Return
AUV path planning in complex 3D underwater environments based on an improved TD3 algorithm
DOI:10.1016/j.oceaneng.2025.123688.png)
Abstract
En 中文
• An enhanced TD3 algorithm integrates reward-sensitive PER, N-step returns, and adaptive noise for robust 3D AUV planning. • The method achieves faster convergence, competitive path length with larger obstacle clearance, and sub-50ms inference. • The proposed model-free framework confirms real-time feasibility for AUV navigation in complex dynamic environments.
Journal
IF:
5.5
Papers:
5.8K
Citations:
7.6W

