摘要
针对目前火电厂锅炉燃烧控制系统的大滞后、强耦合、变工况等突出问题,提出了一种基于卡尔曼(CARIMA)模型的自适应预测函数控制方法。该方法先利用预测模型得到系统未来时刻输出,然后将设定输出值和预测值间的预测误差变化率作为自适应控制器的输入,控制器利用最小二乘算法推理得到控制输出。当被控对象模型参数未知或渐时变时,该方法通过实时辨识过程模型的参数来实时修正预测函数控制器的参数,这样可以进一步提高预测函数控制方法的控制品质,提高锅炉的燃烧效率。仿真实验表明,自适应预测函数控制是一种计算简单、鲁棒性和适应性较强、控制精度高的控制方法。
As the boiler combustion system of thermal power plant has variable operating conditions with long time delay and deep coupling,an adaptive predictive functional control method based on CARIMA ispresented for it,which applies the predictive model to forecast the system output of next step and takes the variation rate of the error between the set value and the predicted value as the input of the adaptive controller.The controller output is deduced by the least squares algorithm.When the parameters of the model for the object to be controlled are unknown or gradual-time-varying,the parameters of the predictive functional controller are amended in real time by the real-time parameter identification of process model to further improve the control quality and enhance the boiler combustion efficiency.Simulative results show that,the adaptive predictive functional control is of simple computation,strong robustness and adaptability and high control precision.
出处
《电力自动化设备》
EI
CSCD
北大核心
2010年第4期127-130,共4页
Electric Power Automation Equipment
