基于三种神经网络的覆冰输电导线气动力系数预测研究  

A Prediction Study of Aerodynamic Coefficients for Iced Conductor Based on Three Neural Networks

作　　者：蔡萌琦 梁俊豪 闵光云 包婉玉 周林抒 胡茂明 CAI Mengqi;LIANG Junhao;MIN Guangyun;BAO Wanyu;ZHOU Linshu;HU Maoming(Failure Mechanics and Engineering Disaster Prevention and Mitigation Key Laboratory of Sichuan Province,Sichuan University,Chengdu 610065,Sichuan,China;School of Architecture and Civil Engineering,Chengdu University,Chengdu 610106,Sichuan,China;School of Mechanical Engineering,Chengdu University,Chengdu 610106,Sichuan,China;Sino-French Institute of Nuclear Engineering and Technology,Sun Yat-Sen University,Zhuhai 519082,Guangdong,China;State Grid Sichuan Integrated Energy Service Co.,Ltd.,Chengdu 610031,Sichuan,China)

机构地区：[1]四川大学破坏力学与工程防灾减灾四川省重点实验室,四川成都610065 [2]成都大学建筑与土木工程学院,四川成都610106 [3]成都大学机械工程学院,四川成都610106 [4]中山大学中法核工程与技术学院,广东珠海519082 [5]国网四川综合能源服务有限公司,四川成都610031

出　　处：《电网与清洁能源》2025年第1期1-9,16,共10页Power System and Clean Energy

基　　金：中国博士后科学基金项目(2021M702371)。

摘　　要：为了对覆冰输电导线气动力系数进行精确预测,基于风洞试验获取了覆冰导线气动力系数,利用SVR(support vector regression,SVR)、BP(back propagation,BP)和RBF(radial basis function,RBF)3种神经网络方法开展了机器学习预测。比较了SVR、BP和RBF神经网络在覆冰输电导线气动力系数预测中的效果。通过对数据集的训练和测试,结果表明:SVR、BP和RBF神经网络均能有效预测气动力系数的数值变化,但RBF神经网络在不同冰厚和风速下的整体预测效果优于SVR和BP神经网络。多组数据的预测实验进一步验证了RBF神经网络模型具有较强的适用性和更高的准确性。RBF神经网络模型可以作为覆冰导线气动力系数预测的有力工具,能有效进行气动力系数预测。To accurately predict the aerodynamic coefficients of iced transmission lines,based on the aerodynamic coefficients of the iced conductor obtained in wind tunnel test,three neural network methods,namely SVR,BP and RBF,are used to carry out machine learning prediction.In this paper,the effects of SVR,BP and RBF neural networks in the prediction of aerodynamic coefficients of iced transmission lines are compared.Through training and testing of the datasets,the results show that all three methods-SVR,BP and RBF neural networks can effectively predict the numerical changes of aerodynamic coefficients,but the overall prediction effect of RBF neural networks under different ice thicknesses and wind speeds is better than that of SVR and BP neural network.The prediction experiment of multiple sets of data further verifies the strong applicability and higher accuracy of the RBF neural network model.The RBF neural network model can be used as a powerful tool for predicting the aerodynamic coefficients of iced conductors,and can effectively predict the aerodynamic coefficients.

关 键 词：SVR BP神经网络 RBF神经网络 气动力系数 数据预测 

分 类 号：TM752[电气工程—电力系统及自动化]