论文部分内容阅读
将单输入单输出的Ｈａｍｍｅｒｓｔｅｉｎ模型推广至多输入多输出系统，提出了广义Ｈａｍｍｅｒｓｔｅｉｎ模型，并给出了其辨识建模方法．进一步以此模型为基础，提出了一种多变量非线性预测控制算法．该算法利用了线性预测控制的成果，避免了通常非线性模型预测控制所需的在线数值寻优计算，大大节约了在线计算时间、提高了算法的可靠性和稳定性．
The Hammerstein model with single-input and single-output is generalized to multiple-input-multiple-output system, a generalized Hammerstein model is proposed and its identification modeling method is given. Based on this model, a multivariable nonlinear predictive control algorithm is proposed. The algorithm takes advantage of the results of linear predictive control and avoids the online numerical optimization required for the general nonlinear model predictive control, saves the online computing time greatly and improves the reliability and stability of the algorithm.