Matlab神经网络工具箱编程和Delphi对其调用

Programming Design of Matlab's Neural Network toolbox and OLE Automation to It with Delphi

针对电弧炉炼钢氧化期是一个复杂的物理化学反应过程，难以建立机理模型，利用Ｍａｔｌａｂ神经网工具箱建立ＢＰ神经网预测模型，并详细讨论该模型的编程设计问题。利用ＯＬＥ自动化，用Ｄｅｌｐｈｉ编写人机交互界面来指导冶炼生产。

In this paper, problems associated with the production of steels during the oxidizing stage in the electric arc furnace are discussed. The design problems (use Matlabs neural network toolbox) to the developments of a backpropagation neural networks predictive model of the oxidizing process is considered and a backpropagation neural network predictive model has been developed, which provides intelligent control capable of learning the characteristics of the oxidizing as it operates, and program interface (use Delphis OLE automation) is designed.