Multi-UAV continuous power inspection scheduling algorithm based on NSGA-III

This paper focuses on addressing the optimization scheduling problem of multi-UAV power inspection tasks. In the context of power inspections, it is often not feasible for a single UAV to complete the inspection due to limitations such as power endurance, UAV hardware constraints, and task settings. Consequently, multiple UAVs are required to perform inspections in a continuous manner. The study models the complex scenario of power line inspection as a multi-objective optimization problem, considering various risk factors associated with UAV operations. By examining these risks, the study proposes a solution based on the Non-dominated Sorting Genetic Algorithm III (NSGA-III) algorithm. The proposed approach aims to minimize both the total risk and the maximum risk faced by UAVs during the inspection process. Comparative experiments are conducted to validate the effectiveness of the proposed algorithm. The results demonstrate that, compared to traditional rule-based methods, the NSGA-III algorithm significantly reduces both total risk and maximum risk, ensuring efficient and safe power inspection. Furthermore, the study also considers the failure continuation case, where the remaining UAVs can effectively take over and complete the inspection tasks in case of failure of one UAV. This contributes to the development of more reliable and robust UAV-based power inspection techniques.