| 494 | 0 | 66 |
| 下载次数 | 被引频次 | 阅读次数 |
针对综合能源系统负荷数据的高波动性导致自身关联性难以有效挖掘、预测精确度不足等问题，提出一种基于相似周和自适应二次分解的多元负荷预测方法。首先，通过综合相似距离法筛选相似周序列，并利用最大互信息数方法筛选出相关性较强的特征，以构建相似周数据集；然后，采用改进的完全自适应噪声集成经验模态分解将负荷数据分解成不同的本征模态函数，针对分解得到的高频分量进行变分模态分解，以降低序列的不平稳性，并引入冠豪猪优化算法对变分模态分解的分解数量及惩罚因子进行优化，实现变分模态分解的自适应性；最后，将各本征模态函数分量与气象信息结合，输入时序卷积网络与双向门控循环单元进行预测。研究结果表明：相较于原始数据集，采用相似周数据集的WMAPE为0.889%，降低了0.513个百分点。
Abstract：Due to the high volatility of load data in an integrated energy system,which makes it difficult to effectively mine the internal correlation and results in an insufficient prediction accuracy,a multivariate load forecasting method based on similar weeks and adaptive quadratic decomposition is proposed in this paper. First,similar week sequences are selected using a comprehensive similarity distance method,and the strongly correlated features are filtered out using the maximal information coefficient to construct the similar week dataset. Second,the load data is decomposed into different intrinsic mode functions(IMFs)using the improved complete ensemble empirical mode decomposition with adaptive noise. The resultant high-frequency components obtained from the decomposition are then processed using variational mode decomposition(VMD)to address the non-stationarity of the sequence. To achieve the adaptability in VMD,the Crested Porcupine Optimizer is applied to optimize the number of decompositions and penalty factors. Finally,each IMF component is combined with the meteorological information and further input into a temporal convolutional network and a bidirectional gated recurrent unit for prediction. Results show that compared with that based on the original dataset,the WMAPEbased on the similar week dataset is 0.889%,which means a reduction of 0.513 percentage point.
[1]李鹏，罗湘淳，孟庆伟，等（Li Peng,Luo Xiangchun,Meng Qingwei,et al）.基于Spearman相关性阈值寻优和VMD-LSTM的用户级综合能源系统超短期负荷预测（Ultra short-term load forecasting of user level integrated energy system based on spearman threshold optimization and variational mode decomposition and long short-term memory）[J].全球能源互联网（Journal of Global Energy Interconnection）,2024,7(4):406-420.
[2]马建鹏，龚文杰，张智晟（Ma Jianpeng,Gong Wenjie,Zhang Zhisheng）.基于Copula理论与KPCA-GRNN结合的区域综合能源系统多元负荷短期预测模型（Shortterm multiple load prediction model for regional integrated energy system based on Copula theory and KPCAGRNN）[J].电工电能新技术（Advanced Technology of Electrical Engineering and Energy）,2020,39(3):24-31.
[3]付文龙，章轩瑞，张海荣，等（Fu Wenlong,Zhang Xuanrui,Zhang Hairong,et al）.多尺度特征提取与非线性融合的综合能源系统多元负荷短期预测（Short-term multivariate load forecasting of integrated energy system based on multiscale feature extraction and nonlinear fusion）[J].电力系统及其自动化学报（Proceedings of the CSU-EPSA）,2023,35(12):89-99.
[4]秦烁，赵健，徐剑，等（Qin Shuo,Zhao Jian,Xu Jian,et al）.基于多任务学习和单任务学习组合模型的综合能源系统多元负荷预测（Multivariate-load forecasting of integrated energy system based on combined multi-task learning and single-task learning model）[J].电网技术（Power System Technology）,2024,48(4):1510-1518.
[5]祁宇轩，范俊岩，吴定会，等（Qi Yuxuan,Fan Junyan,Wu Dinghui,et al）.基于相似日与BiLSTM组合的短期电力负荷预测（Short term power load forecasting based on the combination of similar days and BiLSTM）[J].控制理论与应用（Control Theory&Applications）,2024,41(12):2304-2314.
[6]吴伟杰，吴杰康，雷振，等（Wu Weijie,Wu Jiekang,Lei Zhen,et al）.基于CSO优化深度信念网络的园区能源需求预测方法（Park energy demand forecasting based on CSO optimized deep belief network）[J].电网技术（Power System Technology）,2021,45(10):3859-3868.
[7]张大海，孙锴，和敬涵（Zhang Dahai,Sun Kai,He Jinghan）.基于相似日与多模型融合的短期负荷预测（Short-term load forecasting based on similar day and multi-model fusion）[J].电网技术（Power System Technology）,2023,47(5):1961-1969.
[8]朱刘柱，王绪利，马静，等（Zhu Liuzhu,Wang Xuli,Ma Jing,et al）.基于小波包分解与循环神经网络的综合能源系统短期负荷预测（Short-term load forecast of integrated energy system based on wavelet packet decomposition and recurrent neural network）[J].电力建设（Electric Power Construction）,2020,41(12):131-138.
[9]杨维熙，刘勇，舒勤（Yang Weixi,Liu Yong,Shu Qin）.基于补充集合经验模态分解的短期负荷预测模型（A short-term load forecasting model based on CEEMD）[J].电网技术（Power System Technology）,2022,46(9):3615-3622.
[10]黄璜，张安安（Huang Huang,Zhang An’an）.基于分解算法与元学习结合的综合能源系统负荷预测（Load forecasting of integrated energy system based on combination of decomposition algorithms and meta-learning）[J].电力系统自动化（Automation of Electric Power Systems）,2024,48(10):151-160.
[11]张淑娴，江文韬，陈玉花，等（Zhang Shuxian,Jiang Wentao,Chen Yuhua,et al）.基于二次模态分解的LSTM短期电力负荷预测（Short-term power load forecasting based on quadratic mode decomposition through LSTM）[J].科学技术与工程（Science Technology and Engineering）,2024,24(7):2759-2766.
[12]魏步晗，鲍刚，李振华（Wei Buhan,Bao Gang,Li Zhenhua）.基于支持向量回归预测模型考虑天气因素和分时电价因素的短期电力负荷预测（Short-term electricity load forecasting based on support vector regression forecasting model considering weather factors and time-of-use tariff factors）[J].电网与清洁能源（Power System and Clean Energy）,2023,39(11):9-19.
[13]龙干，黄媚，方力谦，等（Long Gan,Huang Mei,Fang Liqian,et al）.基于改进多元宇宙算法优化ELM的短期电力负荷预测（Short-term power load forecasting based on an improved multi-verse optimizer algorithm optimized extreme learning machine）[J].电力系统保护与控制（Power System Protection and Control）,2022,50(19):99-106.
[14] Yao Xiaotong,Fu Xiaoli,Zong Chaofei. Short-term load forecasting method based on feature preference strategy and LightGBM-XGboost[J]. IEEE Access,2022,10:75257-75268.
[15]朱继忠，董瀚江，李盛林，等（Zhu Jizhong,Dong Hanjiang,Li Shenglin,et al）.数据驱动的综合能源系统负荷预测综述（Review of data-driven load forecasting for integrated energy system）[J].中国电机工程学报（Proceedings of the CSEE）,2021,41(23):7905-7923.
[16]彭曙蓉，陈慧霞，孙万通，等（Peng Shurong,Chen Huixia,Sun Wantong,et al）.基于改进LSTM的光伏发电功率预测方法研究（Research on photovoitaic power prediction method based on improved LSTM）[J].太阳能学报（Acta Energiae Solaris Sinica）,2024,45(11):296-302.
[17]薛东，段立强，高统彤，等（Xue Dong,Duan Liqiang,Gao Tongtong,et al）.考虑多重特征与不确定性度量的综合能源系统负荷预测研究（Study of integrated energy system load forecasting considering multiple characteristics and uncertainty measures）[J].太阳能学报（Acta Energiae Solaris Sinica）,2024,45(7):379-388.
[18]吴晨，姚菁，薛贵元，等（Wu Chen,Yao Jing,Xue Guiyuan,et al）.基于MMoE多任务学习和长短时记忆网络的综合能源系统负荷预测（Load forecasting of integrated energy system based on MMoE multi-task learning and LSTM）[J].电力自动化设备（Electric Power Automation Equipment）,2022,42(7):33-39.
[19] Dragomiretskiy K,Zosso D. Variational mode decomposition[J]. IEEE Transactions on Signal Processing,2014,62(3):531-544.
[20] Mohamed A,Reda M,Mohamed A. Crested porcupine optimizer:A new nature-inspired metaheuristic[J]. Knowledge-Based Systems,2024,284:111257.
[21] ASU. Campus metabolism[EB/OL].(2020-12-31)[2024-12-01]https：//cm.asu.edu/，2024.
[22] NSRDB. Data viewer[EB/OL].(2020-12-31)[2024-12-01]https：//nsrdb.nrel.gov/data-viewer,2024.
基本信息:
DOI：10.19635/j.cnki.csu-epsa.001613
中图分类号:TM715
引用信息:
[1]潘鹏程,孙龙华,胡继岚,等.基于相似周和自适应二次分解的综合能源系统多元负荷预测[J].电力系统及其自动化学报,2025,37(11):62-71.DOI:10.19635/j.cnki.csu-epsa.001613.
基金信息:
国家自然科学基金资助项目（52307109）