基于支持向量机的非线性内模解耦控制

Nonlinear internal-model control based on support vector machine

针对非线性内模控制在应用于多变量系统时逆模型难以建立的问题,提出了支持向量机α阶逆系统的内模解耦控制方法.该方法利用支持向量机辨识非线性系统的逆模型,并将其串连在原系统之前,运用逆系统方法的思想,将一个多变量、非线性、强耦合的复杂系统通过反馈线性化解耦成多个相互独立的单输入单输出的伪线性复合子系统.对求得的伪线性系统采用内模控制方法进行控制.仿真试验表明该方法不需要系统精确的数学模型,较一般的逆系统方法鲁棒稳定性好,设计简单,跟踪精度高,是解决多变量非线性系统控制的一种可行的理论方法.

To deal with the difficulties of inverse modeling in the nonlinear multi-variable internal-model control, we propose a new internal-model control method based on the support vector machine -order inverse system method. This method uses support vector machine to identify the inverse model of the system and then cascades the-order inverse model with the original system to decouple the multi-variable, nonlinear and strong coupling system into several composite pseudo-linear subsystems. The internal-model control method is applied to the pseudo-linear subsystems. Simulation results show that the combined method does not require an accurate mathematical model and has the characteristics of better robustness stability, easier application and higher tracking accuracy. It is really one of the available methods for designing multi-variable nonlinear systems.