超超临界二次再热机组再热汽温系统模型辨识

摘要

Due to the complexity and particularity of double-reheat unit's boiler construction,the dynamic characteristics of reheat steam temperature changes greatly compared with the unit with single reheat cycle,and the stable operation is more difficult.Therefore,it is of great significance to establish the model of reheat steam temperature with good accuracy for the stable operation of the unit and the optimization of the control strategy.Based on the historical data of Guodian Taizhou 3#1000MW ultra-supercritical double reheat unit,the maximum information coefficient was adopted to analyse the correlation between the parameters.Subspace principal component recursive algorithm was used to analyse the main factors influencing reheat steam temperature system at different time.And combined with the subspace identification algorithm,the multi variable model of rolling reheat steam temperature system were obtained.The research shows that the rolling multivariable model identification method can reflect the thermal dynamic characteristics of reheat steam temperature system of ultra-supercritical double reheat unit,and these models can be used to study the control optimization of reheat steam temperature system.%由于超超临界二次再热机组结构的复杂,使得再热汽温系统动态特性与一次再热机组差异较大,这也使得机组稳定运行的难度也随之增加,因此建立精度良好的再热汽温模型对优化控制策略的制定具有重大的意义.基于国电泰州3#1000MW超超临界二次再热机组运行历史数据,采用最大信息系数对参数进行相关性分析,用子空间跟踪主元递推算法分析不同时刻影响再热汽温系统的主要因素,并结合子空间辨识算法,得到了再热汽温系统的滚动多变量模型.仿真结果表明,滚动递推辨识方法得到的滚动多变量模型能够反映超超临界二次再热机组再热汽温的热工动态特性,可用于再热汽温系统的优化控制的研究.