非线性最优化动力配煤技术的研究  

NONLINEAR PROGRAMMING TECHNOLOGY OF COAL BLENDING FOR POWER PLANT

作　　者：汤龙华[1] 周俊虎[1] 曹道卿 殷春根[1] 骆仲泱 姚强[1] 曹欣玉[1] 岑可法[1] 

机构地区：[1]浙江大学

出　　处：《煤炭学报》1997年第5期455-459,共5页Journal of China Coal Society

基　　金："八五"国家重点科技攻关项目

摘　　要：针对动力配煤局限于线性规划的现状及其缺点，研究了用非线性规划实现动力优化配煤的新技术．实验研究及其理论分析表明，混煤与单煤特性指标之间的关系是非线性的．用神经网络方法、线性拟合等多种手段来定量地描述这种非线性规律，从而建立了动力配煤的非线性优化数学模型，并将模型应用于正在建设中的杭州配煤场，得到了与以往线性优化模型不同的重要结果，显示了用非线性优化技术实现动力配煤的重要意义．To counter the shortcoming of limiting to linear programing of steam coal blending a new nonlinear programming technology of coal blending is studied. The experimental research and theoretical analysis shows that the relationship between properties of blended coal and single coal is nonlinear. The nonlinear low is quantitatively described by multi means such as neural network and linear fitting, thus, nonlinear programming mathematical model of steam coal blending is set up and used in the Hangzhou Coal Blending Plant, the important results different from former linear programming model is obtained and shows the important significance of using nonlinear programming technology to conduct steam coal blending.

关 键 词：动力煤 配煤 非线性 最佳化 神经网络 

分 类 号：TQ520.62[化学工程—煤化学工程]