摘要
由于斜拉索索力的准确计算与测量对桥梁建设与养护具有重要的工程意义,提出一种自适应模糊神经网络索力估计与拟合分析方法。基于弦振动理论与切比雪夫级数方法推导弹性边界条件下拉索振动频率方程,建立索力与频率间的函数关系,利用自适应模糊神经网络对索力-基频关系进行拟合修正。通过搭建索力实验平台来验证该方法的有效性。实验结果表明:设计的推理系统使得训练平均误差小于0.15 N,测试平均误差小于0.30 N,其相对误差分别为0.00347%和0.00727%。表明基于实验数据的自适应神经网络系统可以保证基频法索力评估的准确度,为工程应用索力计算提供有效的解决途径。
The accurate calculation and measurement of cable tension has important engineering significance for bridge construction and maintenance.An adaptive fuzzy neural network cable force estimation and fitting analysis method is proposed.Based on the string vibration theory and the Chebyshev series method,the vibration frequency equation of the elastic boundary condition pull-down cable was derived,and the relationship between the cable force and the frequency was established.The adaptive fuzzy neural network was used to fit the cable-foundation frequency relationship.Finally,the effectiveness of the method was verified by building a cable force test platform.The experimental results show that the designed inference system makes the training average error less than 0.15 N,the test average error is less than 0.30 N,and the relative errors are 0.00347%and 0.00727%,respectively.It is shown that the adaptive neural network system based on experimental data can ensure the accuracy of the fundamental frequency method and provide an effective solution for engineering cable force calculation.
作者
王振
闫伟
张刚
陈跃华
Wang Zhen;Yan Wei;Zhang Gang;Chen Yuehua(Faculty of Maritime and Transportation,Ningbo University,Ningbo 315211,Zhejiang,China)
出处
《计算机应用与软件》
北大核心
2021年第10期53-60,共8页
Computer Applications and Software
基金
国家自然科学基金项目(51675286)
国家自然科学基金青年项目(51505237)。
关键词
索力计算
弹性边界
基频法
切比雪夫级数
自适应神经网络
Cable force calculation
Elastic boundary
Fundamental frequency method
Chebyshev series
Adaptive neural network
