丁迅, 张忠, 夏兆俊, 范洋洋, 张颖, 孔亮. 基于非侵入式负荷监测的家庭智慧用能管理研究[J]. 现代电力, 2022, 39(4): 496-504. DOI: 10.19725/j.cnki.1007-2322.2021.0140
引用本文: 丁迅, 张忠, 夏兆俊, 范洋洋, 张颖, 孔亮. 基于非侵入式负荷监测的家庭智慧用能管理研究[J]. 现代电力, 2022, 39(4): 496-504. DOI: 10.19725/j.cnki.1007-2322.2021.0140
DING Xun, ZHANG Zhong, XIA Zhaojun, FAN Yangyang, ZHANG Ying, KONG Liang. Research on the Home Intelligent Energy Management System Based on Noninvasive Load Monitoring[J]. Modern Electric Power, 2022, 39(4): 496-504. DOI: 10.19725/j.cnki.1007-2322.2021.0140
Citation: DING Xun, ZHANG Zhong, XIA Zhaojun, FAN Yangyang, ZHANG Ying, KONG Liang. Research on the Home Intelligent Energy Management System Based on Noninvasive Load Monitoring[J]. Modern Electric Power, 2022, 39(4): 496-504. DOI: 10.19725/j.cnki.1007-2322.2021.0140

基于非侵入式负荷监测的家庭智慧用能管理研究

Research on the Home Intelligent Energy Management System Based on Noninvasive Load Monitoring

  • 摘要: 随着多能源网络的融合和能源互联网技术的快速发展,家庭用能管理在解决各个能源网络节点供需问题上扮演着重要的角色。现有的大多数家庭用能管理针对已知用电负荷进行优化,未考虑用电设备类型的多样化和用电设备突增的情形。基于非侵入式负荷监测(Noninvasive Load Monitoring,NILM)算法可以有效获取家庭用电负荷、规律和用电信息,为家庭智慧用能管理提供数据支撑。文中以家庭用电成本、温度、时间、舒适度为目标函数建立家庭智慧用能多目标优化模型,对可控负荷、电动汽车、储能系统进行分析建立数学模型,利用粒子群算法对模型进行求解。仿真结果表明,基于NILM监测算法,考虑用电成本和舒适度家庭用电成本降低至72.5%;当用户可控用电负荷增加时,NILM算法可以实时更新控制策略降低用户用电成本;对不同用户进行多次计算,净成本和计算时间波动较小,证明了算法的合理性、可靠性。

     

    Abstract: With the integration of multi-energy networks and the rapid development of energy Internet technology, household energy management plays an important role in solving the problem of supply and demand of each energy network node. Most of the existing household energy consumption management is optimized for the known power load, while the diversification of the types of electrical equipments and the sudden increase of electrical equipments are not considered. On the basis of the noninvasive load monitoring (abbr. NILM) algorithm, the household load electricity consumption law and information that provide data support for household smart energy management can be effectively obtained. A multi-objective optimization model of smart home energy consumption, in which the household electricity cost, temperature, time and comfort level were taken as objective functions, was established, and the controllable load, EV and energy storage system were analyzed and corresponding mathematical models were proposed and solved by particle swarm algorithm. Simulation results show that based on NILM algorithm and only considering electricity cost and comfort level, the home power utilization cost can reduced by 72.5%. When user-controllable power load increases, the control strategy can be updated by NILM in realtime to decrease user’s electricity utilization cost. Results of multiple calculations for different users show that the net cost and computing time fluctuate slightly, thus the rationality and reliability of the NILM algorithm can meet the requirements of family intelligent energy consumption.

     

/

返回文章
返回