The accurate estimation of solar radiation is crucial for optimizing solar energy deployment and advancing the global energy transition. This study pioneers the development of a hybrid model combining Recurrent Neural Networks (RNNs) and the Weather Research and Forecasting (WRF) model to estimate solar radiation in tropical regions characterized by scarce and low-quality data. Using datasets from Sumedang and Jakarta across five locations in West Java, Indonesia, the RNN model achieved moderate accuracy, with R² values of 0.68 and 0.53 and RMSE values of 159.87 W/m² and 125.53 W/m², respectively. Additional metrics, such as Mean Bias Error (MBE) and relative MBE (rMBE), highlight limitations due to input data constraints. Incorporating spatially resolved GHI data from the WRF model into the RNN framework significantly enhanced accuracy under both clear and cloudy conditions, accounting for the region’s complex topography. While the results are not yet comparable to best practices in data-rich regions, they demonstrate promising potential for advancing solar radiation modeling in tropical climates. This study establishes a critical foundation for future research on hybrid solar radiation estimation techniques in challenging environments, supporting the growth of renewable energy applications in the tropics.

Energy and power management systems (EMS/PMS) are essential for electric-propulsion ships, affecting propulsion performance, fuel consumption, emissions, and component lifetime. As shipboard power systems integrate heterogeneous energy resources and face nonlinearity, uncertain load demand, and multi-source interactions, deep reinforcement learning (DRL) has emerged as a promising adaptive, sequential decision-making tool in shipboard EMS/PMS. This perspective reviews DRL studies through a hierarchical decision-making framework comprising power dispatch, energy coordination, and operational strategy. Most research focuses on real-time power dispatch, while emerging research addresses energy coordination via multi-source cooperation, multi-objective operation, degradation awareness, and uncertainty handling. However, operational strategy remains underexplored, despite its role in speed control, route-aware planning, predictive operation, and voyage scheduling. This paper argues that future shipboard EMS/PMS adopt integrated hierarchical DRL frameworks across all three decision layers, leveraging DRL’s strengths in sequential policy learning in dynamic environments and supporting multi-time-scale decision-making. This paper clarifies current research trends, identifies gaps, and outlines future directions toward adaptive, reliable, and autonomous shipboard EMS/PMS in next-generation electric-propulsion ships.