航空学报 > 2022, Vol. 43 Issue (3): 325193-325193   doi: 10.7527/S1000-6893.2021.25193

网络攻击下无人机信息物理系统的自适应状态估计

李笑宇1, 冯肖雪1, 潘峰1,2, 蒲宁3   

  1. 1. 北京理工大学 自动化学院, 北京 100081;
    2. 北京理工大学 昆明产业技术研究院, 昆明 650106;
    3. 云南省遥感中心, 昆明 650034
  • 收稿日期:2020-12-31 修回日期:2021-01-23 出版日期:2022-03-15 发布日期:2021-03-01
  • 通讯作者: 冯肖雪 E-mail:fengxiaoxue@bit.edu.cn
  • 基金资助:
    国家自然科学基金(61433003)

Adaptive state estimate for unmanned aircraft cyber-physical system with cyber attack

LI Xiaoyu1, FENG Xiaoxue1, PAN Feng1,2, PU Ning3   

  1. 1. School of Automation, Beijing Institute of Technology, Beijing 100081, China;
    2. Kunming Industry Technology Research Institute INC, Beijing Institute of Technology, Kunming 650106, China;
    3. Yunnan Remote Sensing Center, Kunming 650034, China
  • Received:2020-12-31 Revised:2021-01-23 Online:2022-03-15 Published:2021-03-01
  • Supported by:
    National Natural Science Foundation of China (61433003)

摘要: 针对网络攻击下无人机信息物理系统(CPS)的安全状态估计问题,提出了一种基于自适应方差极小化的递推状态估计器(AVMRE)。通过将针对控制输入和传感器数据的恶意攻击分别建模为状态和量测方程中的未知干扰项,建立了未知干扰解耦状态递推估计器,实现滤波误差中的量测未知干扰解耦,利用滤波残差设计自适应调整因子对估计误差上界进行极小化,应用最小方差估计准则求解出算法中的量测增益反馈矩阵。同时引入事件触发机制,使得系统在保持一定估计精度的情况下节省通信资源。此外,给出了滤波误差指数有界性的充分条件。无人机飞行模型仿真验证了本文算法相比传统算法的有效性和优越性。

关键词: 未知干扰输入, 事件触发, 自适应方差极小化, 状态估计, 干扰解耦, 网络攻击, 信息物理系统

Abstract: An Adaptive-Variance-Minimization-based Recursive Estimator (AVMRE) is proposed for the problem of secure state estimation of Unmanned Aerial Vehicle (UAV) Cyber-Physical System (CPS).By modeling the data attacks on control commands and sensors as the unknown disturbances in the state and measurement equations respectively, an unknown disturbances decoupling state recursive estimator is established, which realizes the unknown measurement disturbances decoupling in filter error.Then, the adaptive adjust factor is designed by the filter residual to minimize the upper bound of the estimate error.Finally, the measurement gain feedback matrix in the algorithm is deduced based on the minimum variance estimation criterion.The distributed event-trigger mechanism is also considered, so that the system can save communication resources in the case of maintaining certain estimation accuracy.In addition, a sufficient condition for the filter error exponential boundedness is given.Simulation results of the UAV flight controller show the effectiveness and superiority of the proposed algorithm over traditional methods.

Key words: unknown disturbance input, event-triggered, adaptive variance minimization, state estimate, disturbance decouple, cyber attack, Cyber-Physical System (CPS)

中图分类号: