SOM神经网络辅助EIS分析有机涂层防护性能

Study on Protective Performance of Organic Coatings Based on EIS Assisted with SOM Neural Network

利用电化学阻抗谱(Electrochemical Impedance Spectroscopy,EIS)技术对某型车辆灰色有机涂层全浸泡条件下的失效过程进行分析,提取了特征参数——高频相位角变化率k(f),并以此对涂层性能进行初步评价。以k(f)为自组织神经网络(Self-Organizing Maps,SOM)的输入训练样本,对涂层防护性能变化进行了辅助分析。k(f)变化规律与SOM神经网络分类结果均表明,涂层失效过程可分为三个阶段,验证了SOM神经网络辅助分析车辆有机涂层浸泡性能的有效性。

Under immersion state,the corrosion behavior of gray organic coating has been studied using electrochemical impedance spectroscopy（ EIS）. This paper has selected the value rate of high-frequency phase angle k（ f）that was the characteristic parameter from the EIS plot to evaluate roughly the protective performance of the coating. Assisted by self-organizing feature map（ SOM） network,k（ f） has been trained as its input sample to acquire the whole process of coating failure that can be divided into three stages. The SOM neural network classification results were consistent with those obtained by the variation law of k（ f）,which verified the effectiveness of the SOM neural network in the analysis of organic coating immersion performance.