论文部分内容阅读
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 further study the failure process of aircraft structural coating protection, the accelerated test results of shot peen + primer protective coating were analyzed using Kohonen neural network and electrochemical impedance spectroscopy (EIS). Accelerated tests simulate thermal shock and salt spray environments, with specimen data measured during the test. The impedance change rate of the specimen per cycle is taken as the input data of the neural network. The neural network is trained with the test data of three sets of test pieces and tested with another set of test piece data. The Kohonen neural network divides the coating failure process into five sub-processes, which gives more corrosion information than the traditional three-process classification. The classification results of neural network valid data are in good agreement with the low-frequency impedance spectra and the corrosion states, which shows that Kohonen neural network can effectively predict the coating failure process.