基于SVD-LSTM的高校学生宿舍空调负荷预测

doi:10.16180/j.cnki.issn1007-7820.2020.11.012

摘要/Abstract

摘要：

准确预测高校空调负荷是保证高校安全用电和电力高峰期区域配电网稳定运行的前提。文中以高校空调负荷中具有代表性的学生宿舍空调负荷为对象,建立了基于奇异值分解-长短期记忆网络的高校学生宿舍空调负荷预测模型。该方法以高校学生宿舍空调负荷特性为基础,使用奇异值分解进行降噪处理,通过长短期记忆网络对高校学生宿舍空调负荷进行预测。文中以武汉某高校的真实数据为样本进行了分析验证,通过与传统预测模型对比,证明所提预测模型的预测效果和精度优于传统预测方法。

关键词: 高校负荷, 空调负荷特性, 奇异值分解, 长短期记忆网络, 负荷预测, 相关性系数

Abstract:

Accurate prediction of air conditioning load in colleges is the premise and basis to ensure the safe electricity consumption and stable operation of regional distribution network during power peak period. In this paper, the student dormitory air conditioning load of college air conditioning is taken as the research object, and an air conditioning load forecasting model based on SVD-LSTM is established. Based on the characteristics of air conditioning load in college dormitory, this model uses SVD to reduce data noise, and predicts the air conditioning load of college students' dormitory through LSTM. The actual data of a university in Wuhan is taken as a sample to analyze and verify the model. It is proves that the prediction result of SVD-LSTM is better by the comparison with traditional prediction model results. The model improves the prediction accuracy. The analysis of a university in Wuhan shows that the prediction effect and accuracy of the proposed prediction model are better than the traditional prediction method.

Key words: college load, air-conditioning load characteristics, SVD, LSTM, load forecasting, correlation coefficient

中图分类号:

TP183

祁鑫,王福忠,张丽,王瑞,王晓慧. 基于SVD-LSTM的高校学生宿舍空调负荷预测[J]. 电子科技, 2020, 33(11): 59-66.

QI Xin,WANG Fuzhong,ZHANG Li,WANG Rui,WANG Xiaohui. Air Conditioning Load Forecast of University Students' Dormitory Based on SVD-LSTM[J]. Electronic Science and Technology, 2020, 33(11): 59-66.

图/表 12

图1

图2

图3

图4

图5

图6

表1

图7

表2

表3

表4

图8

参考文献 24

[1]	Zhao J, Duan Y, Liu X. Uncertainty analysis of weather forecast data for cooling load forecasting based on the Monte Carlo method[J]. Energies, 2018,11(7):1900-1919.
[2]	Dyson M E H, Borgeson S D, Tabone M D, et al. Using smart meter data to estimate demand response potential, with application to solar energy integration[J]. Energy Policy, 2014,73(3):607-619.
[3]	何大四, 张旭. 改进的季节性指数平滑法预测空调负荷分析[J]. 同济大学学报(自然科学版), 2005(12):1672-1676.
	He Dasi, Zhang Xu. Analysis of air conditioning load prediction by modified seasonal exponential smoothing model[J]. Journal of Tongji University (Natural Science), 2005(12):1672-1676.
[4]	任志超, 杜新伟, 王海燕, 等. 调温负荷的估算方法及影响因素研究[J]. 现代电力, 2014,31(3):80-85.
	Ren Zhichao, Du Xinwei, Wang Haiyan, et al. Research on estimation method of temperature- adjusting load and its influence factors[J]. Modern Electric Power, 2014,31(3):80-85.
[5]	王东, 史晓霞, 尹交英. 不同核函数的支持向量机用于空调负荷预测的对比研究[J]. 电工技术学报, 2015,30(S1):531-535.
	Wang Dong, Shi Xiaoxia, Yin Jiaoying. Prediction on hourly load of air conditioning by RBF support vector machine[J]. Transactions of China Electrotechnical Society, 2015,30(S1):531-535.
[6]	赵超, 戴坤成. 自适应加权最小二乘支持向量机的空调负荷预测[J]. 重庆大学学报, 2016,39(1):55-64.
	Zhao Chao, Dai Kuncheng. Modeling air-conditioning load forecasting based on adaptive weighted least squares support vector machine[J]. Journal of Chongqing University, 2016,39(1):55-64.
[7]	Liao G C. Hybrid improved differential evolution and wavelet neural network with load forecasting problem of air conditioning[J]. International Journal of Electrical Power & Energy Systems, 2014,61(1):673-682.
[8]	李帆, 曲世琳, 于丹, 等. 基于运行数据人工神经网络的空调系统逐时负荷预测[J]. 建筑科学, 2014,30(2):72-75.
	Li Fan, Qu Shilin, Yu Dan, et al. Prediction on hourly load of air conditioning system based on operating data and artificial neural network[J]. Building Science, 2014,30(2):72-75.
[9]	Fu G. Deep belief network based ensemble approach for cooling load forecasting of air-conditioning system[J]. Energy, 2018,148(1):269-282. doi: 10.1016/j.energy.2018.01.180
[10]	Palchak D, Suryanarayanan S, Zimmerle D. An artificial neural network in short-term electrical load forecasting of a university campus: a case study[J]. Journal of Energy Resources Technology, 2013,135(3):032001-032018.
[11]	Wang S, Wang X, Wang S, et al. Bi-directional long short-term memory method based on attention mechanism and rolling update for short-term load forecasting[J]. International Journal of Electrical Power & Energy Systems, 2019,109(2):470-479.
[12]	Kong W, Dong Z Y, Jia Y, et al. Short-term residential load forecasting based on LSTM recurrent neural network[J]. IEEE Transactions on Smart Grid, 2017,10(1):841-851.
[13]	Jiao R, Zhang T, Jiang Y, et al. Short-term non-residential load forecasting based on multiple sequences LSTM recurrent neural network[J]. IEEE Access, 2018(6):59438-59448.
[14]	郭远晶, 魏燕定, 周晓军, 等. S变换时频谱SVD降噪的冲击特征提取方法[J]. 振动工程学报, 2014,27(4):621-628.
	Guo Yuanjing, Wei Yanding, Zhou Xiaojun, et al. Impact feature extracting method based on S transform time-frequency spectrum denoised by SVD[J]. Journal of Vibration Engineering, 2014,27(4):621-628.
[15]	池传国, 黄国勇, 孙磊. M估计的强跟踪SVD-UKF算法在组合导航中的应用[J]. 电子科技, 2018,31(7):42-45,54.
	Chi Chuanguo, Huang Guoyong, Sun Lei. Application of strong tracking SVD-UKF algorithm based on m estimation in integrated navigation system[J]. Electronic Science and Technology, 2018,31(7):42-45,54.
[16]	陶维青, 王乐勤, 顾芝瑕, 等. SVD和神经网络在孤岛检测中的应用[J]. 电力系统保护与控制, 2017,45(2):28-34.
	Tao Weiqing, Wang Leqin, Gu Zhixia, et al. Application of SVD and neural network in islanding detection[J]. Power System Protection and Control, 2017,45(2):28-34.
[17]	王天豪. 高校公共建筑能耗影响因素与预测模型构建研究[D]. 西安:西安建筑科技大学, 2018.
	Wang Tianhao. Research on influencing factors and prediction model of public building energy consumption in colleges and universities[D]. Xi’an:Xi’an University of Architecture and Technology, 2018.
[18]	苏凡军, 唐启桂. SBHCF:基于奇异值分解的混合协同过滤推荐算法[J]. 电子科技, 2016,29(1):44-47.
	Su Fanjun, Tang Qigui. SBHCF:An SVD-based hybrid collaborative filtering recommendation algorithm[J]. Electronic Science and Technology, 2016,29(1):44-47.
[19]	隋秀凛, 陈云壮, 葛江华, 等. 基于相关系数的时频矩阵SVD降噪方法[J]. 控制工程, 2018,25(10):1934-1939.
	Sui Xiuling, Chen Yunzhuang, Ge Jianghua, et al. Time frequency matrix SVD de-noising method based on correlation coefficient[J]. Control Engineering of China, 2018,25(10):1934-1939.
[20]	曾鸣, 杨宇, 郑近德, 等. μ-SVD降噪算法及其在齿轮故障诊断中的应用[J]. 机械工程学报, 2015,51(3):95-103.
	Zeng Ming, Yang Yu, Zheng Jinde, et al. μ-SVD based denoising method and its application to gear fault diagnosis[J]. Journal of Mechanical Engineering, 2015,51(3):95-103.
[21]	陆继翔, 张琪培, 杨志宏, 等. 基于CNN-LSTM混合神经网络模型的短期负荷预测方法[J]. 电力系统自动化, 2019,43(8):131-137.
	Lu Jixiang, Zhang Qipei, Yang Zhihong, et al. Short-term load forecasting method based on CNN- LSTM hybrid neural network model[J]. Automation of Electric Power Systems, 2019,43(8):131-137.
[22]	代杰杰, 宋辉, 盛戈皞, 等. 采用LSTM网络的电力变压器运行状态预测方法研究[J]. 高电压技术, 2018,44(4):1099-1106.
	Dai Jiejie, Song Hui, Sheng Gehao, et al. Prediction method for power transformer running state based on LSTM network[J]. High Voltage Engineering, 2018,44(4):1099-1106.
[23]	魏昱洲, 许西宁. 基于LSTM长短期记忆网络的超短期风速预测[J]. 电子测量与仪器学报, 2019,33(2):64-71.
	Wei Yuzhou, Xu Xining. Ultra-short-term wind speed prediction model using LSTM networks[J]. Journal of Electronic Measurement and Instrumentation, 2019,33(2):64-71.
[24]	夏宇彬, 郑建立, 赵逸凡, 等. 基于深度学习的电子病历命名实体识别[J]. 电子科技, 2018,31(11):31-34,37.
	Xia Yubin, Zheng Jianli, Zhao Yifan, et al. Deeplearning based named entity recognition of electronic medical record[J]. Electronic Science and Technology, 2018,31(11):31-34,37.

楼层	寝室数目/间	占总寝室数比例	抽样寝室数目/间
1层	16	12%	3
2层	35	26%	7
3层	39	29%	7
4层	44	33%	9
总计	134	100%	26

隐藏层数	Block数目/个	迭代次数/次	每组输入数据数/个	RMSE	MAE
1	1	1 600	12	0.066 46	0.095 6
1	2	1 600	12	0.066 52	0.095 5
1	3	1 600	12	0.066 72	0.095 8
1	4	1 600	12	0.065 71	0.095 1
1	5	1 600	12	0.065 95	0.094 5

隐藏层数	Block数目/个	迭代次数/次	每组输入数据数/个	RMSE	MAE
1	4	1 600	12	0.065 71	0.095 1
2	4	1 600	12	0.066 51	0.095 3
3	4	1 600	12	0.065 31	0.094 9
4	4	1 600	12	0.065 57	0.095 3
5	4	1 600	12	0.067 08	0.094 9

模型	RMSE	MAE
ARIMA	0.105 00	0.113 3
LSTM	0.065 31	0.094 9
SVD-LSTM	0.064 37	0.093 3