Improved A* Algorithm AUV Path Planning Based on Multi-Thread Parallelism and CUDA Optimization

In the era of rapid technological advancement, Autonomous Underwater Vehicles (AUVs) have become crucial tools in domains such as oceanographic research, underwater archaeology, and marine biodiversity surveys. However, despite these advancements, a persistent challenge lies in achieving optimal path planning. Given constraining factors like limited battery technology and high-energy consumption of the equipment, it is necessary to find a more efficient way of energy usage. To address the optimal path planning problem for AUVs amidst these challenges, our research introduces an Enhanced A* algorithm based on multi-threading and CUDA optimization in academic discussions and evaluates it through an energy equation. The new algorithm can accurately predict motion cost, accelerate convergence speed to reduce planning time, relieve CPU pressure, and prevent memory crashes. Simultaneously, through the utilization of multi-threading and CUDA optimization, this improved algorithm significantly speeds up the convergence rate, substantially reduces the execution time, and markedly boosts efficiency. Overall, our research presents innovative methods for efficient, safe, and reliable AUV operations, which hold representative practical value for real-world applications.