脱硫系统多变量扰动抑制预测控制方法

Predictive Control Method for Multivariable Disturbance Rejection in Desulphurization Systems

脱除烟气中的二氧化硫是燃煤电厂烟气处理不可缺少的一环。由于脱硫系统具有多变量、强耦合的特点,只根据解耦控制或针对主对角元设计的扰动,观测器难以实现对扰动的准确估计,尤其在模型偏差较大时,将对象间的耦合关系当做扰动处理往往会导致控制效果恶化,难以实现预期效果。因此提出一种多变量扰动观测器,基于多变量离散状态空间模型,结合历史运行数据估计前一时刻的扰动,并进行补偿。这种扰动观测器可以考虑不同通道之间的耦合关系,保证了扰动估计的准确性和合理性,然后通过仿真验证了该方法的有效性。

Removing sulfur dioxide from flue gas is an indispensable part of flue gas treatment in coal-fired power plants.Due to the multivariable and strong coupling characteristics of desulfurization systems,it is difficult to accurately estimate disturbances based solely on decoupling control or disturbance observers designed for the main diagonal elements.Especially when the model deviation is large,treating the coupling relationship between objects as disturbances often leads to deterioration of control effectiveness and difficulty in achieving the expected results.Therefore,this paper proposes a multivariable disturbance observer,which is based on a multivariable discrete state space model and combines historical operating data to estimate the previous disturbance and compensate for it.This disturbance observer can consider the coupling relationship between different channels,ensuring the accuracy and rationality of disturbance estimation.Then,the effectiveness of this method is verified through simulation.