针对非平稳频率方位量测的水下目标定距技术

To address the challenges of nonstationary measurement noise and unknown statistical characteristics in underwater environments, a measurement anomaly detection and correction mechanism that integrates motion variation constraints with adaptive noise estimation was developed. Given that autonomous underwater vehicles are characterized by low-speed navigation, a motion model was constructed and a dynamic thresholding strategy based on innovation analysis was introduced to enable real-time identification and correction of abrupt measurement anomalies. The Sage-Husa algorithm was also incorporated to enhance the adaptability of the filter to environments with unknown noise statistics, allowing online adjustment of the measurement noise under dynamic conditions. Both simulation and experimental results demonstrate that the proposed method achieves a 13. 16% reduction in ranging error compared to the traditional approach, indicating enhanced robustness and adaptability undernon-stationary measurement noise conditions.