摘要
本文提出一种粗糙集理论和动态前馈神经网络相结合的神经网络构造方法。充分发挥了粗糙集理论和神经网络的优势，弥补了各自的缺点。并应用于实际工业过程，在乙烯装置裂解炉燃料气热值控制中取得了良好的应用效果。
In this paper, the rough set theory and dynamic feed-forward neural network hybrid system was introduced. The advantage of rough set and neural network has been fully bonded and the disadvantage was avoided. Such soft sensor technique is applied to ethylene device cracking furnace, and the result of bunting-heat control of fuel gas demonstrated the soft sensor and intelligent control system are feasible and effective.
出处
《通讯和计算机（中英文版）》
2005年第2期7-13,57,共8页
Journal of Communication and Computer
