摘要
提出了一种基于模态应变能识别复合材料结构损伤的遗传-神经网络法。采用单元模态应变能改变率作为结构损伤标识量,对损伤结构进行损伤识别仿真。采用混合编码遗传算法对BP神经网络进行拓扑优化,构成遗传-神经网络,并使用其识别复合材料机翼结构损伤的位置和程度.算例表明,本方法具有高效、稳定的特点,有着很好的推广应用价值。
A GA-ANN methodology of composite structures damage detection based on modal strain energy was presented in this paper. The change ratio of the element modal strain energy, which acted as the indicator for structure damage detection, can successful find the location and severity of the damage structure. A mixed coding genetic algorithm was used to optimal the topology of ANN, which combined as the GA-ANN, was used as identification of the composite wing structure. The example shows that the methodology presented in this paper has high efficient and good robust, also has a wide area of application.
出处
《强度与环境》
2008年第3期14-19,共6页
Structure & Environment Engineering
关键词
结构损伤
模态应变能
神经网络
遗传算法
structure damage
modal strain energy
neural network
genetic algorithm
