Adaptive Fault-Tolerant Federated Filter for Micro POS Based on Fuzzy Logic

Micro position and orientation system (POS) is the key motion compensation equipment of low-cost airborne remote sensing systems, whose precision significantly constrains the imaging performance of remote sensing. However, micro POS (MPOS) suffers from harsh environments such as electromagnetic interference and signal obstruction, which leads to measurement outliers and severely affects the motion measurement accuracy. To solve this problem, an adaptive fault-tolerant federated filter based on fuzzy logic is proposed. First, the comprehensive performance evaluation factor is designed based on fuzzy logic to accurately reflect the credibility of measurement information. Then, the adaptive fault-tolerant federated filter algorithm is developed based on the comprehensive performance evaluation factor to dynamically adjust the data fusion weights, allowing for the sequential isolation of outliers and the improved utilization of reliable measurements. The innovation of the proposed method lies in the integration of fuzzy logic into the federated filter. Compared to traditional methods, it significantly improves the utilization of measurement data while effectively isolating outliers. Finally, the experimental results indicate that the proposed method can enhance fault tolerance and have higher motion measurement accuracy compared to traditional methods.