Gu Zhou, Ahn Choon Ki, Yue Dong, Xie Xiangpeng
IEEE Trans Cybern. 2022 Jun;52(6):5311-5321. doi: 10.1109/TCYB.2020.3030028. Epub 2022 Jun 16.
This article focuses on the problem of resilient H filtering for Takagi-Sugeno fuzzy-model-based nonlinear networked systems with multisensors. A weighted fusion approach is adopted before information from multisensors is transmitted over the network. A novel event-triggered mechanism is proposed, which allows us not only to reduce the data-releasing rate but also to prevent abnormal data being potentially transmitted over the network due to sensor measurement or other practical factors. The problem of denial-of-service (DoS) attacks, which often occurs in a communication network, is also considered, where the DoS attack model is based on an assumption that the periodic attack includes active periods and sleeping periods. By employing the idea of the switching model for filtering error systems to deal with DoS attacks, sufficient conditions are derived to guarantee that the filtering error system is exponentially stable. Simulation results are given to demonstrate the effectiveness of the theoretical analysis and design method.
本文聚焦于具有多传感器的基于Takagi-Sugeno模糊模型的非线性网络系统的弹性H滤波问题。在多传感器信息通过网络传输之前，采用加权融合方法。提出了一种新颖的事件触发机制，该机制不仅能降低数据释放率，还能防止由于传感器测量或其他实际因素导致异常数据潜在地在网络上传输。还考虑了通信网络中经常出现的拒绝服务（DoS）攻击问题，其中DoS攻击模型基于周期性攻击包括活跃期和休眠期的假设。通过采用滤波误差系统的切换模型思想来处理DoS攻击，推导了充分条件以保证滤波误差系统指数稳定。给出了仿真结果以证明理论分析和设计方法的有效性。
