变约束高效预测控制

摘要

To improve the optimality of the off-line synthesis algorithm for robust model predictive control, we introduce the idea of on-line variable terminal constraint set. Based on the off-line results, the convex combination coefficients for the state feedback matrix are obtained through on-line optimization; and then, the current control is generated. Compared with the original method, the improved one increases both on-line freedom and optimality. On this basis, we further consider the variable constraints for the efficient predictive control algorithm(VC-MPC). The corresponding control constraints are determined according to the elliptical stability domain where the current states are lying in. The on-line computational load of VC-MPC is reduced significantly while achieving the same optimal performance with the constant constraint algorithm. Finally, the simulation results illustrate the effectiveness of the proposed methods.%为改善离线鲁棒预测控制算法的最优性引入了变终端约束集的思想,即在原离线方法基础上通过在线优化方法获得构建当前反馈矩阵的凸组合系数,然后再产生当前控制量.与原方法相比增加了在线求解的自由度,改善了系统的最优性.在此基础上,又提出了变约束预测控制(VC-MPC)算法,根据状态所处的不同稳定椭圆域确定其对应的控制约束,在取得相同优化性能情况下较恒定约束算法明显节省了计算时间.仿真实例验证了所提方法的有效性.