基于向量量化网络的火电厂炉膛火焰检测

Furnace Flame Detection Based on LVQ in Thermal Power Plant

通过对火力发电厂锅炉炉膛火焰检测装置及炉膛燃烧器的分布位置进行分析,结合神经网络的LVQ结构及MATLAB编程,介绍一种火焰检测的新方法。该方法首先通过火焰检测装置把采集到火焰有无的光信号转换成电信号,然后再经过滤波放大等环节把电信号转换成数字信号,送入微机用MATLAB编写的LVQ程序进行检测,根据仿真的数据及曲线可以看出此种方法能够及时准确地为运行人员提供报警信息,达到预先设定的目标。

By the analysis of the furnace chamber flame detective devices and the distributive location of the chamber burning devices, a new method of flame detection was introduced. This method was based on the combination of the LVQ structure of neural network and programming in MATLAB. First the light signals gathered by flame detective devices were transformed into the electrical signals, then these electrical signals were transformed into the digital signals through the filter amplifier, finally digital signals were sent to the microcomputer and were dealed with by LVQ program in MATLAB language. According to the results, this method could pro- vide alarm informations accurately and promptly for the operators, and the setting goal could be reached.