Resilient fusion navigation based on failure influence level evaluation
Fault subsystems are entirely isolated on the fault-detected epoch in traditional fault-tolerant integrated navigation systems. However, this strategy ignores the asymptotic change and component differences of the soft failure influence. This study proposes a resilient fusion navigation algorithm based on the failure influence level evaluation. The failure influence mapped on the local estimation components is modelled. Then, the influential level of soft failure on the locally estimated states is evaluated. The failure influence levels are used for asymptotic resilience tuning in the global fusion. The simulation results indicate that the fault-tolerant performance under soft failure is improved significantly when using the proposed resilient fusion navigation algorithm based on the failure influence level evaluation.