基于EEMD-CBAM-BiLSTM的牵引负荷超短期预测

doi:10.19595/j.cnki.1000-6753.tces.231515

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (1830 KB)
输出: BibTeX | EndNote (RIS)

摘要

针对电气化铁路牵引负荷难以预测的问题,构建了一种由集合经验模态分解（Ensemble Empirical Mode Decomposition,EEMD）、改进型卷积块注意力模块（Convolutional Block Attention Module,CBAM）和双向长短期神经网络（Bidirectional Long Short-Term Memory Neural Network,BiLSTM）组合成的EEMD-CBAM-BILSTM预测方法,有效降低了牵引负荷超短期预测误差与计算成本。首先,通过EEMD将牵引负荷数据分解为多个稳定、有规律的时序模态函数,突出负荷数据的时序特征;其次,将分解好的各分量整体通入由卷积神经网络（Convolutional Neural Networks, CNN）和改进型CBAM组成的特征提取模块提取全局时序特征;最后,利用贝叶斯优化（Bayesian Optimization,BO）搜寻BiLSTM最优参数,并将全局特征通入优化后的神经网络进行超短期时序预测。仿真算例表明,本文所提预测框架在各预测步长下均能很好地把握牵引负荷变化趋势,显著提升了牵引负荷预测的精度。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	钟吴君
	李培强
	涂春鸣

关键词 ：牵引负荷预测, 集合经验模态分解, 双向长短期神经网络, 贝叶斯优化, 卷积块注意力模块, 卷积神经网络

Abstract：

China has the world's largest rail transit network, with a total mileage of more than 150,000 kilometers, the electrification rate of more than 70 percent, and railway energy consumption is the largest single load category. With the rapid development of China's electrified railway network, the railway transportation represented by high-speed rail has become an important part of China's transportation system. With the development of "net-source-storage-vehicle" collaborative energy supply technology, the internal energy management and collaborative control of the system need the technical support of accurate second-level ultra-short-term prediction on both sides of the source and load. At the same time, accurate ultra-short-term prediction of traction load can also provide data source support for research on power quality analysis of rail transit, optimal scheduling of traction substation, location and capacity determination of traction substation, etc.
This paper constructs an ultra-short-term forecast framework for traction load based on EEMD-CBAM-BILSTM, aiming at the problem of difficulty in predicting rail transit traction load due to strong mutability and volatility: Firstly, based on the analysis of the time series characteristics of rail transit traction load, the data of rail transit traction load is decomposed into several stable and regular time series mode functions by ensemble empirical mode decomposition (EEMD) to highlight the time series characteristics of load data. Secondly, the decomposed components are integrated into the feature extraction module composed of convolutional neural network (CNN) and improved Convolutional block Attention module (CBAM) to extract the global timing features. After that, Bayesian optimization (BO) is used to search the optimal parameters of BiLSTM neural network to make the network structure reach the best state.In this paper, the traction load data in a single day (with a resolution of 1s) is selected to build a prediction model, and the model is compared with the mainstream single prediction model and composite model by starting from three output modes: predicting the load in the next 4 seconds, the load in the next 8 seconds, and the load in the next 16 seconds.
The calculation results show that compared to other models, the framework in this paper reduces MAE by 31.2%, 39.1%, 47.5%, and 45.8% in the 40-16 prediction mode, respectively; RMSE decreased by 40.0%, 42.6%, 53.6%, and 47.5% respectively; R² increased by 5.2%, 5.9%, 10.65%, and 7.7% respectively. Therefore, the following conclusion can be drawn: EEMD has good performance in avoiding modal aliasing and better highlighting the deep data relationships and temporal features hidden in time series. The improved CBAM feature extraction module and global feature extraction structure can significantly reduce the computational workload caused by the "decomposition prediction" method, improve the representation ability of hidden temporal features in traction load data, better capture key features in the data, and improve the accuracy of subsequent neural network predictions. After Bayesian optimization of hyperparameters, the BiLSTM model can better capture information in the front and back temporal directions when processing long feature sequences, fully considering the past and future temporal information based on the current point as the reference point. This article framework combines the advantages and characteristics of EEMD, CBAM, CNN, BO, and BiLSTM technologies, and can significantly improve the accuracy of ultra short term prediction of electrified railway traction loads at various prediction steps, which has good engineering application value.

Key words： Traction load forecasting EEMD BiLSTM BO CBAM CNN

收稿日期: 2023-09-13

PACS:

TM922.3

基金资助:

国家重点研发计划项目（2021YFB2601504）和国家自然科学基金项目（52377097）资助

通讯作者: 钟吴君男,1994年生,博士研究生,研究方向为负荷与新能源预测。E-mail：395236969@qq.com

作者简介: 李培强男,1975年生,博士,博士生导师,研究方向为智能电网及其分布式发电研究。E-mail：lpqcs@hnu.edu.cn

引用本文:

钟吴君, 李培强, 涂春鸣. 基于EEMD-CBAM-BiLSTM的牵引负荷超短期预测[J]. 电工技术学报, 0, (): 231515-. Zhong Wujun, Li Peiqiang, Tu Chunming. Traction Load Ultra-Short-Term Forecasting Framework Based on EEMD-CBAM-BiLSTM. Transactions of China Electrotechnical Society, 0, (): 231515-.

链接本文:

https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.231515 https://dgjsxb.ces-transaction.com/CN/Y0/V/I/231515

[1] 熊嘉阳, 沈志云. 中国高速铁路的崛起和今后的发展[J]. 交通运输工程学报, 2021, 21(5): 6-29.
Xiong Jiayang, Shen Zhiyun.Rise and future development of Chinese high-speed railway[J]. Journal of Traffic and Transportation Engineering, 2021, 21(5): 6-29.
[2] 唐圣德,胡海涛,肖冬华,等.基于阻抗测量信息的电气化铁路“多车共网”系统稳定性分析[J/OL].电工技术学报,1-15[2024-05-31].https://doi.org/10.19595/j.cnki.1000-6753.tces.231294.
Tang Shengde, Hu Haitao, Xiao Donghua, et al. Stability analysis of electrified railway "multi-vehicle common network" system based on impedance measurement information [J/OL]. Transactions of China Electrotechnical Society, 1-15 [2024-05-31]. https://doi.org/10.19595/j.cnki.1000-6753.tces.231294.
[3] 胡海涛, 郑政, 何正友, 等. 交通能源互联网体系架构及关键技术[J]. 中国电机工程学报, 2018, 38(1): 12-24, 339.
Hu Haitao, Zheng Zheng, He Zhengyou, et al.The framework and key technologies of traffic energy Internet[J]. Proceedings of the CSEE, 2018, 38(1): 12-24, 339.
[4] 高仕斌, 罗嘉明, 陈维荣, 等. 轨道交通“网-源-储-车” 协同供能技术体系[J/OL]. 西南交通大学学报, 2022: 1-18. (2022-12-01). https://kns.cnki.net/kcms/detail/51.1277.U.20221130.1759.003.html.
Gao Shibin, Luo Jiaming, Chen Weirong, et al. Rail transit ‘network-source-storage-vehicle’ collaborative energy supply technology system[J/OL]. Journal of Southwest Jiaotong University, 2022: 1-18. (2022-12-01). https://kns.cnki.net/kcms/detail/51.1277.U.20221130.1759.003.html.
[5] 罗嘉明, 高仕斌, 韦晓广, 等. 基于模糊Petri网的“网-源-储-车” 动态阈值能量管理策略研究[J]. 工程科学与技术, 2023, 55(1): 48-58.
Luo Jiaming, Gao Shibin, Wei Xiaoguang, et al.Research on “grid-source-storage-vehicle” dynamic threshold energy management based on fuzzy petri nets[J]. Advanced Engineering Sciences, 2023, 55(1): 48-58.
[6] Liu Yuanli, Chen Minwu, Cheng Zhe, et al.Robust energy management of high-speed railway co-phase traction substation with uncertain PV generation and traction load[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(6): 5079-5091.
[7] 李俊豪, 涂春鸣, 王鑫, 等. 基于“规则+优化” 的电气化铁路站点实时能量管控策略[J]. 电工技术学报, 2024, 39(11): 3339-3352.
Li Junhao, Tu Chunming, Wang Xin, et al.Real-time energy management strategy for electrified railroad stations based on “rules + optimization”[J]. Transactions of China Electrotechnical Society, 2024, 39(11): 3339-3352.
[8] 魏波, 胡海涛, 王科, 等. 基于实测数据和行车运行图的高铁牵引变电站负荷预测方法[J]. 电工技术学报, 2020, 35(1): 179-188.
Wei Bo, Hu Haitao, Wang Ke, et al.Research on traction load forecasting method for high-speed railway traction substation based on measured data and train timetable[J]. Transactions of China Electrotechnical Society, 2020, 35(1): 179-188.
[9] Langerudy A T, Mariscotti A, Abolhassani M A.Power quality conditioning in railway electrification: a comparative study[J]. IEEE Transactions on Vehicular Technology, 2017, 66(8): 6653-6662.
[10] Wang Xiaoyu, Han Ying, Li Luoyi, et al.CVaR quantitative uncertainty-based optimal dispatch for flexible traction power supply system[J]. IEEE Transactions on Transportation Electrification, 2024, 10(1): 1900-1910.
[11] Chen Minwu, Cheng Zhe, Liu Yuanli, et al.Multitime-scale optimal dispatch of railway FTPSS based on model predictive control[J]. IEEE Transactions on Transportation Electrification, 2020, 6(2): 808-820.
[12] 薛艳冰, 马大炜, 王烈. 列车牵引能耗计算方法[J]. 中国铁道科学, 2007, 28(3): 84-87.
Xue Yanbing, Ma Dawei, Wang Lie.Calculation method of energy consumption in train traction[J]. China Railway Science, 2007, 28(3): 84-87.
[13] 王琪. 川藏铁路再生制动能量利用方案研究[D]. 北京: 北京交通大学, 2021.
Wang Qi.Study on energy utilization scheme of regenerative braking in sichuan-tibet railway[D].Beijing: Beijing Jiaotong University, 2021.
[14] 王科, 胡海涛, 魏文婧, 等. 基于列车运行图的高速铁路动态牵引负荷建模方法[J]. 中国铁道科学, 2017, 38(1): 102-110.
Wang Ke, Hu Haitao, Wei Wenjing, et al.Modelling method for dynamic traction load of high speed railway based on train working diagram[J]. China Railway Science, 2017, 38(1): 102-110.
[15] 刘福, 廖启术. 电气化铁路牵引负荷预测研究[J]. 机车电传动, 2023(2): 142-150.
Liu Fu, Liao Qishu.Research on traction load forecasting of electrified railway[J]. Electric Drive for Locomotives, 2023(2): 142-150.
[16] Hong Tao, Gui Min, Baran M E, et al.Modeling and forecasting hourly electric load by multiple linear regression with interactions[C]//IEEE PES General Meeting, Minneapolis, MN, USA, 2010: 1-8.
[17] Lee Chengming, Ko C N.Short-term load forecasting using lifting scheme and ARIMA models[J]. Expert Systems with Applications, 2011, 38(5): 5902-5911.
[18] 刘达. 基于误差校正的中长期负荷预测模型[J]. 电网技术, 2012, 36(8): 243-247.
Liu Da.A model for medium-and long-term power load forecasting based on error correction[J]. Power System Technology, 2012, 36(8): 243-247.
[19] 赵洋, 王瀚墨, 康丽, 等. 基于时间卷积网络的短期电力负荷预测[J]. 电工技术学报, 2022, 37(5): 1242-1251.
Zhao Yang, Wang Hanmo, Kang Li, et al.Temporal convolution network-based short-term electrical load forecasting[J]. Transactions of China Electrotechnical Society, 2022, 37(5): 1242-1251.
[20] 王增平, 赵兵, 纪维佳, 等. 基于GRU-NN模型的短期负荷预测方法[J]. 电力系统自动化, 2019, 43(5): 53-58.
Wang Zengping, Zhao Bing, Ji Weijia, et al.Short-term load forecasting method based on GRU-NN model[J]. Automation of Electric Power Systems, 2019, 43(5): 53-58.
[21] 李元诚, 方廷健, 于尔铿. 短期负荷预测的支持向量机方法研究[J]. 中国电机工程学报, 2003, 23(6): 55-59.
Li Yuancheng, Fang Tingjian, Yu Erkeng.Study of support vector machines for short-term load forecasting[J]. Proceedings of the CSEE, 2003, 23(6): 55-59.
[22] 孔祥玉, 李闯, 郑锋, 等. 基于经验模态分解与特征相关分析的短期负荷预测方法[J]. 电力系统自动化, 2019, 43(5): 46-52.
Kong Xiangyu, Li Chuang, Zheng Feng, et al.Short-term load forecasting method based on empirical mode decomposition and feature correlation analysis[J]. Automation of Electric Power Systems, 2019, 43(5): 46-52.
[23] 张淑清, 李君, 姜安琦, 等. 基于FPA-VMD和BiLSTM神经网络的新型两阶段短期电力负荷预测[J]. 电网技术, 2022, 46(8): 3269-3279.
Zhang Shuqing, Li Jun, Jiang Anqi, et al.A novel two-stage model based on FPA-VMD and BiLSTM neural network for short-term power load forecasting[J]. Power System Technology, 2022, 46(8): 3269-3279.
[24] 邓带雨, 李坚, 张真源, 等. 基于EEMD-GRU-MLR的短期电力负荷预测[J]. 电网技术, 2020, 44(2): 593-602.
Deng Daiyu, Li Jian, Zhang Zhenyuan, et al.Short-term electric load forecasting based on EEMD-GRU-MLR[J]. Power System Technology, 2020, 44(2): 593-602.
[25] 陈映月. 基于数据挖掘的牵引负荷统计方法及应用研究[D]. 成都: 西南交通大学, 2019.
Chen Yingyue.Research on traction load statistics method and application based on data mining[D]. Chengdu: Southwest Jiaotong University, 2019.
[26] 高锋阳, 宋志翔, 高建宁, 等. 计及光伏和储能接入的牵引供电系统能量管理策略[J]. 电工技术学报, 2024, 39(3): 745-757.
Gao Fengyang, Song Zhixiang, Gao Jianning, et al.Energy management strategies for traction power systems with PV and energy storage access[J]. Transactions of China Electrotechnical Society, 2024, 39(3): 745-757.
[27] 张浩, 熊浩清, 陈谦, 等. 电铁牵引负荷的构成分析及其功率模型[J]. 电力系统及其自动化学报, 2015, 27(6): 37-42.
Zhang Hao, Xiong Haoqing, Chen Qian, et al.Component analysis and power model of traction load[J]. Proceedings of the CSU-EPSA, 2015, 27(6): 37-42.
[28] Woo S, Park J, Lee J Y, et al.CBAM: convolutional block attention module[C]//European conference on computer vision. Cham: Springer, 2018: 3-19.
[29] Bergstra J, Bardenet R, Bengio Y, et al.Algorithms for hyper-parameter optimization[J]. Advances in neural information processing systems, 2011, 24.