摘要
针对离散不确定模糊模型,提出了一种鲁棒非线性模型预测控制方法。导出了预测性能指标上界,将稳定性约束、输入约束和输出约束变换成容易求解的线性矩阵不等式（LMI）形式,从而将非线性模型预测控制Min-Max优化问题变换成具有线性矩阵不等式约束的广义特征值问题（GEVP）。对于采用的状态反馈预测控制器,讨论了滚动时域优化的可行性,证明了闭环系统的鲁棒稳定性。仿真结果表明了控制器的有效性。
A robust model predictive control (MPC) scheme based on fuzzy modelling is proposed for a nonlinear system. The upper bound of predictive cost is derived, constraints on inputs and outputs are transformed into linear matrix inequalities (LMI) which are easy to solve, and then minmax optimization of nonlinear model predictive control is transformed to generalized eigen value problem(GEVP) with LMI constraints. State feedback based predictive controller is used in the scheme. Furthermore, feasibility and robust stability of the resulting moving horizon optimization are proved. Simulation shows the validity of the moving horizon controller.
出处
《电机与控制学报》
EI
CSCD
北大核心
2002年第3期229-232,236,共5页
Electric Machines and Control
基金
黑龙江省自然科学基金资助项目
教育部科学技术研究重点资助项目
