| 306 | 2 | 486 |
| 下载次数 | 被引频次 | 阅读次数 |
因风电机组机舱测风仪及风向标代表轮前风况存在较大误差，影响机组功率调节效果。针对上述问题，构建基于LSTM模型的轮前风速映射方法并在河北某风电场进行激光雷达测风实验验证。首先，根据测风实验测得实际轮前风速；其次，分析机组SCADA系统与激光雷达测量风速的相关性；然后，以激光雷达测风数据和SCADA系统数据为基础划分训练集、测试集及验证集，构建LSTM模型来训练得到机舱测速和有效来流风速之间的传递关系。结果表明：映射后得到的风速与实际激光雷达测风装置测得的风速误差很小，实际风电场随机选取的3台风电机组的MAPE值分别为3.664%、3.395%、3.935%，同时与LSSVR及Lenet-5算法计算结果作比较，证明该方法的有效性。该研究对提高风电机组控制系统准确性以及风电场风功率预测等都具有一定的参考意义。
Abstract：Because there is a large error in the wind condition in front of the turbine represented by the wind meter and wind vane in the wind turbine engine cabin, which affects the power regulation effect of the turbine.Aiming at the above problems, a mapping method of front wheel wind speed based on LSTM model is constructed and verified by lidar wind measurement experiment in a wind farm in Hebei Province. Firstly, the actual wind speed in front of the wheel is measured according to the wind measurement experiment. Secondly,the correlation between the turbine SCADA system and the wind speed measured by the laser radar is analyzed. Then, based on the wind measurement data of the laser radar and the data of the SCADA system, the training set, the test set and the verification set are divided, and the LSTM model is constructed to train the transmission relationship between the engine cabin speed measurement and the effective incoming wind speed.The results show that the error between the mapped wind speed and the wind speed measured by the actual lidar wind measuring device is very small. The MAPE values of the three wind turbines randomly selected in the actual wind farm are 3.664%, 3.395% and 3.935% respectively. At the same time, the effectiveness of this method is proved by comparing with the calculation results of LSSVR and Lenet-5 algorithm. This research is of certain reference significance to improve the accuracy of wind turbine control system and wind power prediction of wind farms.
[1]陈琦,李红伟,周海林.考虑风电消纳的电-热综合能源系统经济运行研究[J].中国测试,2022,48(1):116-121.CHEN Q,LI H W,ZHOU H L.Study on economic operation of electricity-heat integrated energy system considering wind power consumption[J].China Measurement&Test,2022,48(1):116-121.
[2]陈泽慧,李博,李博.考虑加权理论的风电场集群风速预测方法[J].国外电子测量技术,2021,40(10):34-39.CHEN Z H,LI B,LI B.Wind speed prediction method for wind farm cluster considering weighted theory[J].Foreign Electronic Measurement Technology,2021,40(10):34-39.
[3]赵文婷.并网型微电网源荷预测及优化运营管理研究[D].太原:太原理工大学,2021.ZHAO W T.Research on power generation forecasting,load forecasting and optimal operation management of gridconnected microgrid[D].Taiyuan:Taiyuan University of Technology,2021.
[4]韩江北,刘志坚.基于虚拟同步控制的双馈式风机组并网控制策略的仿真分析[J].电子测量技术,2019,42(4):47-52.HAN J B,LIU Z J.Simulation analysis of grid connected control strategy of doubly fed wind turbine based on virtual synchronization control[J].Electronic Measurement Technology,2019,42(4):47-52.
[5]PAN L,XIONG Y,ZHU Z,et al.Research on variable pitch control strategy of direct-driven offshore wind turbine using KELM wind speed soft sensor[J].Renewable Energy,2022,184:1002-1017.
[6]GAO X,WANG T,LI B,et al.Investigation of wind turbine performance coupling wake and topography effects based on LiDAR measurements and SCADA data[J].Applied Energy,2019,255:113816.
[7]HAN B,XIE H,SHAN Y,et al.Characteristic curve fitting method of wind speed and wind turbine output based on abnormal data cleaning[J].Journal of Physics:Conference Series,2022,2185(1):012085.
[8]杨明明.基于卷积神经网络的机舱风速修正[J].华电技术,2021,43(5):75-79.YANG M M.Wind speed correction for wind turbine based on convolutional neural network[J].Huadian Technology,2021,43(5):75-79.
[9]KOSANA V,TEEPARTHI K,MADASTHU S.Hybrid wind speed prediction framework using data pre-processing strategy based autoencoder network[J].Electric Power Systems Research,2022,206:107821.
[10]苏文涛,温彩凤,杜乾,等.风电系统功率脉动特性分析及平抑策略探究[J].中国测试,2022,48(11):95-100.SU W T,WEN C F,DU Q,et al.Analysis of power pulsation characteristics of wind power system and exploration of mitigation strategies[J].China Measurement&Test,2022,48(11):95-100.
[11]CHEN Y,WANG Y,DONG Z,et al.2-D regional short-term wind speed forecast based on CNN-LSTM deep learning model[J].Energy Conversion and Management,2021,244(4):114451.
[12]LIANG T,ZHAO Q,LV Q,et al.A novel wind speed prediction strategy based on Bi-LSTM,MOOFADA and transfer learning for centralized control centers[J].Energy,2021,230:120904.
[13]赵文凯,赵世军,单雨龙,等.激光测风雷达风场探测性能评估[J].中国测试,2022,48(1):147-153.ZHAO W K,ZHAO S J,SHAN Y L,et al.Evaluation of wind detection performance based on wind lidar[J].China Measurement&Test,2022,48(1):147-153.
[14]杨芮,文武,徐虹.基于PCC-CNN-GRU的短期风电功率预测[J].成都信息工程大学学报,2022,37(2):165-170.YANG R,WEN W,XU H.Short-term wind power prediction based on PCC-CNN-GRU[J].Journal of Chengdu University of Information Technology,2022,37(2):165-170.
基本信息:
DOI：10.11857/j.issn.1674-5124.2022090169
中图分类号:TM315;TP183
引用信息:
[1]郜宁,高晓霞,吴茂乾,等.基于LSTM的风电机组轮前风速映射方法研究[J].中国测试,2025,51(05):155-161.DOI:10.11857/j.issn.1674-5124.2022090169.
基金信息:
国家自然科学基金项目（52076081）; 国家电网有限公司科技项目资助（SGXJXN00TSJS2200142）
2022-09-28
2022-11-15
2025-05-20
2025-05-20