PID神经网络内模控制在湿法烟气脱硫中的应用

Internal model control based on an improved PID neural network for wet flue gas desulphurization

该文针对石灰石/石膏湿法烟气脱硫工艺中吸收塔浆液pH值变化过程的高度非线性、时滞性以及各种不确定性、常规P ID控制难以达到满意的控制效果,提出了一种基于改进P ID神经网络的内模控制方案,对浆液pH值变化过程进行辨识和控制。仿真结果表明,在改进P ID-NN的内模控制下,吸收塔浆液pH值很好地跟踪了系统的设定输入及其变化,体现了高度的自适应性。同时系统超调量小,稳态精度高,优于常规P ID控制。满足实时控制的要求。

In the limestone-gypsum wet flue gas desulphurization(WFGD) process,changes of the slurry pH in the absorber are nonlinear and time-delayed with a large number of uncertainties,so conventional PID controllers do not give satisfactory results.An internal model control strategy was developed based on an improved PID neural network to identify and control the change of the slurry pH.Simulation results with field data from a WFGD system indicate that the system more quickly tracks and more effectively controls changes of the slurry pH with low overshoot and good accuracy for real-time control.