基于动态模糊神经网络的超临界机组协调控制  

Coordinated Control of Supercritical Unit Based on Dynamic Fuzzy Neural Network

作　　者：马良玉[1] 郑佳奕 MA Liangyu;ZHENG Jiayi(School of Control and Computer Engineering,North China Electric Power University,Baoding 071003,China)

机构地区：[1]华北电力大学控制与计算机工程学院,河北保定071003

出　　处：《华北电力大学学报（自然科学版）》2021年第2期96-103,共8页Journal of North China Electric Power University：Natural Science Edition

基　　金：国家自然科学基金资助项目(61174111)。

摘　　要：为改善超临界机组的协调控制品质,研究了基于动态模糊神经网络(DFNN)的机组负荷与主汽压特性逆模型建模方法,借助火电机组全范围仿真机获取的仿真数据进行离线建模。以上述模型为基础,提出一种模型离线训练与在线校正相结合的协调系统DFNN逆控制方案,编制了实时控制算法。通过与仿真机进行实时双向数据交换,开展了详细的协调控制仿真试验。结果表明:采用DFNN逆控制,机组在大幅变工况下负荷与主汽压响应的快速性与机组原PID控制相比有较为显著的提高,有效改善了机组的协调控制品质。In order to improve the coordinated control quality of supercritical units,this paper studied the inverse modeling method based on dynamic fuzzy neural network(DFNN)for the unit load and main steam pressure characteristics.With data obtained from the full-range simulator of the investigated 600MW supercritical unit,we carried out an off-line modeling experiment with the simulation and proposed a neural network inverse coordinated control scheme.The scheme combines model off-line training with on-line correction.We also developed a real-time control algorithm and conducted detailed control simulation experiments through real-time two-way data exchange with the simulator.The results show that compared with the original PID control,DFNN inverse control greatly improves the response rapidity of load and main steam pressure,as well as the coordinated control quality of the unit.

关 键 词：超临界机组 协调控制 动态模糊神经网络 逆模型 在线校正 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]