摘要
主要针对熔窑温度表现出的非线性、慢时变、大迟延和不确定性等特点,根据径向基函数神经网络具有可以逼近任意非线性映射、较快的学习速度并避免局部极小问题的能力,将RBF网络用于熔窑温度过程非线性模型的在线辨识,仿真结果表明能较好地跟踪温度的实际输出数据,具有较高的学习精度.
This paper offers corresponding plans to control the temperature of melting furnace characterized by nonlinear, slow time - varying, strong delaying and uncertainties, etc. according to the abilities which neural networks of Radial Basis Function can theoretically approach any non - linear relation, high speed of studying or avoid part minimum problem. The RBF neural network is used in nonlinear model identification of melting furnace temperature process and a satisfactory control effect has been obtained. This result indicates that RBF neural network can trace actual output of temperature accurately.
出处
《佳木斯大学学报(自然科学版)》
CAS
2007年第4期440-442,共3页
Journal of Jiamusi University:Natural Science Edition
关键词
RBF
熔窑
温度辨识
RBF
melting furnace
temperature identification