|本期目录/Table of Contents|

[1]陈烈,张永明,齐维贵,等.供热过程时滞递推神经网络解耦器设计[J].电机与控制学报,2009,(增刊1):78-82.
 CHEN Lie,ZHANG Yong-ming,QI Wei-gui,et al.Design of recurrent neural network decoupler with delays for heat supply process[J].,2009,(增刊1):78-82.
点击复制

供热过程时滞递推神经网络解耦器设计(PDF)
分享到:

《电机与控制学报》[ISSN:1007-449X/CN:23-1 408/TM]

卷:
期数:
2009年增刊1
页码:
78-82
栏目:
出版日期:
2010-02-02

文章信息/Info

Title:
Design of recurrent neural network decoupler with delays for heat supply process
作者:
陈烈; 张永明; 齐维贵; 邓盛川; 于德亮
哈尔滨工业大学电气工程及自动化学院
Author(s):
CHEN Lie; ZHANG Yong-ming; QI Wei-gui; DENG Sheng-chuan; YU De-liang
Department of Electrical Engineering and Automation; Harbin Institute of Technology; Harbin 150001; China
关键词:
供热过程 节能控制 神经网络解耦 递推神经网络 嵌入维数预估
Keywords:
heat supply process energy-saving control neural network decoupling recurrent neural network embedding dimension estimation
分类号:
TP183
DOI:
-
文献标志码:
A
摘要:
针对供热过程质调—量调通道耦合特性和节能控制的需要,提出基于时滞递推神经网络的供热解耦方法。通过典型信号响应与最小二乘法结合的方法得到供热过程耦合系统模型,利用改进的假近邻法预估神经网络训练数据的嵌入维数,确定神经网络输入节点个数,引入时滞环节构建神经网络解耦器。采用时滞递推神经网络解耦器对供热耦合系统进行解耦,消除供热过程质调、量调通道间的非线性强耦合作用。仿真结果证明该方法具有良好的动态和静态解耦特性,能够满足供热过程多回路控制的要求,使供热系统能够跟踪节能设定值进行调节,实现供热节能和优质供热。
Abstract:
According to the coupling characteristics of heat supply process and the demands of energy-saving control,a novel decoupling method based on recurrent neural network with delays was proposed.Heat coupling model was founded with typical signal response and least-squares method.After the number of neural network inputs was determined by improved false neighbor method,time-delay blocks were introduced to build the neural network decoupler.With this decoulper,strong influence between quality-adjust and quantity...

参考文献/References:

-

相似文献/References:

备注/Memo

备注/Memo:
国家“十一五”科技支撑计划重大项目(2006BAJ03A05);; 哈尔滨市科技创新人才研究专项资金项目(RC2006XK007001)
更新日期/Last Update: 2010-06-07