• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Energy Conversion & Management >An energy estimation framework for event-based methods in Non-Intrusive Load Monitoring
【24h】

An energy estimation framework for event-based methods in Non-Intrusive Load Monitoring

机译：非侵入式负载监控中基于事件的方法的能量估计框架

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Non-Intrusive Load Monitoring (NILM) is a set of techniques used to estimate the electricity consumed by individual appliances in a building from measurements of the total electrical consumption. Most commonly, NILM works by first attributing any significant change in the total power consumption (also known as an event) to a specific load and subsequently using these attributions (i.e. the labels for the events) to estimate energy for each load. For this last step, most published work in the field makes simplifying assumptions to make the problem more tractable. In this paper, we present a framework for creating appliance models based on classification labels and aggregate power measurements that can help to relax many of these assumptions. Our framework automatically builds models for appliances to perform energy estimation. The model relies on feature extraction, clustering via affinity propagation, perturbation of extracted states to ensure that they mimic appliance behavior, creation of finite state models, correction of any errors in classification that might violate the model, and estimation of energy based on corrected labels. We evaluate our framework on 3 houses from standard datasets in the field and show that the framework can learn data-driven models based on event labels and use that to estimate energy with lower error margins (e.g., 1.1-42.3%) than when using the heuristic models used by others.
机译：非侵入式负载监控（NILM）是一套技术，用于通过测量总用电量来估算建筑物中各个设备的用电量。最常见的情况是，NILM首先将总功耗的任何重大变化（也称为事件）归因于特定负载，然后使用这些归因（即事件的标签）来估算每个负载的能量。对于最后一步，该领域中大多数已发表的工作都简化了假设，使问题更易于处理。在本文中，我们提供了一个用于基于分类标签和聚合功率测量值创建设备模型的框架，该框架可以帮助放松许多假设。我们的框架会自动构建用于家用电器的模型以执行能量估算。该模型依赖于特征提取，通过亲和力传播进行聚类，提取状态的扰动以确保它们模仿设备行为，创建有限状态模型，校正可能违反该模型的分类错误以及根据校正后的标签估算能量。我们从该领域的标准数据集中评估了3栋房屋的框架，并表明该框架可以基于事件标签学习数据驱动的模型，并使用该模型来估计能量，其误差幅度（例如1.1-42.3％）比使用事件时要低。其他人使用的启发式模型。

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