摘要
提出了一种基于燃烧火焰图象特征的回转窑神经网络控制系统 .系统主要由两部分组成 ,一部分是回转窑煅烧带火焰燃烧状态识别系统 ,包括火焰图象获取、预处理、分割、特征提取与识别 ;另一部分是基于高斯基函数神经网络的控制系统 .实际运行结果表明该系统的有效性和实用性 .
A neural network control system for rotary kiln based on features of combustion flame is proposed. The system consists of two main parts. One is status recognition system of calcine band flame, which is composed of image capturing, preprocessing, segmentation, feature extraction and pattern recognition. The other is neural network control system using Gaussian potential function network(GPFN). The practical operating results illustrate effectiveness and practicability of the proposed system.
出处
《自动化学报》
EI
CSCD
北大核心
2002年第4期591-595,共5页
Acta Automatica Sinica
基金
国家"86 3"高技术计划 ( 86 3- 511- 984 5- 0 0 2 )资助
关键词
燃烧火焰图象特征
神经网络
控制系统
水泥回转窑
Algorithms
Control systems
Feature extraction
Functions
Fuzzy sets
Image processing
Neural networks
