Task allocation for inspection drone of power transmission lines based on improved NSGA-II

During the research on unmanned aerial vehicle (UAV) power inspection process, the problem of UAV task allocation was studied, and multi-objective optimization was performed: minimizing the sum of UAV total path lengths and minimizing the longest working time of UAVs. This problem was formulated as the multiobjective multiple traveling salesperson problem (MOMTSP), and a hybrid integer programming multiobjective optimization scheduling model was constructed. To solve this problem, an improved non-dominated sorting genetic algorithm (AE-NSGA-II) was proposed, which can adaptively adjust the crossover and mutation rates and enhance the exploration to find more optimal solutions in both objective function directions. In the experiments, this algorithm was utilized to find a set of optimal solutions on the Pareto front.