基于多尺度相关和机理建模的炉膛压力分析

Analysis of Furnace Pressure Based on Multi-scale Correlation and Mechanism Modeling

借鉴小波分析思想,设计了简单滤波器组,将信号分解为不同频率尺度的分量,对不同信号的各个分量之间进行相关分析,比较不同频率尺度下频率分量间的相关性特征,从而大致判断出信号间的影响关系及类型.利用多尺度频率相关分析方法对电站锅炉炉膛压力等信号进行分析,发现炉膛压力低、中频段分量分别与燃料量、送风量的中、高频段分量呈现出明显的正相关性.建立了燃料量、送风量和引风量与炉膛温度和压力间的简化动态模型,并进行机理分析.结果表明:炉膛压力信号与锅炉燃烧率之间具有带通滤波特性,炉膛压力能够在较宽的负荷范围和频带内反映锅炉燃烧强度的变化.

Using wavelet analysis ideas for reference,simple filter banks were designed to decompose the signals into frequency components with different scales.Correlation analysis was made among components of different signals,and the correlation characteristics among frequency components with different scales were compared to roughly determine the influencing factors and types of signals.Boiler furnace pressures were analyzed using multi-scale correlation method,during which significant positive correlation was found between the low/medium-frequency components of furnace pressure and the medium/high-frequency components of fuel and air flow.A simplified dynamic model was then built up between the flow of fuel/forced-draft air/induced-draft air and the furnace temperature and pressure.Mechanism analysis results show that bandpass filter characteristics exist between the pressure signal and combustion efficiency;the furnace pressure can reflect the change of combustion intensity in a wide range of load and frequency.