【摘要】 电厂过热汽温对象受诸多因素相互作用,传统建模方法难以处理此类复杂系统问题。将基于统计学理论的新方法支持向量机应用于过热汽温建模,避开复杂的机理建模,使用历史数据对汽温与各操作参数间的关系进行描述,建立回归模型。以某电厂200 MW机组锅炉为例,采集各相关变量的现场运行数据,使用支持向量机建立过热汽温回归模型。实际结果与实际对象有较高的相关度。该模型对汽温调节、参数优化及机组运行具有指导意义和参考价值。更多还原

【Abstract】 The superheated steam temperature in a power plant is affected by many factors,leading to the difficulty to deal with such complex system problems with conventional modeling methods.The support vector machine algorithm based on statistical theory was used to build the model of the superheated steam temperature.It avoided the complex mechanism modeling,and used the historical data to describe the relationship between the superheated steam temperature and various operation parameters to build the regression model.Taking a 200 MW power plant unit for example,the field operation data of relevant variables was collected,and the regression model of the superheated steam temperature was built based on support vector machine.The results had high correlation with the actual object.The model could be used as guidance and reference for the steam temperature regulation,parameter optimization and actual operation.更多还原