基于多面体不变集的变终端约束集RMPC

Robust Model Predictive Control with Time-varying Terminal Constraint Set Based on Polyhedron Invariant Set

针对一类状态和输入受约束的多胞不确定线性时变系统,提出了一种基于多面体不变集的变终端约束集鲁棒模型预测控制算法。首先采用基于状态反馈增益的多面体不变集计算方法,给出了一种新的控制不变集序列构造方法,然后以控制不变集序列的并集作为终端约束集,结合在线优化和增益切换,实施变终端约束集双模鲁棒预测控制。该算法不仅有效地扩大了终端约束集,而且采用离线方式计算不变集序列降低了在线运算量。最后仿真结果验证了算法的有效性。

A robust model predictive control algorithm with a time-varying terminal constraint set is presented for a polytopic uncertain linear time-varying system with input and state constraints. A method for constructing a sequence of polyhedron invariant sets is presen- ted by using the polyhedron invariant set algorithm with state feedback gain. And then the union of the sequence of control invariant sets is treated as the terminal constraint set, and combined the on-line optimization and the gain switching, the dual-mode predictive control algorithm with a time-varying constraint set is designed. The proposed method can enlarge the terminal constraint set effectively, and reduce the computation load by computing polyhedron invariant set off-line. Simulation result shows the advantages of the proposed method.