线性参数时变可测系统的混合反馈预测控制

针对参数可测的线性参数时变（LPV）系统，通过综合增益调度控制思想和多步控制集方法，设计鲁棒反馈预测控制器.借鉴增益调度控制扩大设计自由度，改善系统控制性能的优点，设计基于增益调度控制和普通反馈控制律序列的多步混合反馈预测控制策略.在首步椭圆集采用增益调度控制律，有效地利用可测的参数信息提高系统控制性能；在余下的椭圆集系列采用多步控制律，降低在线优化算法的计算量，在控制性能和计算负担之间进行有效平衡.利用该混合反馈预测控制器可以有效地利用当前可测的参数信息，且以较低的在线计算量取得了较优的控制性能.仿真结果验证了该方法的有效性.

Mixed multi-step feedback model predictive control for linear varying systems with measurable parameters

A mixed feedback robust model predictive control (RMPC) approach was developed based on multi-step control sets and scheduling control for linear parameter varying (LPV) systems with the varying parameters measurable. The mixed feedback RMPC consists of a feedback control sequence and the parameter-dependent control, which has the merits of enlarged freedom of design and improved control performance. A parameter-dependent feedback control was applied to utilize the information on parameter value and enhance the control performance for the first step in the ellipsoid sets. A single feedback control was adopted to reduce the computational burden and achieve balance between the computational burden and the control performance for the remained ellipsoid sets. The proposed method achieved high control performance with lower on-line computational burden through use of the measured value of varying parameter. The effectiveness of the proposed approach was validated by a simulation.