A novel Rauch-Tung-Streibel smoothing scheme based on the factor graph for autonomous underwater vehicles

Area of Science:

• Robotics
• Navigation Systems
• Signal Processing

Background:

• Autonomous underwater vehicles (AUVs) require precise navigation for surveying and mapping.
• Traditional Kalman filters struggle with multi-sensor data fusion, especially with asynchronous data or sensor failures.
• Offline navigation solutions are crucial for refining AUV mission data.

Purpose of the Study:

• To develop an improved offline navigation solution for AUVs.
• To enhance multi-sensor data fusion by addressing asynchronous data and sensor failure issues.
• To combine the strengths of factor graphs and Rauch-Tung-Streibel smoothing for superior navigation accuracy.

Main Methods:

• Application of the factor graph method for optimal multi-sensor information fusion.
• Implementation of a revised Rauch-Tung-Streibel smoothing (RTSS) as a post-mission algorithm.
• Recursive updating of smoothed state and covariance using backward data processing.

Main Results:

• The factor graph offers plug-and-play capability and improves real-time navigation accuracy over standard Kalman filtering.
• The RTSS significantly enhances position, velocity, and attitude accuracy and smoothness.
• The combined factor graph and RTSS method demonstrates superior performance, particularly during signal loss or sensor malfunctions.

Conclusions:

• The proposed hybrid factor graph and RTSS scheme effectively improves AUV offline navigation.
• The method is reliable and effective, validated through semi-physical experiments.
• This approach offers a robust solution for challenging underwater navigation scenarios.