基于AGA-Smith预估补偿PID的脱硝系统控制  

Denitration System Control Based on AGA-Smith Predictive Compensation PID

作　　者：孟宏君 王尚尚 张凯奇 李丽锋 MENG Hong-jun;WANG Shang-shang;ZHANG Kai-qi;LI Li-feng(College of Automation and Software,Shanxi University,Taiyuan Shanxi 030013,China;College of Mathematical Sciences,Shanxi University,Taiyuan Shanxi 030006,China;Shanxi Hepo Power Generation Co.,Ltd.,Yangquan Shanxi 045011,China)

机构地区：[1]山西大学自动化与软件学院,山西太原030013 [2]山西大学数学科学学院,山西太原030006 [3]山西河坡发电有限责任公司,山西阳泉045011

出　　处：《计算机仿真》2023年第2期94-101,共8页Computer Simulation

基　　金：国家自然科学青年基金(51605321);山西省自然科学基金(201701D221144)。

摘　　要：为解决SNCR脱硝系统模型精度不高及脱硝控制效果不佳的现状,采集现场运行数据并预处理,利用IPSO算法分别辨识出系统在典型工况170MW和260MW下尿素溶液流量到NO x排放浓度过程的传递函数模型,辨识输出与原始输出的均方根误差值分别为3.13×10^(-2)、7.11×10^(-2)。在电站现场原有单回路PID控制策略基础上,将AGA-Smith预估补偿控制策略引入。仿真结果表明,在两种典型工况下,AGA-Smith预估补偿控制超调量更小,抵抗外来扰动的能力更强,且模型适配能力强于单回路PID控制,为电站现场SNCR脱硝控制提供了良好的技术参考。In order to solve the current situation of poor model accuracy and denitrification control effect of SNCR denitrification system,field operation data were collected and pre-processed to identify the process transfer function model from urea solution flow to NO x emission concentration under typical operating conditions of 170MW and 260MW by IPSO algorithm,and the root mean square error between the identification output and the original output are 3.13×10^(-2)and 7.11×10^(-2),respectively.Based on the original single-loop PID control strategy in the power plant site,the AGA-Smith predictive compensation control was introduced.The simulation results show that the AGA-Smith prediction compensation control has less overshoot and better resistance to external disturbances under the two typical operating conditions,and the model adaptation capability is better than that of the single-loop PID control,which provides a good technical reference for the on-site SNCR denitrification control.

关 键 词：脱硝 模型辨识 预估补偿 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]