In this paper, we extend unconstrained model predictive control (MPC) from setpoint stabilization to dynamic reference tracking for continuous-time nonlinear systems. In particular, we focus on the case when the reference cannot be perfectly tracked by the system due to dynamics and/or constraints. Under the incremental stabilizability assumption and an additional dissipativity assumption, the practical stability of tracking the unknown optimal reachable reference trajectory is proved even though the controller does not know such a reference explicitly.

中文翻译：

---

连续时间非线性系统的无约束跟踪MPC

在本文中，我们将无约束模型预测控制（MPC）从设定值稳定扩展到了连续时间非线性系统的动态参考跟踪。特别地，我们关注由于动力学和/或约束而系统无法完美跟踪参考的情况。在增量稳定性假设和附加耗散性假设下，即使控制器不明确知道这种参考，也证明了跟踪未知最佳可达到参考轨迹的实际稳定性。