摘要
针对汽轮机排汽焓的计算是火电机组热经济性在线分析的难点,提出了采用遗传算法(GA)对基于支持向量回归机(SVR)的预测模型参数进行优化,利用优化后的模型(GA-SVR)对汽轮机排汽焓进行预测研究。以某300 MW汽轮机组为例进行了排汽焓的在线计算,并与常规SVR模型和BP-ANN模型进行对比。结果表明,该方法能够较为准确地在线预测汽轮机排汽焓值,可为火电机组的在线性能监测提供有效的手段。
As the calculation of steam turbine exhaust enthalpy is a difficult point for online economic analysis of thermal power plant,an optimized GA-SVR model for prediction and study of steam turbine exhaust enthalpy is being proposed,which is optimized from SVR model using GA algorithm.Taking a 300MW steam turbine unit as an example,online calculation of its turbine exhaust enthalpy was performed,of which the results were compared with that of conventional SVR model and BP-ANN model.Results show that this method can accurately predict steam turbine exhaust enthalpy,and therefore may be used for online performance monitoring of thermal power plants.
出处
《发电设备》
2010年第6期425-429,共5页
Power Equipment
关键词
汽轮机
排汽焓
模型
参数优化
在线预测
steam turbine
exhaust enthalpy
model parameter optimization
online prediction