论文部分内容阅读
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。
The traditional research methods of phosphating exist many defects, such as long test period, fee, pollution and harmfulness, which make it difficult to produce excellent phosphating solution. Therefore, for a particular application, to quickly generate a reasonable formula for phosphating solution has been the focus of current research. Based on the formula of zinc-manganese-nickel phosphating solution, the main factors affecting the quality of phosphating film were analyzed to determine the composition of phosphating solution. The neural network was used to establish the mathematical model of simulated phosphate film’s weight and corrosion resistance. And MATLAB integrated technology for the preparation of the entire network training and forecasting process, developed a phosphide liquid formula experimental process of intelligent simulation system. The system can be a user of any combination of a set of experimental data, calculate the corresponding weight and corrosion resistance results. According to the calculation results, choose the best combination of programs to arrange experiments to verify. The intelligent simulation method makes up for the technical defects of the traditional experimental research methods, which not only solves the huge experimental problems, but also realizes the optimal data selection and saves the experimental resources.