摘要
传统的磷化处理研究方法存在着实验周期过长、费料、污染、有害等缺陷,导致优良的磷化液配方难以快速的生成。因此,针对某一项应用,快速生成合理的磷化液配方一直是目前的研究焦点。本文以锌锰镍系磷化液配方为基础,分析影响磷化膜质量的主要因素,确定磷化液配方成分;用神经网络建立模拟磷化膜膜重和耐腐蚀性的数学模型;用VB和MATLAB集成技术编制了整个网络的训练及预测程序,研发了磷化液配方实验过程的智能模拟系统。该系统可以对用户任意组合的一组实验数据,计算出相应的膜重及耐腐蚀性结果。根据计算结果,选择最优的组合方案安排实验进行验证。该智能模拟方法弥补了传统实验研究方法的技术缺陷,既解决了实验量巨大的问题,又实现了最优数据的选取,节约了实验资源。
Traditional phosphate research methods have unfavorable factors,such as time-consuming,material costs,pollution,poor safety and so on,which leading to excellent phosphate solution formulations difficult to generate rapidly.Therefore,for a particular application, generate a reasonable formula has been the focus of current research.In the research,based on Nickel-Zinc phosphate solution formula,analysis the main factors of influence phosphate film,determine the phosphate solution formulation ingredients.By neural network modeling,intelligent simulation system on membrane weight and corrosion is constructed.Programming the network training and predicting system by VB MATLAB.The formulations experiment intelligent simulation system has been developed.For any combination of a set of experimental data,the system can calculate the membrane weight and corrosion resistance results.Based on the calculating results,users can choose an optimizing combination data to arrange experiment for verification.The intelligent simulation methods make up traditional experiment defects,not only resolve the huge experiments problems,but also achieve the optimal data selection, at the same time saving experiment＇s resource.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2010年第7期941-944,共4页
Computers and Applied Chemistry
基金
河北省科技厅科技攻关项目(07212173)
关键词
膜重
耐腐蚀性
智能模拟
神经网络
membrane weight
corrosion resistance
intelligent simulation
neural network
