材料期刊网

采用面向对象的编程语言Visual Basic(VB)编制了金属材料海洋环境腐蚀数据咨询管理和预测诊断系统，系统中包含常用金属材料实海腐蚀数据和腐蚀形貌图谱的图文数据库。除常用的数据库管理功能外，还具有一定的腐蚀预测和腐蚀形貌诊断功能。腐蚀预测可分为人工神经网络模型预测和灰色模型预测，人工神经网络模型根据材料的合金成分或海水环境因素对材料的腐蚀进行预测，而灰色模型用于对材料的长期腐蚀数据进行计算并给出灰色模型参数。由扫描得到的金属材料的腐蚀形貌图像的灰度值分布或分形特征值和对应的腐蚀形貌作为知识库，通过模糊模式识别理论建立了腐蚀形貌分析诊断系统，可以对材料的腐蚀形貌进行诊断。

A consultation and forecast diagnosis system of metal exposed in marine environment has been developed by using the object-oriented programming language, Visual Basic (VB). The system comprises three major modules: database and management system, corrosion prediction and corrosion morphology analysis. In the first part, the common data management function such as data input, data modify, data delete and data refresh are discussed. There are two different method used in the corrosion prediction module, the artificial neural network (ANN) and gray model. Using the ANN method, the corrosion depth of new materials in different marine environment can be predicted. The gray model method can give the corrosion rate of materials in the future. The corrosion morphology images of metallic material in seawater, which are acquired by scanner, are stored in the database. Taking the gray distribution or fractal characters of metallic samples and their corrosion morphology as the knowledge base respectively, the diagnosing system identifying corrosion morphology of metallic material in seawater was established according to the theory of fuzzy pattern recognition. The corrosion morphology of sample can be identified by gray distribution or fractal characters of corrosion images.