基于参数自适应SVR的火电厂热工参数软测量

影响火电厂经济运行的一个重要因素是许多重要技术参数和经济参数难以进行在线实时测量。提出了一种新型软测量模型,它基于支持向量回归(SVR)算法,利用训练数据性能信息获得模型训练参数,减少人为因素对模型精度的影响,提高了建模效率和模型精度。某火电厂历史实测数据仿真结果表明,文中提出的方法能有效实现热工过程参数的软测量,有较大实用价值。

Soft measure for power plant thermal parameter based on self adaptive SVR

There are some important factors impact the electric power plant to go to economic operation model,such as a lot of important technical parameters and economic parameters cannot be real-time measured.This paper presents a new soft-sensing model based on the Support Vector Regression(SVR).The new model gets the training parameter form training the data and it can reduce the effect of human factor to the model accuracy and advance the modeling efficiency and accuracy.The result of the simulation experiment with one electric power plant history true data shows that the method presented in this paper can achieve excellent soft-sensing for electric power plant thermal parameter and has great use value.