Stochastic Event-Triggered Fault Detection and Isolation Based on Kalman Filter
针对有色测量噪声和未知输入的系统，提出一种基于随机事件触发调度的鲁棒故障检测与隔离卡尔曼滤波方法，通过凸优化选择参数以降低通信率并提升性能，在三相感应电机故障中验证。
In this article, the robust fault detection and isolation (FDI) Kalman filter is extended based on stochastic event-triggered schedulers for discrete linear systems consisting of deterministic/stochastic unknown inputs with nonzero mean and colored measurement noise. In the proposed FDI method, first, a subspace of the main system that significantly attenuates disturbance effects is proposed. After that, the fusion method is proposed for dealing with the colored measurement-noise problem in designing the Kalman filter and preventing from leading to measurement noise with zero mean and zero covariance. The stochastic event-triggered schedulers are considered as Gaussian functions. Therefore, the Gaussian property of innovation sequence is preserved, and consequently, the recursive equation of the Kalman filter need not be extended based on approximation techniques. Finally, design parameters are chosen based on the convex optimization problem such that the lowest communication rate between the sensor node and FDI filter, and also the best FDI performance are achieved. The proposed FDI method is evaluated to detect and isolate stator interturn short circuit and broken rotor bars faults with unbalanced voltage as disturbance in three-phase induction motors.