Reliability-aware short-term roll prediction for unmanned surface vehicles via multi-task learning and adaptive centralization

2026-09-30

AI Summary

This study addresses the insufficient quantification of reliability in unmanned surface vehicle (USV) roll motion prediction, which compromises navigational safety. To this end, a reliability-aware prediction paradigm is proposed. Methodologically, a multi-task learning dual-head architecture built upon a shared feature extraction backbone is constructed to simultaneously achieve accurate prediction and confidence estimation. Furthermore, a real-time adaptive centering strategy is introduced to enhance model generalization under varying operating conditions. Validation using real-sea data demonstrates that the proposed approach effectively quantifies predictive uncertainty and provides actionable risk indicators. It maintains superior prediction accuracy and robustness across diverse sea states, thereby offering reliable technical support for the safe navigation of USVs.

Abstract

Reliable roll prediction of unmanned surface vehicles (USVs) is essential for ensuring navi?gational safety and enhancing autonomous decision-making. While existing studies primarily focus on improving prediction accuracy, the quantification of prediction reliability remains insufficiently addressed. To bridge this gap, this paper proposes a reliability-aware prediction paradigm that integrates confidence assessment into the predictive pipeline. The architecture utilizes a multi-task learning structure where a shared feature extraction backbone feeds into dual heads: a regression head for precise roll prediction and a quantification head for confidence scoring. This configuration provides accurate prediction and corresponding confidence for risk?sensitive downstream tasks. In addition, an adaptive centralization strategy tailored for short?term real-time roll prediction is introduced to improve model generalization under varying operational conditions. Experiments conducted on a real-sea dataset demonstrate that the proposed method effectively quantifies the reliability of prediction results and maintains superior generalization under varying conditions, offering significant potential for practical engineering applications.

Problem

• unmanned surface vehicles
• roll prediction
• reliability quantification
• confidence assessment
• generalization

Innovation

• Reliability-aware prediction
• Multi-task learning
• Adaptive centralization
• Roll prediction
• Unmanned surface vehicles