| 1,249 | 55 | 32 |
| 下载次数 | 被引频次 | 阅读次数 |
针对短期电力负荷预测目前存在的难点与问题，提出了一种基于麻雀搜索优化的注意力门控循环单元预测方法。首先，应用注意力机制对输入序列进行权重分配；然后，输入门控循环单元组合网络对内部特征进行学习，并输出预测时间负荷值；最后，使用麻雀搜索算法对网络超参数进行组合优化，以验证集损失最小为目标函数获取最优化网络结构超参数。该方法实现了原始输入序列结构权重分配与组合网络超参数的最优化。算例分析表明，所提方法比传统预测模型精确度更高。
Abstract：Aimed at the existing difficulties and problems in short-term power load forecasting,a forecasting method based on sparrow search algorithm(SSA)optimized Attention-gated recurrent unit(GRU)is proposed. First,the Attention mechanism is applied to assign weights to input sequences. Then,the updated sequences are imported into the combined GRU network to learn the internal characteristics,and the forecasted load values in the corresponding time are output. Finally,SSA is used to optimize the combination of network hyperparameters,and the optimal network structure hyperparameters are obtained through minimizing the loss of the validation set. The proposed method realizes structural weights distribution of the original input sequences and the optimization of combined network hyperparameters.The analysis of a numerical example shows that compared with the traditional forecasting models,the proposed method has a higher accuracy.
[1]李焱，贾雅君，李磊，等(Li Yan,Jia Yajun,Li Lei,et al).基于随机森林算法的短期电力负荷预测(Short term power load forecasting based on a stochastic forest algorithm)[J].电力系统保护与控制(Power System Protection and Control),2020,48(21):117-124.
[2]Song Kyung-Bin,Baek Young-Sik,Hong Dug-Hun,et al.Short-term load forecasting for the holidays using fuzzy linear regression method[J].IEEE Trans on Power Systems,2005,20(1):96-101.
[3]Xu Wenquan,Hu Hui,Yang Wei.Energy time series forecasting based on empirical mode decomposition and FRBF-AR model[J].IEEE Access,2019,7:36540-36548.
[4]Huang Shyh-Jier,Shih Kuang-Rong.Short-term load forecasting via ARMA model identification including nonGaussian process considerations[J].IEEE Transactions on Power Systems,2003,18(2):673-679.
[5]Ervin C,Vladimir C,Adrijan Baric.A strategy for shortterm load forecasting by support vector regression machines[J].IEEE Trans on Power Systems,2013,28(4):4356-4364.
[6]Luis H,Carlos B,Javier A,et al.Artificial neural network for short-term load forecasting in distribution systems[J].Energies,2014,7(3):1576-1598.
[7]王静，李维德(Wang Jing,Li Weide).基于CEEMD和GWO的超短期风速预测(Ultra-short-term forecasting of wind speed based on CEEMD and GWO)[J].电力系统保护与控制(Power System Protection and Control),2018,46(9):69-74.
[8]王增平，赵兵，纪维佳，等(Wang Zengping,Zhao Bing,Ji Weijia,et al).基于GRU-NN模型的短期负荷预测方法(Short-term load forecasting method based on GRU-NNmodel)[J].电力系统自动化(Automation of Electric Power Systems),2019,43(5):53-58.
[9]赵兵，王增平，纪维佳，等(Zhao Bing,Wang Zengping,Ji Weijia,et al).基于注意力机制的CNN-GRU短期电力负荷预测方法(A short-term power load forecasting method based on attention mechanism of CNN-GRU)[J].电网技术(Power System Technology),2019,43(12):4370-4376.
[10]孔祥玉，郑锋，鄂志君，等(Kong Xiangyu,Zheng Feng,EZhijun,et al).基于深度信念网络的短期负荷预测方法(Short-term load forecasting based on deep belief network)[J].电力系统自动化(Automation of Electric Power Systems),2018,42(5):133-139.
[11]张健(Zhang Jian).基于自适应K-means与DNN的短期负荷预测研究分析(Research and analysis of short-term load forecasting based on adaptive K-means and DNN)[J].电子测量技术(Electronic Measurement Technology),2020,43(17):58-61.
[12]陆继翔，张琪培，杨志宏，等(Lu Jixiang,Zhang Qipei,Yang Zhihong,et al).基于CNN-LSTM混合神经网络模型的短期负荷预测方法(Short-term load forecasting method based on CNN-LSTM hybrid neural network model)[J].电力系统自动化(Automation of Electric Power Systems),2019,43(8):131-137.
[13]郑瑞骁，张姝，肖先勇，等(Zheng Ruixiao,Zhang Shu,Xiao Xianyong,et al).考虑温度模糊化的多层长短时记忆神经网络短期负荷预测(Short-term load forecasting of multi-layer long short-term memory neural network considering temperature fuzziness)[J].电力自动化设备(Electric Power Automation Equipment),2020,40(10):181-186.
[14]田浩含，张智晟，于道林(Tian Haohan,Zhang Zhisheng,Yu Daolin).基于改进LSTM的区域综合能源系统多元负荷短期预测研究(Research on multi-load short-term forecasting of regional integrated energy system based on improved LSTM)[J].电力系统及其自动化学报(Proceedings of the CSU-EPSA),2021,33(9):130-137.
[15]牛哲文，余泽远，李波，等(Niu Zhewen,Yu Zeyuan,Li Bo,et al).基于深度门控循环单元神经网络的短期风功率预测模型(Short-term wind power forecasting model based on deep gated recurrent unit neural network)[J].电力自动化设备(Electric Power Automation Equipment),2018,38(5):36-42.
[16]庞昊，高金峰，杜耀恒(Pang Hao,Gao Jinfeng,Du Yaoheng.)基于多神经网络融合的短期负荷预测方法(Short-term load forecasting method based on fusion of multiple neural networks).[J].电力自动化设备(Electric Power Automation Equipment),2020,40(6):37-42.
[17]冯斌，郭亦宗，陈页，等(Feng Bin,Guo Yizong,Chen Ye,et al).基于GRU多步预测技术的云储能充放电策略(Charging and discharging strategy of cloud energy storage based on GRU multi-step prediction technology)[J].电力系统自动化(Automation of Electric Power Systems),2021,45(9):46-54.
[18]Vaswani A,Shazeer N,Parmar N,et al.Attention is all you need[C]//Conference on Advances in Neural Information Processing.Long Beach,USA,2017:5999-6009.
[19]Yang Zichao,Yang Diyi,Dyer Chris,et al.Hierarchical attention networks for document classification[C]//Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies.San Diego,USA,2016:1480-1489.
[20]Xue Jiankai,Shen Bo.A novel swarm intelligence optimization approach:sparrow search algorithm[J].Systems Science and Control Engineering,2020,8(1):22-34.--------
基本信息:
DOI：10.19635/j.cnki.csu-epsa.000853
中图分类号:TP18;TM715
引用信息:
[1]刘可真,阮俊枭,赵现平,等.基于麻雀搜索优化的Attention-GRU短期负荷预测方法[J].电力系统及其自动化学报,2022,34(04):99-106.DOI:10.19635/j.cnki.csu-epsa.000853.
基金信息:
国家自然科学基金资助项目（51477100）; 云南电网有限责任公司科技项目（YNKJXM20180736）
2021-10-11
2021-10-11
2021-10-11