基于VMD-SSA-Transformer-GRU模型的高压输电塔钢管微风振动预测

程燕军; 苏仁斌; 熊卫红; 刘先珊; 黄子宣; 曹伊婷

doi:10.3724/j.gyjzG25041405

基于VMD-SSA-Transformer-GRU模型的高压输电塔钢管微风振动预测

doi: 10.3724/j.gyjzG25041405

1. 国家电网有限公司华中分部, 武汉 430072;
2. 重庆大学土木工程学院, 重庆 400045

基金项目:

国家电网公司华中分部科技项目（52140024000K）。

详细信息

作者简介:
程燕军，高级工程师，主要从事电力系统自动化方向研究，chengyj@cc.sgcc.com.cn。

通讯作者:
刘先珊，博士，教授，主要从事岩土工程风险评估及数值方法方向研究，lzmoumou@163.com。

计量
- 文章访问数: 21
- HTML全文浏览量: 7
- PDF下载量: 0
- 被引次数: 0
出版历程
- 收稿日期: 2025-04-14
- 网络出版日期: 2026-08-31
- 刊出日期: 2026-07-20

Prediction of Breeze Vibration with VMD-SSA-Transformer-GRU Model of Steel Tubular Members for a UHV Transmission Tower

1. Centralchina Branch of State Grid Corporation of China, Wuhan 430072, China;
2. School of Civil Engineering, Chongqing University, Chongqing 400045, China

摘要

摘要: 高压钢管输电塔涡激振动时有发生，诱导构件疲劳破坏，安全隐患大。鉴于传统风洞试验和数值模拟方法的局限性，基于典型输电塔钢管构件的涡激振动时间序列，创新性地提出VMD-SSA-Transformer-GRU模型，以此开展微风环境中输电塔钢管构件的涡激振动预测研究，更好地捕捉振动时序的大范围波动过程。为此，基于三类模型进行了不同训练样本数的正弦函数仿真研究，5个评价指标的对比分析说明所提出的模型能更快收敛于最优解。建立了对应的钢管构件涡激振动预测模型，对于单一时间序列预测，其平均绝对误差MAE、均方差MSE、根均方差RMSE、决定系数R²的数值分别为0.0029653、2.6566×10^-5、0.005154、0.96143，阐明了该模型对高维非线性微风振动时序预测的可靠性；并将其用于多时间序列预测，得到MAE、MSE、RMSE、R²的数值分别为0.0020449、1.7791×10^-5、0.004218、0.97373，该结果相对于单变量模型的精度更高，对比分析了时间序列的频率曲线，进而验证了多变量模型的高精度、可靠性及泛化能力。综上所述，多变量VMD-SSA-Transformer-GRU模型能精准捕捉输电塔钢管微风振动时序的特征，实现不同工况的微风振动响应预测研究，且在计算效率上相对计算流体力学方法（CFD）具有明显的优势。
- 高压输电塔 /
- 微风振动 /
- 钢管构件 /
- 时间序列 /
- VMD-SSA-Transformer-GRU模型
Abstract: Vortex-induced vibrations frequently occur in ultra-high voltage （UHV） steel tubular towers， inducing fatigue damage in structural components and posing significant safety hazards， particularly given the limitations of traditional wind tunnel testing and numerical simulation method， time series induced by the vortex-induced vibrations for a typical steel tubular member from a large-span transmission tower， as a case study， VMD-SSA-Transformer-GRU has been creatively proposed to predict the vortex-induced vibration to better capture the wide-range fluctuations in vortex-induced vibration monitoring data of transmission towers. Firstly， based on the three models with different groups and iterations， the sine functions with different sample sizes have been simulated， indicating that the simulated values based on the proposed model are more accurate and have faster convergence to approach the theoretical values when compared with the five assessment indices. Furthermore， a prediction model has been proposed to predict the vortex-induced vibration， as for a single series， Mean Absolute Error （MAE）， the Mean Square Error （MSE）， the Root Mean Square Error （RMSE） and higher coefficient of determination （R²） are 0.0029653，2.6566×10^-5，0.005154 and 0.96143 respectively， which indicating the reliability of the proposed model to predict the vibration with high dimensions and nonlinearity. And then， the proposed model has been considered for multi-variable series， MAE， MSE， RMSE and R² are 0.0020449，1.7791×10^-5，0.004218 and 0.97373， respectively， it can be observed that the predictions is better than that with single series， and also comparing with the frequency curves of the time series， deeply verifying the high accuracy and reliability of the multivariable VMD-SSA-Transformer-GRU model. In summary， the multivariable model can effectively capture the fluctuation of steel tubular members for a UHV transmission tower to realize the vibration prediction considering different conditions， and has obvious superiority in calculation efficiency for CFD methods.
- UHV transmission towers /
- breeze vibration /
- steel tubular members /
- time series /
- VMD-SSA-Transformer-GRU model

HTML全文

参考文献(24)

[1]	国家能源局电力规划设计总院.中国电力发展报告2024[R/OL].[2024-07-16 ]. https://www.nea.gov.cn/2024-07/19/c_1310782066.htm.
[2]	杨靖波，李正.输电线路钢管塔微风振动及其对结构安全性的影响[J].振动、测试与诊断，2008，3（4）:23-30.
[3]	MANNINI C. Incorporation of turbulence in a nonlinear wake-oscillator model for the prediction of unsteady galloping response[J]. Journal of Wind Engineering and Industrial Aerodynamics，2020，200:104141.
[4]	赵张峰，邓洪洲.输电塔钢管构件涡振疲劳临界起振风速的探讨[J].特种结构，2015，32（5）:54-61.
[5]	HUANG M F，ZHANG B Y，GUO Y，et al. Prediction and suppression of vortex-induced vibration for steel tubes with bolted joints in tubular transmission towers[J]. Journal of Structural Engineering，2021，147（9）:04021128
[6]	DENG H Z，JIANG Q，LI F，et al. Vortex-induced vibration tests of circular cylinders connected with typical joints in transmission towers[J]. Journal of Wind Engineering and Industrial Aerodynamics，2011，99（10）:1069-1078.
[7]	邓洪洲，赵张峰.输电塔钢管构件涡激振动数值模拟[J].同济大学学报（自然科学版），2017，45（1）:9-15.
[8]	LIAO H，MEI H，HU G，et al. Machine learning strategy for predicting flutter performance of streamlined box girders[J]. Journal of Wind Engineering and Industrial Aerodynamics，2021，209:104493.
[9]	LI S，LAIMA S，LI H. Physics-guided deep learning framework for predictive modeling of bridge vortex induced vibrations from field monitoring[J]. Journal of Physics of Fluids，2021，33（3）:037113.
[10]	JIN X，CHENG P，CHEN W L，et al. Prediction model of velocity field around circular cylinder over various Reynolds numbers by fusion convolutional neural networks based on pressure on the cylinder[J]. Journal of Physics of Fluids，2018，30（4）:047105.
[11]	CHENG C，MENG H，LI Y Z，et al. Deep learning based on PINN for solving 2 DOF vortex induced vibration of cylinder[J]. Journal of Ocean Engineering，2021，240:109932.
[12]	高喜峰，翟立宾，徐万海.应用BP神经网络的柔性圆柱涡激振动预报[J].哈尔滨工程大学学报，2020，41（8）:1150-1155.
[13]	FOSTER J A，DE TROYER T，RUNACRES M C，et al. Modelling vortex-induced loads using recurrent neural networks[J]. IFAC-Paperson Line，2021，54（20）:32-37.
[14]	DENG Z，CHEN Y，LIU Y，et al. Time-resolved turbulent velocity field reconstruction using a long short-term memory（LSTM）-based artificial intelligence framework[J]. Physics of Fluids，2019，31（7）:075108.
[15]	李佳鸿，李正良，王涛.基于神经网络的输电塔钢管构件涡激振动幅值预测方法[J].工程力学，2024，41（1）:64-75
[16]	LI Z X，LIU H Y. A multivariate time series forecasting method incorporating global and serial features[J]. Journal of Computing，2023，46（1）:70-84.
[17]	李兆玺，刘红岩.融合全局和序列特征的多变量时间序列预测方法[J].计算机学报，2023，46（1）:70-84.
[18]	GALASSI A，LIPPI M，TORRONI P. Attention in natural language processing[J]. IEEE Transactions on Neural Networks and Learning Systems，2020，32（10）:4291-4308.
[19]	ZHANG T. BMT-Net:broad multitask transformer network for sentiment analysis[J]. IEEE Transactions on Cybernetics，2021，52（7）:6232-6243.
[20]	ARIK S O，YODER N C，PFISTER T. Self-adaptive forecasting for improved deep learning on non-stationary time-series[J]. arXiv preprint arXiv:2202.02403，2022.
[21]	WU H，XU J，WANG J，et al. Autoformer:decomposition transformers with auto-correlation for long-term series forecasting[J]. Advances in Neural Information Processing Systems，2021，34:22419-22430.
[22]	CHEN H，JIANG D，SAHLI H. Transformer encoder with multi-modal multi-head attention for continuous affect recognition[J]. IEEE Transactions on Multimedia，2020，23:4171-4183.
[23]	李宏扬，高丙朋.基于改进 VMD和 SAS-Attention-GRU的短期光伏发电功率预测[J].太阳能学报，2023，44（8）:292-300.
[24]	国家能源局.架空输电线路杆塔结构设计技术规程:DL/T 5486—2020[S].北京:中国计划出版社，2020.