基于Mamdani模糊神经网络浆液pH值控制研究

摘要

在热电厂石灰石/石膏湿法烟气脱硫系统中,浆液pH值对脱硫效率影响较大.针对浆液pH值的控制过程具有非线性、大滞后和变增益的特点,从脱硫效率的方向出发对浆液pH值控制进行分析与研究.选用了Mamdani模糊神经网络来建立控制器并用于浆液pH值控制中,进一步提高网络的寻优能力以达到更好的控制效果.通过对基于Mamdani模糊神经网络浆液pH值控制进行Matlab设计和仿真,并与常规的PID控制对比,结果证明了该方法的有效性.%In limestone/gypsum wet flue gas desulfurization system in thermal power plant,pH value of slurry has a great influence on desulfurization efficiency.Aiming at disadvantage of nonlinear,large lag and varing gain of slurry pH value control,slurry pH value control is analyzed and studied from desulfurization efficiency.Mamdani fuzzy neural network is used to establish the controller.It is used in the slurry pH value control to further improve search ability of network to achieve better control.Through Matlab design and simulation with control of Mamdani based on fuzzy neural network slurry pH control,as well as comparison with conventional PID control,the results prove effectiveness of the method.