基于GRU-NN模型的短期负荷预测方法
|
摘要
目前基于统计分析和机器学习的预测方法难以同时兼顾负荷数据的时序性和非线性特点。文中提出了一种基于GRU-NN模型的短期电力负荷预测方法。该方法基于深度学习思想处理不同类型的负荷影响因素,引入门控循环单元(GRU)网络处理具有时序性特点的历史负荷序列,建模学习负荷数据内部动态变化规律,其输出结果与其他外部影响因素(天气、日类型等)融合为新的输入特征,使用深度神经网络进行处理,整体分析特征与负荷变化的内在联系,最后完成负荷预测。以美国某公共事业部门提供的公开数据集和中国某地区的负荷数据作为实际算例,该方法预测精度分别达到了97.30%和97.12%,并与长短期记忆神经网络、多层感知机以及GRU神经网络方法进行对比,实验结果表明所提方法具有更高的预测精度和更快的预测速度。
At present, the prediction methods based on statistical analysis and machine learning cannot simultaneously consider the time series and nonlinear characteristics of load data. This paper proposes a short-term power load forecasting method based on GRU-NN model. The method is based on the deep learning idea to deal with different types of load influencing factors, and introduces the gated recurrent unit(GRU) network to process the historical load sequence with time series characteristics. A model is developed to learn the internal dynamic change law of the load data, and its output and other external influence factors(weather, day type) are merged into new input features. The deep neural network is used to process the data. The internal relationship between the characteristics and load changes is analyzed, and the load forecasting is finally completed. Taking the public data set provided by a public utility department in the United States and the load data of a certain region in China as practical examples, the forecasting accuracy of the proposed method is 97.30% and 97.12%, respectively. The proposed method is compared with long short-term memory neural network, multi-layer perceptron and GRU neural network, the experimental results show that the proposed method has higher forecasting accuracy and faster forecasting speed.
At present, the prediction methods based on statistical analysis and machine learning cannot simultaneously consider the time series and nonlinear characteristics of load data. This paper proposes a short-term power load forecasting method based on GRU-NN model. The method is based on the deep learning idea to deal with different types of load influencing factors, and introduces the gated recurrent unit(GRU) network to process the historical load sequence with time series characteristics. A model is developed to learn the internal dynamic change law of the load data, and its output and other external influence factors(weather, day type) are merged into new input features. The deep neural network is used to process the data. The internal relationship between the characteristics and load changes is analyzed, and the load forecasting is finally completed. Taking the public data set provided by a public utility department in the United States and the load data of a certain region in China as practical examples, the forecasting accuracy of the proposed method is 97.30% and 97.12%, respectively. The proposed method is compared with long short-term memory neural network, multi-layer perceptron and GRU neural network, the experimental results show that the proposed method has higher forecasting accuracy and faster forecasting speed.
引文
[1] 程宇也.基于人工神经网络的短期电力负荷预测研究[D].杭州:浙江大学,2017. CHENG Yuye. Short-term electricity demand forecasting based on artificial neural network[D]. Hangzhou: Zhejiang University, 2017.
[2] CAPRIO U D, GENESIO R, POZZI S, et al. Short term load forecasting in electric power systems: a comparison of ARMA models and extended wiener filtering[J]. Journal of Forecasting, 2010, 2(1): 59-76.
[3] NURY A H, HASAN K, ALAM M J B. Comparative study of wavelet-ARIMA and wavelet-ANN models for temperature time series data in Northeastern Bangladesh[J]. Journal of King Saud University - Science, 2017, 29(1): 47-61.
[4] YING L C, PAN M C. Using adaptive network based fuzzy inference system to forecast regional electricity loads[J]. Energy Conversion and Management, 2008, 49(2): 205-211.
[5] HERNáNDEZ L, BALADRóN C, AGUIAR J M, et al. Artificial neural network for short-term load forecasting in distribution systems[J]. Energies, 2014, 7(3): 1576-1598.
[6] LIN W M, TU C S, YANG R F, et al. Particle swarm optimization aided least-square support vector machine for load forecast with spikes[J]. IET Generation, Transmission & Distribution, 2016, 10(5): 1145-1153.
[7] 吴倩红,高军,侯广松,等.实现影响因素多源异构融合的短期负荷预测支持向量机算法[J].电力系统自动化,2016,40(15):67-72.WU Qianhong, GAO Jun, HOU Guangsong, et al. Short-term load forecasting support vector machine algorithm based on multi-source heterogeneous fusion of load factors[J]. Automation of Electric Power Systems, 2016, 40(15): 67-72.
[8] GARCíA-ASCANIO C, MATé C. Electric power demand forecasting using interval time series: a comparison between VAR and iMLP[J]. Energy Policy, 2010, 38(2): 715-725.
[9] LI Y, WANG J. The load forecasting model based on Bayes-GRNN[J]. Journal of Software, 2012, 7(6): 1273-1280.
[10] TONG Chao, LI Jun, LANG Chao, et al. An efficient deep model for day-ahead electricity load forecasting with stacked denoising auto-encoders[J]. Journal of Parallel and Distributed Computing, 2018, 117: 267-273.
[11] 孔祥玉,郑锋,鄂志君,等.基于深度信念网络的短期负荷预测方法[J].电力系统自动化,2018,42(5):133-139.DOI:10.7500/AEPS20170826002.KONG Xiangyu, ZHENG Feng, E Zhijun, et al. Short-term load forecasting based on deep belief network[J]. Automation of Electric Power Systems, 2018, 42(5): 133-139. DOI: 10.7500/AEPS20170826002.
[12] HOSEIN S, HOSEIN P. Load forecasting using deep neural networks[C]// IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT), April 23-26, 2017, Washington, USA: 1-5.
[13] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[14] HE W. Load forecasting via deep neural networks[J]. Procedia Computer Science, 2017, 122: 308-314.
[15] NARAYAN A, HIPEL K W. Long short term memory networks for short-term electric load forecasting[C]// IEEE International Conference on Systems, Man, and Cybernetics (SMC), October 5-8, 2017, Banff, Canada: 2573-2578.
[16] CHO K, MERRIENBOER B V, BAHDANAU D, et al. On the properties of neural machine translation: encoder-decoder approaches[EB/OL]. [2018-06-05]. https://arxiv.org/pdf/1409.1259.pdf.
[17] GUPTA A, GURRALA G, SASTRY P S. Instability prediction in power systems using recurrent neural networks[C]// Twenty-sixth International Joint Conference on Artificial Intelligence, August 19-25, 2017, Melbourne, Australia: 1795-1801.
[18] FU Rui, ZHANG Zuo, LI Li. Using LSTM and GRU neural network methods for traffic flow prediction[C]// 31st Youth Academic Annual Conference of Chinese Association of Automation (YAC), November 11-13 , 2016, Wuhan, China: 324-328.
[19] GRAVES A. Generating sequences with recurrent neural networks[EB/OL]. [2018-06-05]. https://arxiv.org/pdf/1308.0850.pdf.
[20] MAAS A L, QI Peng, XIE Z, et al. Building DNN acoustic models for large vocabulary speech recognition[J]. Computer Speech and Language, 2017, 41: 195-213.
[21] KINGMA D P, BA J. Adam: a method for stochastic optimization[EB/OL]. [2018-06-05]. https://arxiv.org/pdf/1412.6980.pdf.
[22] KONG Weicong, DONG Zhaoyang, HILL D J, et al. Short-term residential load forecasting based on resident behavior learning[J]. IEEE Transactions on Power Systems, 2018, 33(1): 1087-1088.
[2] CAPRIO U D, GENESIO R, POZZI S, et al. Short term load forecasting in electric power systems: a comparison of ARMA models and extended wiener filtering[J]. Journal of Forecasting, 2010, 2(1): 59-76.
[3] NURY A H, HASAN K, ALAM M J B. Comparative study of wavelet-ARIMA and wavelet-ANN models for temperature time series data in Northeastern Bangladesh[J]. Journal of King Saud University - Science, 2017, 29(1): 47-61.
[4] YING L C, PAN M C. Using adaptive network based fuzzy inference system to forecast regional electricity loads[J]. Energy Conversion and Management, 2008, 49(2): 205-211.
[5] HERNáNDEZ L, BALADRóN C, AGUIAR J M, et al. Artificial neural network for short-term load forecasting in distribution systems[J]. Energies, 2014, 7(3): 1576-1598.
[6] LIN W M, TU C S, YANG R F, et al. Particle swarm optimization aided least-square support vector machine for load forecast with spikes[J]. IET Generation, Transmission & Distribution, 2016, 10(5): 1145-1153.
[7] 吴倩红,高军,侯广松,等.实现影响因素多源异构融合的短期负荷预测支持向量机算法[J].电力系统自动化,2016,40(15):67-72.WU Qianhong, GAO Jun, HOU Guangsong, et al. Short-term load forecasting support vector machine algorithm based on multi-source heterogeneous fusion of load factors[J]. Automation of Electric Power Systems, 2016, 40(15): 67-72.
[8] GARCíA-ASCANIO C, MATé C. Electric power demand forecasting using interval time series: a comparison between VAR and iMLP[J]. Energy Policy, 2010, 38(2): 715-725.
[9] LI Y, WANG J. The load forecasting model based on Bayes-GRNN[J]. Journal of Software, 2012, 7(6): 1273-1280.
[10] TONG Chao, LI Jun, LANG Chao, et al. An efficient deep model for day-ahead electricity load forecasting with stacked denoising auto-encoders[J]. Journal of Parallel and Distributed Computing, 2018, 117: 267-273.
[11] 孔祥玉,郑锋,鄂志君,等.基于深度信念网络的短期负荷预测方法[J].电力系统自动化,2018,42(5):133-139.DOI:10.7500/AEPS20170826002.KONG Xiangyu, ZHENG Feng, E Zhijun, et al. Short-term load forecasting based on deep belief network[J]. Automation of Electric Power Systems, 2018, 42(5): 133-139. DOI: 10.7500/AEPS20170826002.
[12] HOSEIN S, HOSEIN P. Load forecasting using deep neural networks[C]// IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT), April 23-26, 2017, Washington, USA: 1-5.
[13] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[14] HE W. Load forecasting via deep neural networks[J]. Procedia Computer Science, 2017, 122: 308-314.
[15] NARAYAN A, HIPEL K W. Long short term memory networks for short-term electric load forecasting[C]// IEEE International Conference on Systems, Man, and Cybernetics (SMC), October 5-8, 2017, Banff, Canada: 2573-2578.
[16] CHO K, MERRIENBOER B V, BAHDANAU D, et al. On the properties of neural machine translation: encoder-decoder approaches[EB/OL]. [2018-06-05]. https://arxiv.org/pdf/1409.1259.pdf.
[17] GUPTA A, GURRALA G, SASTRY P S. Instability prediction in power systems using recurrent neural networks[C]// Twenty-sixth International Joint Conference on Artificial Intelligence, August 19-25, 2017, Melbourne, Australia: 1795-1801.
[18] FU Rui, ZHANG Zuo, LI Li. Using LSTM and GRU neural network methods for traffic flow prediction[C]// 31st Youth Academic Annual Conference of Chinese Association of Automation (YAC), November 11-13 , 2016, Wuhan, China: 324-328.
[19] GRAVES A. Generating sequences with recurrent neural networks[EB/OL]. [2018-06-05]. https://arxiv.org/pdf/1308.0850.pdf.
[20] MAAS A L, QI Peng, XIE Z, et al. Building DNN acoustic models for large vocabulary speech recognition[J]. Computer Speech and Language, 2017, 41: 195-213.
[21] KINGMA D P, BA J. Adam: a method for stochastic optimization[EB/OL]. [2018-06-05]. https://arxiv.org/pdf/1412.6980.pdf.
[22] KONG Weicong, DONG Zhaoyang, HILL D J, et al. Short-term residential load forecasting based on resident behavior learning[J]. IEEE Transactions on Power Systems, 2018, 33(1): 1087-1088.