Enhancing UAV Path Planning in Dynamic Environments with RRT, BFOA, and Generative AI-based Predictive Models

Globally, the UAVs (drones) are showing its potential across different applications starting from agriculture to military applications. Due to the drone technology, various real time industrial applications turns high productivity. One of the major tasks with UAV is path planning and reaching out the destination without hitting any obstacles. Various path planning algorithms have emerged with lots of features and advantages. However, most of the algorithms failed in the real time environment due to the dynamic changes of the obstacles in the environment. A promising algorithm is proposed using the latest Generative AI (GAN) combing with the traditional RRT and BFOA to predict the paths for UAV in the dynamic environment. The proposed working model is implemented on four different modes as static mode, linear motion mode, circular motion mode and serpentine motion mode. The algorithm turns out with 94.7% accuracy in the path estimation and 96.1% accuracy in obstacle avoidance. While compared with the RRT path optimization alone with the hybridization of RRT and BFOA, the path exploration turns to 98% accuracy. The enhanced model will be very much suited for the dynamic environments for the UAVs.