An Effective Wind-aware Path Planning Heuristics for UAV-assisted Bridge Inspection

Abstract Utilizing unmanned aerial vehicles (UAVs) for bridge inspections offers significant flexibility. However, weather conditions can significantly impact the phenomenon in question, with meteorological factors such as wind and rain being particularly influential. To facilitate effective bridge inspections, this paper investigates the path planning problem for the UAV-assisted bridge inspection system, with consideration given to the impact of wind conditions. A heuristic utilizing simulated annealing is proposed to address the problem under study. The algorithm accounts for wind factors, simultaneously considering various wind speeds and directions on UAVs’ path planning. To begin with, a dedicated environment model designed for UAV bridge inspection is constructed, considering the diverse constraints and objective functions for UAVs. Subsequently, a more complex and exact mathematical model, building upon this model, is developed to facilitate efficient path planning of UAVs. Secondly, an innovative application of the bridge separation planning model is implemented, along with multiple parameters being simulated, which include adjusting the initial value of temperature. The experimental findings indicate that the proposed method reduces costs by 30.05% and significantly enhances effectiveness compared to traditional local search algorithms. In contrast to the methods of path planning that overlook factors related to wind, the approach presented in this paper delivers more pragmatic and feasible outcomes for the applications of UAVs.