基于AR模型与卡尔曼滤波的火检算法研究

摘要

In the boiler fire flame detection, in order to find the flame burning characteristics, real-time flame spectrum is analyzed by using Kalman filtering and AR model, the results show that this method can effectively extract the flame of each frequency component to respond to fire information and prevent gaussian white noise interference to the flame. It can improve the frequency resolution and calculation precision, it has a good application prospect in the inspection of the flame.%在锅炉火焰检测中,为更好地提取火焰的燃烧特征,有效地判断锅炉燃烧状态,文中采用卡尔曼滤波和AR模型相结合的算法,对仿真信号和实时火焰进行频谱分析.结果表明,该方法能从复杂的背景信号中有效地获得火焰的各频率组分,提高了对火焰频率的分辨率,减小了背景高斯白噪声对燃烧状态分析的干扰,为快速分析燃烧火焰信息提供了一种高效的方法,在锅炉火焰检测中有很好的应用前景.