy. Finally, the simulation test is carried out on the IEEE-39 bus system, and compared with the traditional feature extraction method under different data loss levels. The accuracy, rapidity and good generalization ability of the proposed method are verified.
机译：随着广域测量系统的广泛覆盖范围，已经实现了对整个网络的相当大的实时监控，使得可以快速准确地识别电力系统干扰的类型。然而，许多因素，例如同步信号丢失和通信协议错误可能导致PMU数据丢失问题，这可能影响基于PMU的干扰识别。针对这个问题，提出了一种基于堆叠的去噪自身（SDAE）和随机森林的电力系统扰动方法。 SDAE建立了损坏的样本和完整样本之间的非线性映射关系。这避免了基于人类经验构建特征的局限性，并且能够提取对包含缺失数据的样本更强大的高级特征表示。此外，提出了一种基于随机林的分类方法，具有综合学习思路，提高了识别准确性。最后，在IEEE-39总线系统上执行仿真测试，并与不同数据丢失水平下的传统特征提取方法进行比较。验证了所提出的方法的准确性，快速和良好的泛化能力。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号