基于径向基函数温度控制系统的模型辨识  

Model Identification of the Temperature Control System Based on the Radial Basis Function (RBF)

作　　者：张红朋[1] 张一[1] 李涛[1] 杨静[1] 

机构地区：[1]辽宁石油化工大学信息与控制工程学院,辽宁抚顺113001

出　　处：《石油化工自动化》2009年第1期40-41,54,共3页Automation in Petro-chemical Industry

摘　　要：针对一些带复杂参数辨识性能指标的模糊建模无法采用经典辨识方法解决的情况,提出了采用径向基函数(RBF)神经网络模型对被控对象进行参数辨识,该方法具有无需具体数学模型的特点,解决了模糊模型辨识中辨识指标表达式过于复杂甚至非解析所带来的难题,从而实现了系统的参数辨识。仿真研究表明,利用该辨识方法得到的模糊模型能够很好地逼近真实系统,并且对初始条件和输入信号不敏感,具有鲁棒性。The radial basis function （RBF） neural network model is proposed to carry out parameters identification for controlled objects because some blurred modeling with complicated parameters identification index can not be solved by classical identification methods. The method has the characteristic do not need actual mathematic model, solves the difficulty that the identification index formula is too complicated and even in the blurred model identification, therefore realized the parameter identification. The simulation experiments demonstrate that the blurred model by the identification method can be close real system very well and is not sensitive to original conditions and input signal, has robustness.

关 键 词：神经网络 径向基函数 模型辨识 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]