Heuristic Based Online Information Path Planning

Autonomous exploration is the basic research of UAV in many fields, which requires UAV to automatically explore unknown space and obtain complete environmental information. Sampling based method is one of the classical methods of independent exploration. However, although the existing sampling-based methods can complete the task, there are still problems of low detection efficiency caused by UAV maneuvering and low path quality. Current research is focused on improving the efficiency of independent exploration. Therefore, in order to overcome the above problems, this paper proposes a heuristic online information planning method based on sampling. In this method, a heuristic path decision method based on RRT* is proposed, which can better consider the overall environmental information and generate high-quality exploration paths compared with traditional sampling-based methods. Secondly, considering the acquisition of the whole environment information, the heuristic evaluation function is designed, and high-quality viewpoint is continuously selected as the next local navigation target to optimize the path and generate a smooth path. We conducted several simulation experiments to test our approach, and the experimental results show that the algorithm outperforms the benchmark algorithm in terms of both flight time and total path length.