基于观测器与小波变换的UAV作动器故障检测

Observer and Wavelet Transform based Actuator Fault Detection for UAV

为提高无人机飞行控制系统的安全性和可靠性,将观测器方法与小波变换相结合,针对无人机飞行控制系统作动器故障检测问题开展研究。在基于观测器构造残差的基础上,通过引入小波变换,产生一组包含不同频率作动器故障的残差,将残差产生器设计归结为小波基函数选取和H_2/H_2性能指标最优化问题。利用小波变换的时频局部化特性,在保证性能指标最优的同时,可减小故障检测延时,且产生的一组残差能够检测较宽频带内的作动器故障。基于某型无人机的Simulink仿真结果表明,该方法的检测快速性好,可检测故障的频率范围宽,且对闭环飞行控制系统中的微小故障具有较高的灵敏度,有效降低了漏报率。

For improving the security and reliability of unmanned aerial vehicle（ UAV） flight control system（ FCS）,this paper combined the observer-based method w ith w avelet transform（ WT） for the UAV FCS actuator fault detection. By introducing WT to observer,a group of residuals containing different frequency of actuator faults are obtained,the design of residual generator is converted to the selection of w avelet basis and H_2/H_2 optimization problem. For the good location of WT in time and frequency domain,the performance index is optimal,the detection time-delay is smaller,and the group of residuals can detect faults w ithin a broader frequency band. The simulation results of Simulink w ith a UAV show that,this method can achieve high detection speed,broad frequency band of detectable faults,and be sensitive to small faults in closed- loop FCS,low ering the miss detection rate effectively.