基于约束软化的满意容错控制-输出反馈情形

Satisfactory Fault-Tolerant Control with Soften Constraints-Output Feedback Case

现有的满意容错控制成果难以处理非凸非线性约束,无法用于在线重构容错控制器设计,针对一类离散系统的传感器故障,研究容错系统具有上述特性的闭环区域极点指标、二次型保性能指标和控制输入等约束的满意容错控制问题.借鉴预测控制对多种约束的软化处理策略,利用被控输出反馈控制结构和在线非线性优化技术,给出一种根据故障估计信息进行递推控制的主动满意容错控制器的数值化设计算法,并分析闭环容错系统的稳定性.据此设计的满意容错控制器可使闭环容错系统在保证稳定性和控制输入约束条件下具有满意的优化性能.仿真算例验证文中设计算法的有效性.

In view of the difficulty for the existing satisfactory fault-tolerant control in dealing with nonlinear and non-convex constraints, it is not appealing to the online reconfiguration of active faulttolerant control. Hence, the active satisfactory fault-tolerant control was studied for a class of discretetime systems subject to sensor faults. The study involved the non-convex and nonlinear performance constraints on regional closed-loop poles, quadratic guaranteed performance and control input. Referring to the constraints softening technique in predictive control, a numerical algorithm was developed for the faulttolerant controller reconfiguration by using online nonlinear programming. The time-varying controller with static-output feedback structure was designed from the estimated fault information, which regulated the faulty system in the moving horizon way. The stability of the closed-loop system was also analyzed. Thus, the satisfactory multiple performance of the resulted fault-tolerant system was guaranteed. A numerical example was illustrated to indicate the validity of our proposed method.