BIBO稳定的网络控制系统的随机容错控制

Stochastic Fault-tolerant Control for Networked Control Systems with BIBO Stability

针对一类包含网络诱导时延、数据丢包及错序等非理想因素的离散时间网络化控制系统,研究系统在均方有界输入-有界输出(BIBO)稳定条件下,执行器或传感器发生失效故障时的随机容错控制问题.由于传感器与执行器失效与否都是随机的,鉴于此考虑引入相互独立的Bernoulli随机变量序列作为开关矩阵,改进了之前的0,1开关矩阵.为了降低保守性,通过对连接权矩阵进行分解,利用离散形式的牛顿-莱布尼兹公式,依据Lyapunov稳定性理论,结合线性矩阵不等式(LMIs)方法,分别给出传感器失效故障与执行器失效故障情况下的系统均方BIBO稳定条件,基于此条件求解出反馈控制增益矩阵,最后设计出随机容错控制器.

For a discrete time networked control systems with time-delay, date packet losses, wrong-sequence and so, on non-ideal ele- ments, the stochastic fault-tolerant control problem of sensor or actuator failures is investigated under the system having mean square bounded input bounded output （BIBO） stability. Due to the sensor and actuator failure or not are random, in view of this consider, the independent Bernoulli random sequences as switching matrices are introduced, which improves the 0,1 switch matrix. For reduc- ing the conservatism, by decomposing connect weighting matrix, using discrete Newton-Leibniz formula, combining with Lyapunov stability theory and linear matrix inequality （LMIs） method, the mean square BIBO stability conditions are obtained for the system has sensor or actuator failures, based on this stability conditions , we can solve the feedback control gain matrix. At last, the stochas- tic fault-tolerant controller is designed.