基于Azure机器学习平台的大学校园用电分析与预测

Abstract
Figure/Table
References
Related Citation (1)

Download: PDF (5670 KB) HTML (1 KB)
Export: BibTeX | EndNote (RIS)

Abstract The construction of campus power consumption supervision platform in China provides data support for the analysis and prediction of campus power consumption. With the help of Microsoft Azure Machine Learning platform, this paper analyzes the power consumption data of Fuzhou University, and then derives two impact factors of power consumption: air temperature and daily schedules. According to the analysis results, this paper defines the air temperature that leads to the sharp growth of power consumption, and then proposes a segmented prediction algorithm based on this idea. The experimental results prove that the segmented prediction algorithm can effectively improve the prediction accuracy of power consumption.

Key words： machine learning analysis of campus power consumption power consumption prediction segmentedprediction

Received: 20 October 2017 Published: 21 May 2018

	Service

	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Xiong Tian
	Zheng Song
	Xu Zhezhuang
	Xie Renxu
	Ge Yongle

Cite this article:

Xiong Tian,Zheng Song,Xu Zhezhuang等. The analysis and prediction of power consumption in university campus based on azure machine learning platform[J]. Electrical Engineering, 2018, 19(5): 5-9.

URL:

http://dqjs.cesmedia.cn/EN/Y2018/V19/I5/5

[1] 陈晶晶, 李红娇, 许智. 基于随机森林的用电行为分析[J]. 上海电力学院学报, 2017(4): 331-336.
[2] 王仁祥, 王小曼. 终端用户分布式新能源接入智能配电网技术研究[J]. 电气技术, 2010, 11(8): 58-62.
[3] Stroombergen A, Tait A, Patterson K, et al.The relationship between New Zealand's climate, power, and the economy to 2025[J]. New Zealand Journal of Social Sciences, 2006, 13(1): 139-160.
[4] Wang Zhiyong, Cao Yijia.Mutual information and non-fixed ANNs for daily peak load forecasting[J]. Power Sys-temsConference and Expoxision, 2006(5): 1523-1528.
[5] 黄海新, 邓丽, 张路. 基于需求响应的实时电价研究综述[J]. 电气技术, 2015, 16(11): 1-6.
[6] 曲朝阳, 张率, 刘洪涛. 基于用电影响因素回归的小区用电预测模型[J]. 东北电力大学学报, 2015(01): 73-77.
[7] 张棪, 曹健. 面向大数据分析的决策树算法[J]. 计算机科学, 2016(S1): 374-379, 383.
[8] 王桂玲, 韩燕波, 张仲妹, 等. 基于云计算的流数据集成与服务[J]. 计算机学报, 2017(1): 107-125.
[9] 王永康. Azure云平台对Twitter推文关键字实时大数据分析[J]. 电脑编程技巧与维护, 2015(12): 68-72.
[10] Xiao Laisheng, Wang Zhengxia. Cloud computing: A new business paradigmfor E-learning[C]//Inter- national Conference on Measuring Technology and Mechatronics Automation (ICMTMA2011) 3rd, 2011: Shanghai, China.
[11] 王颖, 赵航宇, 赵洪山. 配电网自动化建设的现状与若干建议[J]. 电工技术, 2015(11): 82-83.
[12] 何春光, 卢志明, 姜春莹, 等. 移动式应急配变的研制[J]. 电工技术, 2016(2): 6-7.
[13] Brandon Butler.2017年必须关注的10大云趋势[J]. 计算机世界, 2017(4).
[14] 易植. Windows Azure 新服务, 让机器学习触手可及[J]. 英才, 2014(9).
[15] 韩阳, 吕由, 潘宇航, 等. SVM、BP神经网络、线性回归的比较研究[J]. 河北联合大学学报(自然科学版), 2017, 39(2).
[16] Song Y, Wang H, He X.Adapting deep RankNet for personalized search[C]//ACM International Conference on Web Search and Data Mining, 2014: 83-92.
[17] 柯国霖. 梯度提升决策树(GBDT)并行学习算法研究[D]. 厦门: 厦门大学, 2016.
[18] 王天华. 基于改进的GBDT算法的乘客出行预测研究[D]. 大连: 大连理工大学, 2016.