Dual-Attention CNN Architectures for Multi-parameter SSE
摘要
The pursuit of ship intelligence aims to strengthen efficiency, innovation, and adaptability across the maritime and offshore sectors, ensuring their readiness for future operational demands. As a cornerstone of the digital transformation framework, ship intelligence constitutes a key element within Europe’s strategic vision for sustainable growth [1]. In recent years, research and industrial efforts have increasingly focused on the conception and deployment of autonomous vessels, which employ advanced intelligent technologies to enhance decision-making accuracy, optimize control performance, lower fuel expenditure, and broaden operational scope [2]. Nevertheless, compared with autonomous ground vehicles, unmanned ships operate in far more intricate and unpredictable environments. The fluctuating interactions of wind and wave dynamics pose the greatest source of uncertainty. Hence, developing a real-time and resilient mechanism for precise SSE has become essential to ensure reliable situational awareness and informed decision-making on board.