【作者】 盛杨；

【导师】 刘禾；

【作者基本信息】 华北电力大学（北京） ， 模式识别与智能系统， 2007， 硕士

【摘要】 当今电站锅炉炉膛灭火判别对锅炉安全和经济运行有着重要意义。本文针对炉膛火焰图像监测系统存在的灭火判别问题,提出了一种炉膛火焰灭火判别方法。该方法通过对火焰图像的分析,提取了用于灭火判别的两个特征值,然后对这个特征空间使用神经网络和支持向量机进行识别分类,结果表明:特征量提取是成功的,这种判别方法有很高的正确率,能够正确对火焰图像进行灭火判别;支持向量机方法用于灭火判别是可行的。在此基础上讨论了根据序列图像如何判别单煤粉燃烧器有灭火情况发生,以及利用支持向量机算法如何判别火焰燃烧稳定性的问题。更多还原

【Abstract】 To solve the problem of judging outfire in flame image monitor system, the way of judging furnace outfire was presented. First, the two features of the flame image were extracted. Then the feature space consisting of two features were classified with the neural network and Support Vector Machine. Results show that the feature extraction is correct. This method has high accuracy, and can correctly judge the state of the combustion. Support Vector Machine is effective for outfire judgment. At last, this article discusses how to judge outfire situation of one coal burner by flame image sequence and how to judge the combustion stability by SVM.更多还原