In this study, a new robust predictive control technique is investigated for uncertain fractional-order descriptor systems. Using the properties of fractional calculus and the construction of an appropriate Lyapunov function, the sufficient conditions to guarantee the existence of a robust predictive controller are given by minimizing the worst-case optimization problem. The new robust predictive controller is presented by using Shur complement, linear matrix inequality toolbox and cone-complement linear algorithm, and it is proved that the feasible solution of the optimization problem at the initial time can guarantee the admissibility of fractional-order descriptor closed-loop system.

在本研究中，针对不确定的分数阶描述符系统研究了一种新的鲁棒预测控制技术。利用分数阶微积分的性质和适当的李亚普诺夫函数的构造，通过最小化最坏情况优化问题，给出了保证鲁棒预测控制器存在的充分条件。利用Shur补、线性矩阵不等式工具箱和锥补线性算法提出了新的鲁棒预测控制器，并证明了优化问题初始时的可行解能够保证分数阶描述符闭集的可接受性。循环系统。