两种数学模型在评判燃煤结渣特性中的改进及应用  

Improvement and Application of Two Mathematics Models in Evaluating the State of Deposition and Slagging of Coal Burning Boiler

作　　者：文孝强[1] 刘彦臣[1] 孙媛媛[2] 孙灵芳[3] 

机构地区：[1]东北电力大学电力运行仿真中心,吉林吉林132012 [2]东北电力大学保卫处,吉林吉林132012 [3]东北电力大学自动化工程学院,吉林吉林132012

出　　处：《东北电力大学学报》2007年第6期71-76,共6页Journal of Northeast Electric Power University

摘　　要：针对常规的模式识别模型和BP网络模型在评判燃煤结渣特性的过程中准确率不高的问题,将两个动态指标引入模型中建立了新的评判模型,通过对几个具体电厂进行评判,结果表明所提出的模型具有较高的评判准确率,从而也证明所提出的模型是合理可行的。To solve the problem of low accuracy of normal pattern-recognition and BP neural-network model in predicting and determining the state of deposition and slagging of coal burning boiler, two dynamic norms are applied in the two new models. Through experiments, it proves that the two new models are higher in accuracy than the traditional ones accordingly and also proves that the models are reasonable and practicable.

关 键 词：模式识别 神经网络 评判 结渣 

分 类 号：TK227[动力工程及工程热物理—动力机械及工程]