针对一类具有传感器故障的闭环网络化系统,推导了一种新的容错估计方法。闭环网络化系统意味着在传感器到估计器信道和控制器到执行器信道中,都可能存在由网络引起的随机延迟和数据包丢失。由两组随机变量单独驱动的指示函数用来表征两个通道中网络化所产生的影响。在考虑存在传感器故障的情况下,设计状态估计器。为处理闭环网络化系统中存在的传感器故障,开发了一种无偏最小方差容错估计器。最后,在一个实际的网络化永磁同步电机驱动系统上进行的仿真,验证了所提方法的可行性和有效性。 For a class of closed-loop networked systems with sensor failure, a new fault-tolerant estimation method is derived. Closed-loop networked systems mean there may be network-induced random delays and packet loss in both sensor-estimator channels and controller-to-actuator channels. An indicator function driven by two random variables alone is used to characterize the effects of networking in both channels. The state estimator is designed considering the presence of sensor failure. To deal with the sensor failure in closed-loop networked systems, an unbiased minimum-variance fault-tolerant estimator was developed. Finally, the simulation on a real networked PMSM drive system verifies the feasibility and effectiveness of the proposed method.