基于SIS的典型工况下火焰特征统计分析

电厂的煤粉火焰检测装置只作炉膛有火或无火判断,不仅不能对燃烧稳定程度做出定量表述。以某电厂的监控信息系统(SIS)中的历史数据库作为分析平台,对常规光学火检信号所包含的丰富信息进行深层次挖掘。首先对火检信号数据进行必要的预处理,提取出火焰信号的亮度平均值、火焰亮度方差、火焰亮度峰峰值作为火检信号时域特征量,在前人研究的基础上提出并使用“均匀度”作为火检信号频域特征量。然后分锅炉高负荷稳定、锅炉低负荷稳定、锅炉高负荷波动、锅炉低负荷波动、启磨、停磨、点火与停炉等8种典型燃烧工况对火检强度信号作了大量统计分析。结果表明这些特征量能够较准确地反映不同工况下的火焰燃烧状态,并为今后电站锅炉燃烧优化奠定了基础。

Statistic analysis on features of flame in typical-condition based on SIS

Coal flame detectors of power plant are only used to recognize On-fire or Off-fire, which make it impossible to judge quantitatively the stability of combustion.This paper analyzes the flame signals on-site to mine more information with the help of Supervisory Information System(SIS). Firstly, flame signals are preprocessed.Four features, including mean, variance, peak-to-peak value of time-domain flame signal and uniformity of frequency-domain are extracted.Then we did a lots of data analyses of eight typical combustion conditions, i.e.stable high load condition, stable low load condition, high load fluctuation condition, low load fluctuation condition, starting mill condition, stopping mill condition, igniting condition, stopping boiler condition, and the result shows that these four features are evident to reflect the combustion state under different conditions.The conclusion has laid foundation for the next work of combustion optimization.