Dynamic Path Planning and Multi-Objective Task Execution for Maneuvering Platforms Based on Deep Reinforcement Learning

With the increasing complexity and variability of modern combat environments, path planning and task execution for maneuvering platforms have become critical technologies for enhancing operational efficiency and success rates. This paper proposes a dynamic path planning and multi-objective task execution method for maneuvering platforms, integrating deep reinforcement learning with rule-based decision-making. The rule-based decision component determines the task execution order, while deep reinforcement learning optimizes path planning and dynamic obstacle avoidance, improving the adaptability and efficiency of maneuvering platforms in complex dynamic environments. Experimental results demonstrate that the proposed method enables efficient path planning and obstacle avoidance in multi-objective tasks. Particularly when facing dynamic obstacles, the maneuvering platform can adjust its path in real time, exhibiting strong adaptability and robustness.