Fast Bidirectional A-Star Algorithm Based on Radar Image Data Fusion for Polar Ship Route Planning

Path planning is a significant issue in the current research on ice navigation. Most existing studies utilize sea ice models or ice charts for large-scale polar ship route planning. However, the dynamic trends in weather and ice conditions in polar waters make it challenging for existing research to meet the needs of small-scale, flexible route planning and to accurately and effectively avoid the risk of ice collision. To address this, the study reconstructs the sea ice navigation scene using marine radar imaging technology and applies an optimized bidirectional A-star algorithm to design a polar ice surface ship navigation route planning scheme. The improved bidirectional A-star algorithm takes into account both distance and turning angle, introduces an offset
$p$
to prioritize nodes closer to the destination. Additionally, in this research, the
$Z$
(x,y) function is applied to filter nodes within the closed list, discarding those that fail to satisfy the criteria to prevent irregular paths. To enhance the ship's turning path smoothness, cubic Bezier curves are employed as a replacement for polyline turning paths. Overall, the results of the simulation experiments confirm the effectiveness of the improved bidirectional A-star algorithm in path planning.