动态事件触发的无人机非线性系统故障检测

Dynamic event-triggered fault detection for unmanned aerial vehicles nonlinear system

中文关键词  无人机  非线性系统  故障检测  动态事件触发  Hi/H1优化

英文关键词  unmanned aerial vehicles  nonlinear system  fault detection  dynamic event-triggered  Hi/H1 optimal

中文摘要

      本文在Hi/H1优化框架下, 研究基于动态事件触发的无人机非线性系统故障检测问题. 高空、长航时无人 机需要通过通信网络与地面站进行数据交互, 以实现故障检测等复杂功能. 为了充分利用有限的通信资源, 采用动 态事件触发机制决定是否将测量输入输出数据传送给故障检测模块, 若传输数据不满足触发条件则被丢弃, 因此, 故障检测性能不仅受到干扰和故障影响, 还受到非事件触发时刻数据与实际系统数据误差影响, 即事件触发传输误 差影响. 为此, 针对无人机非线性系统, 提出一种新的动态事件触发Hi/H1故障检测方法. 该方法可以在动态事件触 发条件下, 实现故障检测滤波器残差与事件传输误差完全解耦, 能够避免连续通信和Zeno现象. 在Hi/H1优化框架 下, 通过Riccati方程递归计算, 得到动态事件触发故障检测滤波器的最优解. 最后, 以无人机非线性姿态控制系统为 例, 验证所提方法的有效性和可行性.

英文摘要

      This paper deals with the problem of dynamic event-triggered fault detection (FD) for unmanned aerial vehicle (UAV) nonlinear system in the Hi/H1 optimization framework. High-altitude and long-endurance UAV need to communicate with ground stations through communication networks to realize complex functions such as FD. In order to utilize limited communication resources, a dynamic event-triggered mechanism is considered to determine whether the measurement input and output should be transmitted to the FD module. As such, the FD performance is affected not only by disturbance and fault, but also the error between the data of nonevent time and current actual system data, which is the so-called event-triggered transmission error. To overcome this, a new dynamic event-triggered Hi/H1 optimization FD is presented. The main contribution is the complete decoupling of residuals from event transmission errors in the dynamic event-triggered condition, and the proposed method is able to avoid continuous communication and Zeno phenomenon. In the Hi/H1 optimization framework, the optimal solution of dynamic event-triggered fault detection filter is obtained by recursive calculation of Riccati equation. Finally, an nonlinear attitude control system of the UAV is adopted to demonstrate the effectiveness of the proposed method and feasibility of application.