Addressing Climate Change and Port Emissions : An In-Depth Analysis and Optimization of Maritime Trajectory Reconstruction using Hybrid AI Methodologies

The maritime transportation sector is as important as air and land transportation and must operate sustainability, minimizing waste and negative environmental impact. The optimization of vessel trajectories can achieve decarbonization and sustainability in this sector. This research supports climate finance goals by using data and information to make better decisions that reduce pollution and promote an environmentally friendly approach that improves shipping efficiency. In this methodological study, we review current research on vessel trajectory reconstruction, analyzing their strengths and limitations. We propose a novel approach that combines these current methods with advanced Artificial Intelligence (AI) techniques for better accuracy and efficiency in trajectory reconstruction. It is composed of four main steps; First, denoising the Automatic Identification System (AIS) data using Geohash-based Dynamic Time Warping (DTW) then applying U-Net for refining the trajectory grid. Second, extracting the most relevant trajectory features using Ant Colony Optimization (ACO). Third, reconstructing the trajectory using Geohash-DTW for long gaps and kinematic interpolation for short gaps. Finally, applying a temporal refinement using Bidirectional Gated Recurrent Unit (BiGRU) and optimizing the hyperparameters by applying a Genetic Algorithm (GA). Keywords : Trajectory reconstruction, maritime transportation, decarbonization, AI techniques, computational intelligence, energy transition, climate change, AIS data, Dynamic Time Warping (DTW).