论文部分内容阅读
为了有效提取火焰燃烧特征量,进而对火焰燃烧稳定性进行分析,文章通过理论分析和实验总结,首先提取了火焰强度方差及闪烁频率作为特征量,之后重点运用高阶统计量对火焰信号进行分析,提取了对火焰稳定性有较大影响的相位作为另一个新的特征量。最后考虑到各个特征量之间的相关性及量纲的影响,运用马氏距离作为判别准则,对火焰燃烧稳定性进行了有效的识别。结果表明,提取的3个特征量对火焰燃烧稳定性都有重要影响,而且运用马氏距离能够综合这3个特征量,使得从总体样本中选出的训练样本识别率达到93%,识别效果理想。
In order to extract the characteristics of flame combustion effectively and analyze the combustion stability of the flame, the paper first extracts the variance of flame intensity and the flicker frequency as the characteristic quantities by theoretical analysis and experimental conclusion, and then analyzes the flame signal by using high-order statistics , Extracted a phase that has a greater influence on flame stability as another new feature quantity. Finally, taking into account the correlation between the features and the influence of the dimension, Mahalanobis distance is used as the criterion to effectively identify the flame combustion stability. The results show that the three extracted features have an important influence on the flame combustion stability, and the use of Mahalanobis distance can be integrated these three features, making the training samples selected from the overall sample identification rate of 93% recognition effect ideal.