Ship ocean voyage weather routing optimization method based on weather clustering

In the contemporary shipping industry, enhancing navigation safety and energy efficiency has become increasingly crucial. Medium- to large-sized vessels are particularly susceptible to adverse weather conditions, which can significantly impact both operational safety and fuel consumption. In this study, meteorological data in proximity to planned shipping routes are collected and processed using K-means clustering to identify dominant weather patterns. Suitable indicators are employed to delineate the boundaries of these weather clusters. Based on risk assessment, the identified clusters are categorized into two types: hazardous weather clusters and navigable weather clusters. For hazardous weather clusters, an improved artificial potential field method with an adjusted repulsion radius is employed to generate avoidance waypoints at appropriate distances. Risk contours are subsequently plotted to visualize the spatial distribution of navigation risks. For navigable weather clusters, a fuel consumption model is used to evaluate energy efficiency, identify optimal fuel consumption points, and generate corresponding fuel consumption contours. The two types of waypoints are integrated to form a weather-clustered route, which is further evaluated using the carbon intensity indicator (CII). Experimental cases demonstrate that the proposed method significantly reduces fuel consumption while mitigating weather-related navigation risks. Moreover, as meteorological conditions become more complex, the method adapts well and exhibits excellent energy-saving and emission-reduction effects.