摘要
基于小波变换的时频局域化特性和人工神经网络的非线性映射特性,将两者优点结合发展了一种适用于海洋平台结构损伤识别与定位的新方法。以一典型海洋平台模型数值模拟结构损伤识别效果,结果表明将小波变换和神经网络方法联合应用于海洋平台结构损伤识别与定位是切实可行的。
Combing the advantages of localization character both in time-domain and frequency-domain of wavelet transform and the nonlinear mapping property of neural network, a new approach is developed for damage identification and localization of offshore platforms. A typical offshore platform model is selected as a numerical example to simulate the effect of damage identification. The results show that the method combing wavelet analysis and neural networks is feasible to detect and localize the damages of offshore platforms.
出处
《海洋技术》
北大核心
2009年第4期113-116,共4页
Ocean Technology
基金
国家自然科学基金资助项目(50609009)
江苏省船舶先进设计制造技术重点实验室项目(GJ0807)
关键词
海洋平台
损伤识别
小波分析
神经网络
offshore platform
damage identification
wavelet analysis
neural network
