基于免疫原理的RBF网络在线学习算法及其在热工过程大范围工况建模中的应用.pdfVIP

26 卷 9 期 中 国 电 机 工 程 学 报 Vol.26 No.9 May 2006 2006 年5 月 Proceedings of the CSEE 2006 Chin.Soc.for Elec.Eng. 文章编号：0258-8013 （2006 ）09-0014-06 中图分类号：TK 123 文献标识码：A 学科分类号：47020 基于免疫原理的径向基函数网络在线学习算法及 其在热工过程大范围工况建模中的应用 林金星, 沈 炯, 李益国 （东南大学动力系, 江苏省 南京市 210096 ） A Novel Online Training Algorithm Based on Immune Principles for RBF Neural Network and Its Application in the Thermal Process Wide Range Modeling LIN Jin-xing, SHEN Jiong, LI Yi-guo （Department of Power Engineering, Southeast University, Nanji ng 210096, Jiangsu Province, China ） ABSTRACT: A novel online training algorithm for RBF 记忆新样本的功能, 并将该网络应用于某300MW 火电机组 neural network based on immune principles is presented to 主汽压的多工况辨识.实验结果表明该算法不仅能精简网络 overcome the weaknesses of the traditional neural network 的结构, 而且能很好地适应对象的时变特性. training algorithm in the long time-period identification of 关键词：热工过程；径向基函数神经网络；在线学习；免疫 non-linear dynamical system. With the analysis of the 原理；系统辨识 comparability between the learning of RBF neural network and immune system, this algorithm uses immune memory, clone 0 引言 selection and cell languish mechanisms to adj ust the nodes in 近来, 径向基函数网络(RBF 网) 由于诸多优点 hidden layer dynamically, and determine their core function’s center and width. As a result, the RBF neural network could （如：非线性逼近能力强, 网络结构简单, 学习速度