Distributed Photovoltaic Net Load Forecasting in New Energy Power Systems

doi:10.16183/j.cnki.jsjtu.2021.244

Abstract

Abstract:

To respond to the demand of achieving carbon peaking and carbon neutrality goals, and to construct a complete “source-grid-load-storage” new energy power system, a distributed photovoltaic net load forecasting model based on Hamiltonian Monte Carlo inference for deep Gaussian processes (HMCDGP) is proposed. First, direct and indirect forecasting methods are used to examine the accuracy of the proposed model and to obtain spot forecasting results. Then, the proposed model is used to perform probability forecasting experiments and produce interval prediction results. Finally, the superiority of the proposed model is verified through the comparative experiments based on the net load data of 300 households recorded by Australia Grid. After obtaining the exact net load probabilistic forecasting results, the photovoltaic production can be fully utilized via power dispatch, which can reduce the use of fossil energy and further reduce the carbon emission.

Key words: net load forecasting, photovoltaic production, deep Gaussian process, point forecasting, interval prediction

CLC Number:

TM615
TM715

LIAO Qishu, HU Weihao, CAO Di, HUANG Qi, CHEN Zhe. Distributed Photovoltaic Net Load Forecasting in New Energy Power Systems[J]. Journal of Shanghai Jiao Tong University, 2021, 55(12): 1520-1531.

Figures/Tables 9

Fig.1

Fig.2

Tab.1

Fig.3

Tab.2

Tab.3

Fig.4

Tab.4

Fig.5

References 24

[1]	江泽民. 对中国能源问题的思考[J]. 上海交通大学学报, 2008, 42(3):345-359.
	JIANG Zemin. Reflections on energy issues in China[J]. Journal of Shanghai Jiao Tong University, 2008, 42(3):345-359.
[2]	HONG T, WANG P, WILLIS H L. A naïve multiple linear regression benchmark for short term load forecasting[C]// 2011 IEEE Power and Energy Society General Meeting. Detroit, MI, USA: IEEE, 2011: 1-6.
[3]	肖白, 刘庆永, 牛强, 等. 基于元胞负荷特性分析的RBF神经网络空间负荷预测方法[J]. 电网技术, 2018, 42(1):301-307.
	XIAO Bai, LIU Qingyong, NIU Qiang, et al. A spatial load forecasting method based on RBF neural network and cellular load characteristics analysis[J]. Power System Technology, 2018, 42(1):301-307.
[4]	吴云, 雷建文, 鲍丽山, 等. 基于改进灰色关联分析与蝙蝠优化神经网络的短期负荷预测[J]. 电力系统自动化, 2018, 42(20):67-72.
	WU Yun, LEI Jianwen, BAO Lishan, et al. Short-term load forecasting based on improved grey relational analysis and neural network optimized by bat algorithm[J]. Automation of Electric Power Systems, 2018, 42(20):67-72.
[5]	占勇, 程浩忠, 丁屹峰. 自适应神经网络在短期负荷预测中的应用[J]. 上海交通大学学报, 2005, 39(Sup.1):14-17.
	ZHAN Yong, CHENG Haozhong, DING Yifeng. The application of adaptive neural network in short term load forecasting[J]. Journal of Shanghai Jiao Tong University, 2005, 39(Sup.1):14-17.
[6]	赵佩, 代业明. 基于实时电价和加权灰色关联投影的SVM电力负荷预测[J]. 电网技术, 2020, 44(4):1325-1332.
	ZHAO Pei, DAI Yeming. Power load forecasting of SVM based on real-time price and weighted grey relational projection algorithm[J]. Power System Technology, 2020, 44(4):1325-1332.
[7]	陈振宇, 刘金波, 李晨, 等. 基于LSTM与XGBoost组合模型的超短期电力负荷预测[J]. 电网技术, 2020, 44(2):614-620.
	CHEN Zhenyu, LIU Jinbo, LI Chen, et al. Ultra short-term power load forecasting based on combined LSTM-XGBoost model[J]. Power System Technology, 2020, 44(2):614-620.
[8]	HONG T, FAN S. Probabilistic electric load forecasting: A tutorial review[J]. International Journal of Forecasting, 2016, 32(3):914-938. doi: 10.1016/j.ijforecast.2015.11.011 URL
[9]	LIU B D, NOWOTARSKI J, HONG T, et al. Probabilistic load forecasting via quantile regression averaging on sister forecasts[J]. IEEE Transactions on Smart Grid, 2017, 8(2):730-737.
[10]	何耀耀, 闻才喜, 许启发, 等. 考虑温度因素的中期电力负荷概率密度预测方法[J]. 电网技术, 2015, 39(1):176-181.
	HE Yaoyao, WEN Caixi, XU Qifa, et al. A method to predict probability density of medium-term power load considering temperature factor[J]. Power System Technology, 2015, 39(1):176-181.
[11]	庞昊, 高金峰, 杜耀恒. 基于时间卷积网络分位数回归的短期负荷概率密度预测方法[J]. 电网技术, 2020, 44(4):1343-1350.
	PANG Hao, GAO Jinfeng, DU Yaoheng. A short-term load probability density prediction based on quantile regression of time convolution network[J]. Power System Technology, 2020, 44(4):1343-1350.
[12]	何耀耀, 刘瑞, 撖奥洋. 基于实时电价与支持向量分位数回归的短期电力负荷概率密度预测方法[J]. 中国电机工程学报, 2017, 37(3):768-776.
	HE Yaoyao, LIU Rui, HAN Aoyang. Short-term power load probability density forecasting method based on real time price and support vector quantile regression[J]. Proceedings of the CSEE, 2017, 37(3):768-776.
[13]	何耀耀, 许启发, 杨善林, 等. 基于RBF神经网络分位数回归的电力负荷概率密度预测方法[J]. 中国电机工程学报, 2013, 33(1):93-98.
	HE Yaoyao, XU Qifa, YANG Shanlin, et al. A power load probability density forecasting method based on RBF neural network quantile regression[J]. Proceedings of the CSEE, 2013, 33(1):93-98.
[14]	ALIPOUR M, AGHAEI J, NOROUZI M, et al. A novel electrical net-load forecasting model based on deep neural networks and wavelet transform integration[J]. Energy, 2020, 205:118106.
[15]	方陈, 江兴稳, 周健, 等. 含分布式能源区域电网月最大净负荷概率预测[J]. 水电能源科学, 2018, 36(9):197-200.
	FANG Chen, JIANG Xingwen, ZHOU Jian, et al. Probability prediction of monthly net load for regional power grid with distributed energy penetration[J]. Water Resources and Power, 2018, 36(9):197-200.
[16]	SUN M Y, ZHANG T Q, WANG Y, et al. Using Bayesian deep learning to capture uncertainty for residential net load forecasting[J]. IEEE Transactions on Power Systems, 2020, 35(1):188-201. doi: 10.1109/TPWRS.59 URL
[17]	刘升伟, 王星华, 鲁迪, 等. 基于改进高斯过程回归的短期负荷概率区间预测方法[J]. 电力系统保护与控制, 2020, 48(1):18-25.
	LIU Shengwei, WANG Xinghua, LU Di, et al. Electric load probabilistic interval prediction method based on improved Gaussian process regression[J]. Power System Protection and Control, 2020, 48(1):18-25.
[18]	宗文婷, 卫志农, 孙国强, 等. 基于改进高斯过程回归模型的短期负荷区间预测[J]. 电力系统及其自动化学报, 2017, 29(8):22-28.
	ZONG Wenting, WEI Zhinong, SUN Guoqiang, et al. Short-term load interval prediction based on improved Gaussian process regression model[J]. Proceedings of the CSU-EPSA, 2017, 29(8):22-28.
[19]	彭虹桥, 顾洁, 胡玉, 等. 基于混沌粒子群—高斯过程回归的饱和负荷概率预测模型[J]. 电力系统自动化, 2017, 41(21):25-32.
	PENG Hongqiao, GU Jie, HU Yu, et al. Forecasting model of saturated load based on chaotic particle swarm and optimization-Gaussian process regression[J]. Automation of Electric Power Systems, 2017, 41(21):25-32.
[20]	VAN DER MEER D W, MUNKHAMMAR J, WIDÉN J. Probabilistic forecasting of solar power, electricity consumption and net load: Investigating the effect of seasons, aggregation and penetration on prediction intervals[J]. Solar Energy, 2018, 171:397-413. doi: 10.1016/j.solener.2018.06.103 URL
[21]	HAVASI M, HERNANDEZ-LOBATO J M, MURILLO-FUENTES J J. Inference in deep Gaussian processes using stochastic gradient Hamiltonian Monte Carlo[EB/OL].(2018-11-12)[2021-03-04]. https://arxiv.org/abs/1806.05490.
[22]	DAMIANOU A C. Deep Gaussian processes and variational propagation of uncertainty[EB/OL].(2015-02-10) [2021-03-04]. https://www.researchgate.net/publication/283296556_Deep_Gaussian_Processes_and_Variational_Propagation_of_Uncertainty.
[23]	SPRINGENBERG J T, KLEIN A, FALKNER S, et al. Bayesian optimization with robust Bayesian neural networks[EB/OL]. [2021-03-04] https://xueshu.baidu.com/usercenter/paper/show?paperid=14b338c730619da3fd3ba5160a7cd186.
[24]	PINSON P, NIELSEN H A, MØLLER J K, et al. Non-parametric probabilistic forecasts of wind power: Required properties and evaluation[J]. Wind Energy, 2007, 10(6):497-516. doi: 10.1002/(ISSN)1099-1824 URL

方法	RMSE		MAE		NRMSD
方法	间接	直接	间接	直接	间接	直接
BPN	35.00	35.74	20.22	21.19	0.109	0.111
QR	33.53	33.52	22.06	21.75	0.104	0.104
SVR	32.52	31.42	19.38	20.80	0.100	0.098
LR	31.14	31.04	21.35	22.41	0.096	0.096
SGP	31.34	31.49	20.83	21.05	0.097	0.098
HMCDGP	28.65	28.12	18.67	19.01	0.089	0.087

方法	夏季			冬季
方法	RMSE	MAE	NRMSD	RMSE	MAE	NRMSD
BPN	35.740	21.185	0.1109	41.267	29.484	0.1369
QR	33.522	21.751	0.1040	39.462	25.219	0.1309
SGP	31.490	21.047	0.0977	34.702	24.657	0.1151
SVR	31.424	20.804	0.0975	35.929	25.718	0.1192
LR	31.044	22.411	0.0963	36.272	27.598	0.1203
HMCDGP	28.124	19.014	0.0872	33.950	24.646	0.1126

方法	夏季		冬季
方法	Q^b	Q^a/%	Q^b	Q^a/%
QR	7.533	22.56	12.428	26.18
SGP	8.326	23.17	9.500	28.91
HMCDGP	6.803	19.57	8.712	24.92

方法	Q^a=90%	Q^a=80%	Q^a=70%	Q^a=60%
QR	29.44	22.56	18.09	14.34
SGP	29.51	23.17	18.77	15.32
HMCDGP	25.06	19.57	15.82	12.79