Dynamic Guidance Point Obstacle Avoidance Planning Method and Optimization Based on DWA

Currently, global-local two-layer path planning is usually adopted when robots perform area coverage cleaning tasks in unfamiliar environments. However, due to the limited detection range of the sensors carried by the robot, when the robot encounters unknown obstacles during the planning of local paths, the robot is prone to deviate from the original planned route during obstacle avoidance, resulting in a decrease in the coverage rate of the cleaning task. To solve this problem, this paper proposes a dynamic guided point obstacle avoidance method based on the Dynamic Window Approach (DWA). The method is that when the robot performs the cleaning task, through constantly updating the perception information and environmental data, the system dynamically selects the guide point on the global path as the temporary target point, and then constantly adjusts the driving direction and speed to ensure that the robot can return to the preset global path while avoiding obstacles. In addition, the evaluation function is improved for this application scenario. This enables the robot to safely and smoothly avoid obstacles when operating autonomously in unfamiliar environments, navigate and travel autonomously in the global path, and improve cleaning task coverage.