非线性优化掺烧模型及应用

为满足电站燃煤锅炉在安全稳定运行的前提下降低燃料成本,提出多种煤优化掺烧研究。采用线性回归方法建立掺烧煤的发热量、挥发分、硫分、水分和灰分的预测公式,应用径向基人工神经网络建立掺烧煤的软化温度预测模型,从而建立了新的非线性优化掺烧模型。珠海电厂掺烧试验表明,所建模型能指导电厂多煤种优化掺烧,取得显著的经济效益。

Non- linear Optimization Model for Admixing Combustion and Its Application

In order to cut down the coal consumption on the condition that the power station boiler can operate safely,optimization admixing combustion was put forward.Linear regress formula was used to predict such properties of blended coal as heat value,volatility content,sulfur content,water content and ash content.Radial basis function neural network was applied to describe the relationship of soft temperature of blended coal with that of single coal,and a new non-linear optimization admixing combustion model was set up.Admixing combustion experiments of Zhuhai power station showed that this model could guide optimization coal admixing combustion and gain remarkable economic profits.