非侵入负荷分解技术验证平台的研究与应用
|
摘要
非侵入负荷分解技术作为新兴的用户用电行为感知技术,尚缺乏对分解算法进行评估的验证平台。针对上述问题,以非侵入负荷分解算法运行结果与侵入子系统采集的负荷信息进行对比作为核心思想,构建了非侵入负荷分解技术验证平台。首先提出并设计了模块化组合模式的验证平台硬件架构,非侵入式子系统、侵入子系统和对比平台3个模块独立设计却又功能耦合,各子系统支持共同对负荷信息独立采集,非侵入子系统满足差异化分解算法对采集频率的需求,侵入子系统的数据密度可从分钟级到毫秒级。接着以分层理念为指导开发了验证平台软件架构,平台具备分析非侵入负荷分解算法的启停误差、功率误差和电量误差的功能,并将负荷用电信息以电轨迹图形式进行直观展示,软件的松耦合架构模式方便验证功能的集成与扩展。最后,模拟实验表明,非侵入负荷分解技术验证平台侵入式负荷信息启停时刻统计误差为1 min,功率和电量统计误差小于1%,此误差下,验证平台通过负荷信息对比能够客观评估验证分解算法的准确性,为非侵入负荷分解算法的研发提供了良好的开发环境。
Non-intrusive load decomposition technology as an emerging technology to obtain user power consumption behavior, lacks a verification platform to evaluate decomposition algorithms. In view of the above problems, it is the core idea that comparing the results of non-intrusive load decomposition algorithm with the load information collected by the intrusion subsystem to construct a non-intrusive load decomposition technology verification platform. Firstly, a modular hardware architecture is proposed and designed for the verification platform, the three modules of the non-intrusive subsystem, intrusion subsystem and comparison platform are independently designed but functionally coupled, each subsystem supports independent collection of load information. the non-intrusive subsystem meets the requirements of the differential decomposition algorithm for the acquisition frequency. The data density of the intrusion subsystem can range from minute to milliseconds. Secondly, a layered software architecture is developed for the verification platform which has the ability to analyze the start-stop-moment error, active power error and power consumption error of non-intrusive load decomposition algorithms, and display the load power information visually in the form of electrical track graph, the loosely coupled architecture of software facilitates the integration and extension of verification functions. Lastly, simulation experiment shows that the statistical error of start-stop-moment information obtained by the intrusion subsystem is 1 minute, and the statistical errors of active power and power consumption are both less than 1%, under this error, the verification platform can objectively evaluate the accuracy of the verification decomposition algorithm by comparing the load information, so that it provides a good development environment for the research of non-intrusive load decomposition algorithms.
Non-intrusive load decomposition technology as an emerging technology to obtain user power consumption behavior, lacks a verification platform to evaluate decomposition algorithms. In view of the above problems, it is the core idea that comparing the results of non-intrusive load decomposition algorithm with the load information collected by the intrusion subsystem to construct a non-intrusive load decomposition technology verification platform. Firstly, a modular hardware architecture is proposed and designed for the verification platform, the three modules of the non-intrusive subsystem, intrusion subsystem and comparison platform are independently designed but functionally coupled, each subsystem supports independent collection of load information. the non-intrusive subsystem meets the requirements of the differential decomposition algorithm for the acquisition frequency. The data density of the intrusion subsystem can range from minute to milliseconds. Secondly, a layered software architecture is developed for the verification platform which has the ability to analyze the start-stop-moment error, active power error and power consumption error of non-intrusive load decomposition algorithms, and display the load power information visually in the form of electrical track graph, the loosely coupled architecture of software facilitates the integration and extension of verification functions. Lastly, simulation experiment shows that the statistical error of start-stop-moment information obtained by the intrusion subsystem is 1 minute, and the statistical errors of active power and power consumption are both less than 1%, under this error, the verification platform can objectively evaluate the accuracy of the verification decomposition algorithm by comparing the load information, so that it provides a good development environment for the research of non-intrusive load decomposition algorithms.
引文
[1] 余贻鑫,刘博,栾文鹏.非侵入式居民电力负荷监测与分解技术[J].南方电网技术,2013,7(4):1-5.
[2] 刘世成,刘沅昆,武昕,等.基于大数据的非侵入式负荷分解技术研究[J].电力信息与通信技术,2016,14(12):9-14.
[3] HART G W. Nonintrusive appliance load monitoring [J]. Proceedings of the IEEE, 1992, 80(12): 1870-1891.
[4] 周尤林.关于电力负荷控制系统的发展探索[J].大众用电,2000(4):32-33.
[5] 武昕,祁兵,韩璐,等.基于模板滤波的居民负荷非侵入式快速辨识算法[J].电力系统自动化,2017,41(2):135-141.
[6] WANG Z, ZHENG G. Residential appliances identification and monitoring by a nonintrusive method[J]. IEEE Transactions on Smart Grid, 2012, 3(1):80-92.
[7] PARSON O, GHOSH S, WEAL M, et al. Non-intrusive load monitoring using prior models of general appliance types[C]. AAAI Conference on Artificial Intelligence, 2012:356-362.
[8] 罗显志,杨滕,魏海涛,等.基于MATLAB的卫星导航信号仿真和验证平台[J].系统仿真学报,2009(18):5692-5698,5703.
[9] 吴英攀,于立新,薛可,等.基于层次化验证平台的存储器控制器功能验证[J].微电子学与计算机,2009,26(2):25-28.
[10] 杨冰,陶锋,褚广斌,等.基于USRP的电力线载波技术验证平台设计与实现[J].电测与仪表,2014,51(12):108-113.
[11] 刘志凯,梁成华,王冬,等.基于FPGA的核安全级数字化智能化仿真验证平台[J].仪器仪表用户,2017,24(12):49-52,48.
[12] WANG J, WANG F. An advanced verification platform based on UVM[J]. Applied Mechanics and Materials, 2014, 670-671:6.
[13] ADESH PANWAR. Verification Platform for FPGA Based Architecture[M].Springer India:2012.
[14] 邓宁,宋森森,羿昌宇,等.软件无线电系统综合诊断平台设计与实现[J].电子测量与仪器学报,2016,30(6):968-974.
[15] 孙小进,郭恩全.基于IEEE1232的故障诊断系统的软件架构设计[J].电子测量与仪器学报,2014,28(1):36-42.
[2] 刘世成,刘沅昆,武昕,等.基于大数据的非侵入式负荷分解技术研究[J].电力信息与通信技术,2016,14(12):9-14.
[3] HART G W. Nonintrusive appliance load monitoring [J]. Proceedings of the IEEE, 1992, 80(12): 1870-1891.
[4] 周尤林.关于电力负荷控制系统的发展探索[J].大众用电,2000(4):32-33.
[5] 武昕,祁兵,韩璐,等.基于模板滤波的居民负荷非侵入式快速辨识算法[J].电力系统自动化,2017,41(2):135-141.
[6] WANG Z, ZHENG G. Residential appliances identification and monitoring by a nonintrusive method[J]. IEEE Transactions on Smart Grid, 2012, 3(1):80-92.
[7] PARSON O, GHOSH S, WEAL M, et al. Non-intrusive load monitoring using prior models of general appliance types[C]. AAAI Conference on Artificial Intelligence, 2012:356-362.
[8] 罗显志,杨滕,魏海涛,等.基于MATLAB的卫星导航信号仿真和验证平台[J].系统仿真学报,2009(18):5692-5698,5703.
[9] 吴英攀,于立新,薛可,等.基于层次化验证平台的存储器控制器功能验证[J].微电子学与计算机,2009,26(2):25-28.
[10] 杨冰,陶锋,褚广斌,等.基于USRP的电力线载波技术验证平台设计与实现[J].电测与仪表,2014,51(12):108-113.
[11] 刘志凯,梁成华,王冬,等.基于FPGA的核安全级数字化智能化仿真验证平台[J].仪器仪表用户,2017,24(12):49-52,48.
[12] WANG J, WANG F. An advanced verification platform based on UVM[J]. Applied Mechanics and Materials, 2014, 670-671:6.
[13] ADESH PANWAR. Verification Platform for FPGA Based Architecture[M].Springer India:2012.
[14] 邓宁,宋森森,羿昌宇,等.软件无线电系统综合诊断平台设计与实现[J].电子测量与仪器学报,2016,30(6):968-974.
[15] 孙小进,郭恩全.基于IEEE1232的故障诊断系统的软件架构设计[J].电子测量与仪器学报,2014,28(1):36-42.