无人机多源冗余传感器容错估计算法研究

Research on Fault-tolerant Estimation Algorithm for Multi-source Redundant Sensors of UAV

在无人机应用中,单余度传感器系统在定位导航精度、多任务处理与决策能力以及系统的鲁棒性和可靠性等方面有待进一步提升和改善。鉴于此,设计三余度传感器的冗余配置,对相应的余度系统采用非相似的滤波算法,提出一种多源冗余传感器容错估计算法对系统软/硬故障进行检测与隔离,通过加权平均投票计算出余度系统的权重因子进而有效地隔离故障系统,保证系统姿态信息的精度和可靠性。试验结果表明,该方法能够实现余度系统的软/硬故障的检测、识别与隔离,具有较好的实时性。

In UAV applications,the single-redundancy sensor system needs to be further improved in terms of positioning and navigation accuracy,multi-task processing and decision-making capabilities,and system robustness and reliability.In view of this,this paper designs the redundant configuration of the three-redundancy sensor,and adopts the non-similar filtering algorithm for the corresponding redundancy system.A multi-source redundant sensor fault-tolerant estimation algorithm is proposed to detect and isolate the soft/hard faults of the system.The weighted average voting is used to calculate the weight factor of the redundancy system and then effectively isolate the fault system to ensure the accuracy and reliability of the system attitude information.The experimental results show that the method can realize the detection,identification and isolation of soft/hard faults in the redundancy system,and has good real-time performance.