An Improved Genetic Algorithm for the Dubins Multiple Traveling Salesman Problem with Neighborhoods

In this paper, we focus on the problem of unmanned aerial vehicles (UAVs) engaged in reconnaissance and imaging missions. The aim is to distribute multiple UAVs to diverse target nodes, and ensuring that imaging collection is conducted with high resolution, reliability, and cost-effectiveness. Considering the limits of side-looking imaging mode, UAVs must navigate through target area at precise angles, this problem can be described as Dubins Multiple Traveling Salesman Problem with Neighborhoods (DMTSPN). To solve this, we introduce an improved genetic algorithm (IGA). Regenerate the population when facing local optima. Through simulation experiments, we demonstrate our improved genetic algorithm significantly accomplishes the tasks.