论文部分内容阅读
针对一类不确定非线性可逆系统,提出了一种基于最小二乘支持向量机(LS-SVM)逆系统方法的动态矩阵解耦控制方法.利用LS-SVM辨识得到原系统的逆模型,并采用逆系统思想,将一个非线性系统解耦成多个相互独立的单入单出伪线性子系统.为了克服建模误差影响及外部干扰,对所得的伪线性系统采用动态矩阵预测控制(DMC)方法进行闭环控制.以一个典型的2输入2输出强耦合微型燃机系统为例对该解耦控制方法的有效性进行仿真验证,结果表明该控制方法鲁棒性好,结构简单,易于工程实现.
Aiming at a class of uncertain nonlinear reversible systems, a dynamic matrix decoupling control method based on least square support vector machine (LS-SVM) inverse system is proposed. The inverse model of the original system is obtained by LS-SVM identification. Inverse system theory is used to decouple a nonlinear system into several independent pseudonlinear subsystems with single input and single output.In order to overcome the impact of modeling error and external interference, the dynamic matrix predictive control (DMC ) Method for closed-loop control.The effectiveness of the decoupling control method is verified by a typical 2-input-2 output strong coupling micro-turbine system.The results show that this control method is robust, simple and easy to implement achieve.