基于SOM技术的有机涂层电化学阻抗谱的自动识别分析  

Automatic Recognition and Analysis of Electrochemical Impedance Spectra of Organic Coatings Based on SOM Technology

作　　者：丁锐[1,2] 桂泰江[2] 蒋健明[2] 余海斌[1] DING Rui GUI Tai-jiang JIANG Jian-ming YU Hai-bin(Ningbo Institute of Materials Technology, and Engineering, Chinese Academy of Sciences, Ningbo 315201, Zhejiang, China Marine Chemical Research Institute Co.,Ltd. State Key Laboratory of Marine Coatings, Qingdao 266071, Shandong, China)

机构地区：[1]中国科学院宁波材料技术与工程研究所,浙江宁波315201 [2]海洋化工研究院有限公司,海洋涂料国家重点实验室,山东青岛266071

出　　处：《中国涂料》2016年第11期53-61,65,共10页China Coatings

摘　　要：本实验建立了用于识别分析有机涂层电化学阻抗谱的SOM人工神经网络。应用该网络对几种常见的有机涂层电化学阻抗谱进行了识别分析,实验表明3种数据组合θ和|Z|、θ和d|Z|/dlogf、dθ/dlogf和|Z|既包含了阻抗谱形状的变化,又可以反映研究对象阻抗大小的变化,可以获得较好的识别效果,其中θ和d|Z|/dlogf的数据组合效果最好。该网络技术可以为涂料研究机构提供比较便捷的阻抗谱分析手段,用于考察涂层状态的变化。In this experiment, the SOM artificial neural network was founded to recognize and analyze the electrochemical impedance spectra of organic coatings. After the recognition and analysis of the electrochemical impedance spectra of some common organic coatings, it indicated that the data combinations of θ ｜Z｜, θ d｜Z｜/dlogf and dθ/dlogf ｜Z｜ not only included the changes in shapes of impedance spectra but also reflected the changes in impedance size of the researched subject. Three combinations obtained better recognition result, and the combination of θ （d｜Z｜）/dlogf was the best. This network technology can provide coatings research institutions with convenient analysis methods of electrochemical impedance spectra. What＇s more, it also can be used to investigate the state changes of coatings.

关 键 词：SOM神经网络 有机涂层 电化学阻抗谱 

分 类 号：TQ630.72[化学工程—精细化工] TG376[金属学及工艺—金属压力加工]