基于NUIO的无人机作动器故障检测

针对受未知气流干扰与随机噪声影响的无人机纵向系统进行作动器故障检测研究.在建立固定翼式无人机非线性系统纵向模型的基础上,设计了基于容积卡尔曼滤波(CKF)的非线性未知输入观测器(NUIO).通过构造未知输入观测器结构来解耦未知气流干扰对残差的影响,同时,CKF被算法用于求解观测器增益矩阵,实现了在未知气流干扰解耦情况下残差对随机噪声的鲁棒性.最后,利用残差χ2检验方法判断故障是否发生.仿真结果表明:此方法能有效解耦未知干扰对残差的影响,并快速、准确地检测出了无人机作动器故障. 

NUIO based actuator fault detection for a UAV

The actuator fault detection for an unmanned aerial vehicle (UAV) longitudinal system with unknown atmospheric disturbances and stochastic noise was studied. Based on introducing a nonlinear longitudinal model of the fixed UAV, a residual generation was designed by employing a nonlinear unknown input observer (NUIO) which is based on cubature Kalman filter (CKF). The unknown input observer structure was constructed to decouple the unknown disturbances from residual. At the same time, the CKF was applied to calculate the gain matrix to achieve the requirement of robustness to noise. Finally, the occurrence of fault can be detected based on chi-square test about the residual sequence. The simulation results show that the proposed method can decouple the unknown disturbances from residual effectively and achieve the fast and accurate actuator fault detection.