基于Kohonen神经网络的飞机结构涂层失效分析  

Failure Analysis of Protection System of Aircraft Structures Using Kohonen Artificial Network

作　　者：徐元铭[1] 冉峻塽 XU Yuan-ming RAN Jun-shuang(School of Aeronautic Science and Engineering, Beijing University of Aeronautics and Astronautics, Beijing 100191, China)

机构地区：[1]北京航空航天大学航空科学与工程学院,北京100191

出　　处：《飞机设计》2017年第4期37-41,共5页Aircraft Design

摘　　要：为了深入研究飞机结构涂层防护的失效过程,使用Kohonen神经网络结合电化学阻抗谱技术(EIS),对喷丸+喷底漆防护涂层的加速试验结果进行了分析。加速试验模拟了热冲击和盐雾环境,试件数据都是在试验过程中测得。将试件每周期的阻抗变化率作为神经网络的输入数据,用3组试件的试验数据对神经网络进行训练,用另一组试件数据进行测试。Kohonen神经网络将涂层失效过程分成了5个子过程,相比传统的3个过程分类可以得到更多的腐蚀信息。神经网络有效数据的分类结果与低频阻抗谱和腐蚀状态十分吻合,说明了Kohonen神经网络可以有效地预测涂层失效过程。In order to study the failure process of protection system of aircraft structures,accelerated life test results on shot peening ＋ spray primer protective coating system have been analyzed by using Kohonen artificial network and EIS method. Accelerated experiments considered general atmospheric environmental factors such as thermal shock and salt fog. All the sample data were collected during exposure to accelerated corrosion environments. And it took the changing rate of impedance of each cycle as an input data. Neural network has been trained using three sets of samples and tested using another set of sample data. Compared with traditional classification,Kohonen artificial network method classifies corrosion process into five sub-processes which is refinement of three typical corrosion processes. The classification results of Kohonen artificial network are highly consistent with the predictions based on impedance magnitude at lowfrequency and corrosion stages,which illustrates that the Kohonen artificial network is an effective method to predict the failure process of coatings.

关 键 词：涂层 腐蚀 EIS KOHONEN神经网络 分类 

分 类 号：TF111.52[冶金工程—冶金物理化学]