[1]张顺淼、陈铭龙、洪茂雄.基于优化多类Adaboost的非侵入式负荷监测[J].福建工程学院学报,2019,17(04):352-358.[doi:10.3969/j.issn.1672-4348.2019.04.008]
　ZHANG Shunmiao,CHEN Minglong,HUNG Maosiung.Non-intrusive load monitoring based on optimized multi-class Adaboost[J].Journal of FuJian University of Technology,2019,17(04):352-358.[doi:10.3969/j.issn.1672-4348.2019.04.008]
点击复制

基于优化多类Adaboost的非侵入式负荷监测()
分享到：

《福建工程学院学报》[ISSN:2097-3853/CN:35-1351/Z]

卷:
第17卷
期数:
2019年04期
页码:
352-358
栏目:
出版日期:
2019-08-25

文章信息/Info

Title:
Non-intrusive load monitoring based on optimized multi-class Adaboost
作者:
张顺淼、陈铭龙、 102;  102); font-family: Arial;  Verdana;  sans-serif; font-size: 12px; background-color: rgb(255;  255;  255);">洪茂雄
福建工程学院信息科学与工程学院
Author(s):
ZHANG Shunmiao;  CHEN Minglong;  HUNG Maosiung
School of Information Science and Engineering, Fujian University of Technology
关键词:
非侵入式负荷监测技术; 多类Adaboost; 多状态辨识; 遗传算法
Keywords:
non-intrusive load monitoring;  multi-class Adaboost;  multi-state identification;  genetic algorithm
分类号:
TM714
DOI:
10.3969/j.issn.1672-4348.2019.04.008
文献标志码:
A
摘要:
针对非侵入式负荷监测技术在多状态设备的工作状态辨识研究较少及精度不高的问题，提出了一种基于遗传算法优化的多类Adaboost的非侵入式负荷监测技术。首先提取原始数据集有效特征(电流有效值及其变化量、有功功率及其变化量、无功功率)。其次利用遗传算法优化多类Adaboost中的五个参数，得到最优强分类器。最后通过第六届“泰迪杯”数据挖掘挑战赛A题数据对同时运行两个设备(九阳热水壶、激光打印机)所有状态进行识别。实验结果表明，该算法识别能力优于决策树算法和SVM算法。
Abstract:
Aiming at the lack of research and low accuracy of non-intrusive load monitoring technology in working status identification of multi-state equipment, non-intrusive load monitoring technology based on genetic algorithm optimization for multi-class Adaboost is proposed. Firstly, the effective features(effective value of current and its variation, active power and its variation, reactive power) of the original data set are extracted. Secondly, the genetic algorithm is used to optimize the five parameters of multi-class Adaboost to obtain the optimal strong classifier. Finally, all the states of two devices (Joyoung kettle and FUJI laser printer) running simultaneously are identified by the data set of item A of the 6th “Teddy Cup” Data Mining Race. The experimental results show that the recognition ability of this algorithm is better than that of decision tree algorithm and SVM algorithm.

参考文献/References:

[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] HART G W. Nonintrusive appliance load monitoring[J]. Proceedings of the IEEE, 1992, 80(12): 1870-1891. [3] FIGUEIREDO M, DE ALMEIDA A, RIBEIRO B. Home electrical signal disaggregation for non-intrusive load monitoring (NILM) systems[J]. Neurocomputing, 2012, 96(3): 66-73. [4] 孙毅, 崔灿, 陆俊, 等. 基于遗传优化的非侵入式家居负荷分解方法[J]. 电网技术, 2016, 40(12): 3912-3917.[5] GUEDES J D S, FERREIRA D D, BARBOSA B H G. A non-intrusive approach to classify electrical appliances based on higher-order statistics and genetic algorithm: a smart grid perspective[J]. Electric Power Systems Research, 2016, 140: 65-69.[6] 王晓换, 李如意, 周东国, 等. 基于决策融合的非侵入式电力负荷辨识方法及应用[J]. 电力系统保护与控制, 2016, 44(7): 115-121.[7] BASU K, DEBUSSCHERE V, BACHA S, et al. A generic data driven approach for low sampling load disaggregation[J]. Sustainable Energy, Grids and Networks, 2017, 9: 118-127.[8] DE BAETS L, RUYSSINCK J, DEVELDER C, et al. Appliance classification using VI trajectories and convolutional neural networks[J]. Energy and Buildings, 2018, 158: 32-36. [9] 王守相, 刘天宇. 计及用电模式的居民负荷梯度提升树分类识别方法[J]. 电力系统及其自动化学报, 2017, 29(9): 27-33.[10] 宋旭帆, 周明, 涂京, 等. 基于k-NN结合核Fisher判别的非侵入式负荷监测方法[J]. 电力系统自动化, 2018, 42(6): 73-80.[11] 周明, 宋旭帆, 涂京, 等. 基于非侵入式负荷监测的居民用电行为分析[J]. 电网技术, 2018, 42(10): 3268-3276.[12] HASTIE T, ROSSET S, ZHU J, et al. Multi-class AdaBoost[J]. Statistics and Its Interface, 2009, 2(3): 349-360.[13] 翟夕阳, 王晓丹, 雷蕾, 等. 基于多类指数损失函数逐步添加模型的改进多分类AdaBoost算法[J]. 计算机应用, 2017, 37(6): 1692-1696.[14] 黄新波, 李文君子, 宋桐, 等. 采用遗传算法优化装袋分类回归树组合算法的变压器故障诊断[J]. 高电压技术, 2016, 42(5): 1617-1623.

更新日期/Last Update: 2019-08-25