A Method with Roulette Selection Strategy for Path Planning in UAV-Based Waste Monitoring Systems

This paper generally targets the problem of unmanned aerial vehicles (UAVs) path planning for usage in various scenarios in urban environments. The proposed method of UAV path planning is designed for the urban waste monitoring scenario. The system is designed to monitor urban green areas, public containers, and unregulated landfills with the goal of detecting possible violations of garbage disposal rules. The proposed method and the system are created for real-time monitoring of the area using drones with programmable path and the detection of irregularities in garbage disposal. This paper is focused on single UAV path planning and path optimization. The path planning method is applicable for a single drone, as well as for a fleet of drones. In this research, the possible methods for path planning based on the Traveling Salesman Problem (TSP) with Genetic Algorithm (GA) and Vehicle Routing Problem (VRP) are initially considered. As a result, the TSP with GA method expanded with the Roulette selection strategy is defined in this paper. The results of the simulation are presented and discussed in this paper, giving the conclusion that proposed model is applicable in proposed scenario.