A Path Planning Decision System for Unknown Pipeline Detection Using Uavs

Abstract

The laying of pipelines is characterized by particularly narrow, wide coverage and sensory degradation, making manual inspection inconvenient. Therefore, we present a path planning method of pipeline detection based on UAV, which includes path planning of the UAV and map construction in the unknown pipeline environment. Due to the limited endurance capacity of the UAV’s battery, the efficiency of UAV exploration should be enhanced in order to explore a larger-scale environment, encompassing the reduction of time spent on path planning, augmentation of the maximum exploration speed permitted by UAV path planning, and minimization of the length of planning paths. Based on the general characteristics of the pipeline environment, the planner provides a future collision free path. At the same time, by setting and updating the anchor points proposed by us, we can avoid UAV turning back to the explored area when it approach the elbow of the pipeline or narrow blocked area. To demonstrate the algorithm's superiority, we conducted a simulation experiment in a virtual environment involving UAV maneuvers such as turning within pipelines or navigating narrow areas. Various data were collected and compared with those obtained using state-of-the-art algorithms.

Keywords: Unknown environment detection, Path planning, UAVs, Kalman filtering.