约束非线性系统的一个准无限时域预测控制方案(英文)

A QuasiInfinite Horizon Predictive Control Schemefor Constrained Nonlinear Systems*

提出一个具有准无限预测时域的模型预测控制方案．该方案可用于输入和状态约束非线性系统的控制．用优化柔性状态约束条件代替了硬性状态约束，以避免优化问题的不可解．开环优化问题含有附加的终端代价项和终端约束条件，这样预测时域延伸至准无限，而需在线优化获得的控制函数仅为有限时域．如果在最初时刻优化问题有解，则闭环系统具有保证稳定性．

We present in this paper a quasiinfinite horizon model predictive control scheme for stable and unstable nonlinear systems subject to input and state constraints.Hard state constraints are relaxed in an optimal way to avoid infeasibility. With an additional terminal cost in the standard finite horizon objective functional and a terminal inequality constraint,the prediction is expanding quasi to infinity but the control profile to be determined by online optimization is only of finite horizon.Closedloop stability is guaranteed by the feasibility of the optimization problem at the very beginning.