Modifiye bir gezgin satıcı problemi için açgözlü bir sezgisel ile maliyet minimizasyonu

Photographing a large area in an instant is only made possible by satellites. Satellites are able to help reducing the time required for the photographing/monitoring process, however, satellite images are not available all the time, and even if they exist, they are not easy to evaluate for decision makers. So decision makers use surveillance drones, also called unmanned aerial vehicles to take photos of the predefined region. Given a set of nodes defining an area, each node should be monitored and analyzed. In that sense, the problem may be defined as a variant of Traveling Salesman Problem. It is not practical, however, just to go to every node and take a shot. Instead, one can make use of the concept of relative heights, meaning if there is a node in a higher or more appropriate position than that of another node, drones can go to that higher positioned node, take a photo and are able to monitor the other node that is 'seen' by the current node. In this study, we provide a mathematical model for this modified TSP, in which we should cover all the nodes either by photographing or physical visits and minimize the total travel cost. Then, we provide a greedy heuristic to find solutions and compare the values with optimal solutions as well as lower bounds to evaluate performance. We observe that in low photo costs, our algorithm may provide solutions within 1% of the solutions or lower bounds on hand, and in high photo costs the algorithm is still able to provide good solutions up to 5-10% of the solutions or lower bounds on hand.