基于神经网络算法的海上风机结构状态监测研究

针对目前海上风机结构状态监测研究很少涉及包括基础在内的整个结构体系,而现有模态分析方法某些情况下精度较差,且未经过工程验证的现状,本文采用静力分析与风机结构模态分析相结合的方法,分析并提取了各种环境荷载对风机结构响应显著的特征要数,建立了BP神经网络,提出了传感器的布置原则,并利用克里格插值得到了结构的分布云图。通过对比,神经网络与有限元计算误差在5%左右,具有较高的精度,且计算速度快,可为海上风机结构状态在线监测提供参考。

Research of offshore wind turbine status monitoring based on neural network algorithm

Nowadays, research on status monitoring of offshore wind turbine is a hot topic, and most of it focuses on the blade and tower,rather than the whole structure including foundation. Modal analysis is the most common method, which is invalid in some cases for poor precision,and not-applied in any projec-tion. In this paper,the static analysis and modal analysis will be used in the study of structure state moni-toring,based on neural network algorithm. This paper presents the installation principle of sensors,and ex-tracts the characteristics of various environment loads which have remarkable effect on the whole structural. Compared with the finite element calculation, the results show that error between neural network and finite element calculation is around 5%, and can provide reference for the on-line monitoring of the structure state of offshore wind turbine.