Research on a path optimization algorithm based on enhanced DDPG for continuous dynamic obstacle environments

This paper proposes a Continuous Dynamic Obstacle Environment Path Optimization Algorithm (DDAPG) based on Enhanced DDPG, which focuses on solving the problem of path planning for robots in continuous action spaces. The algorithm effectively guides the robot to move towards the target position and punishes collision behavior by designing a reward reshaping method based on potential energy. At the same time, an adaptive noise model is used to enhance exploration ability and avoid falling into local optima. The experimental results show that the DDAPG algorithm outperforms the baseline algorithm in terms of convergence speed, collision rate reduction, and adaptability, and can achieve efficient and safe path planning in different scenarios. This study provides strong support for the application of embodied intelligence in the field of robot path planning and promotes the development of autonomous decision-making technology for robots.