Energy-Efficient UAV routing problem based on approximate cellular decomposition for geohazards monitoring

The emerging use of unmanned aerial vehicles (UAVs) in monitoring geohazard-prone areas for collecting information and preventing disasters has gained popularity. Due to the limited battery capacity, the main concern for such applications is how to precisely measure the energy consumption of UAVs and plan the most efficient patrolling paths. Moreover, onboard cameras or sensors equipped on UAVs are typically unable to capture the entire geohazard-prone site in one shoot screen at a specific altitude. Therefore, it is necessary to divide geohazard-prone areas into smaller sections and then consolidate images for risk evaluation. In this study, we investigate an energy-efficient UAV routing problem (EURP) for monitoring the dispersed geohazard-prone sites. The objective of the problem is to determine the most efficient UAVs’ routes with minimal energy consumption while ensuring coverage of all target areas. We build an energy consumption model for UAVs in different flight modes, such as straight-line flight, turning maneuvers, and hovering. An approximate cellular decomposition technique is then introduced to discretize geohazard-prone areas into square grids to obtain image information with specific accuracy. To solve the proposed EURP, we develop an energy grid hybrid metaheuristic (EGHM) utilizing the large neighborhood search (LNS) for solution exploration and employing the variable neighborhood descent (VND) for post-optimization, respectively. A set of efficient destruction and repair operators are customized in the LNS based on the features of UAV energy consumption and geohazard-prone area gridding. Finally, the proposed approach is evaluated on benchmark instances and dataset of geohazard-prone areas in Shaanxi Province, China.