Research on Autonomous Obstacle Avoidance Algorithm of Forest Fire Reconnaissance UAV Based on Improved DWA Algorithm

Unmanned Aerial Vehicles (UAVs) play a crucial role in forest fire prevention due to their flexibility, timeliness, and cost-effectiveness. However, the complex and dynamic nature of forest environments can result in mission failures and potential damage to UAVs if they are unable to avoid obstacles in real time. In this paper, autonomous obstacle avoidance technology for UAV reconnaissance is studied for forest fire prevention scenarios. In order to solve the problem that original dynamic window approach (DWA) algorithm is easy to fall into local optimization, an improved DWA algorithm is proposed in this paper. A new target distance function is added to enhance global information acquisition, a safety distance function is introduced to enhance dynamic environmental adaptability, and fusion is carried out in the total evaluation function. Simulation experiments for UAV obstacle avoidance were conducted on the MATLAB platform. The improved DWA algorithm was compared with both the original DWA algorithm and the improved artificial potential field (APF) algorithm. The results indicate that the improved DWA algorithm outperforms the others in terms of flight distance and time consumption under static obstacles, single dynamic obstacles, and multiple dynamic obstacles, thereby validating the effectiveness of the improved algorithm.