一种适应变工况的火电机组过热汽温控制方法  

A Control Method for Superheated Steam Temperature of Thermal Power Units Adapted to Off Design Condition

作　　者：王印松[1] 郭迈豪 翁疆 WANG Yin-song;GUO Mai-hao;WENG Jiang(Department of Automation,North China Electric Power University,Baoding Hebei 071003,China;Fujian Huadian Electric Power Engineering Co.,Ltd,Fuzhou Fujian 350002,China)

机构地区：[1]华北电力大学自动化系,河北保定071003 [2]福建华电电力工程有限公司,福建福州350002

出　　处：《计算机仿真》2022年第10期124-129,183,共7页Computer Simulation

基　　金：百万千瓦超超临界机组的精细状态监测、故障诊断与自愈调控关键技术研究(U1709211)。

摘　　要：变工况下超临界火电机组过热蒸汽温度控制系统的动态特性具有时变性、非线性、大惯性及时滞性,常规串级PID控制无法保证其控制品质。提出一种具有数据库更新算法的数据驱动控制器,数据库更新算法采用数据驱动的扩展虚拟参考迭代整定算法。上述控制器能够从数据库中选出最贴切当前状态的数据信息,并预测得到控制器参数。以某容量为600MW的超临界机组的典型工况作为测试对象,仿真结果表明,深度调峰下出现工况大范围变化造成的模型失配问题时,上述控制器的控制性能优于常规串级PID控制。The dynamic characteristics of superheated steam temperature control system of supercritical thermal power unit under variable working conditions are time-varying, non-linear, large inertia and time lag, and the conventional cascade PID control can not guarantee its control quality. Based on this, a data-driven control strategy with database updating algorithm is proposed, which can select the most appropriate data information from the database and predict the parameters of the controller. The database updating algorithm adopts data driven extended fictitious reference iterative tuning algorithm. Taking a 600 MW supercritical unit as the test object, the simulation results show that the control performance of the controller is better than that of the conventional cascade PID control when the model mismatch problem caused by a wide range of operating conditions changes occurs under the deep peak load regulation.

关 键 词：超临界机组 变工况 过热蒸汽温度 在线优化 数据驱动控制 扩展虚拟参考迭代整定 

分 类 号：TM743[电气工程—电力系统及自动化]