Construction of a UAV flight trajectory for monitoring agricultural land

When using agricultural lands, one of the tasks is to monitor their condition, in particular soil moisture and the quality of crops. For monitoring, various types of sensors are used, data from which must be periodically transmitted to the processing center. It is not economically feasible to create a wired network of sensors; wireless technologies are used. The lifetime of the sensor network is determined by the battery discharge rate. The least energy consuming is data transmission on demand. In conditions of extensive farmland, such collection can be organized using unmanned aerial vehicles (UAVs). In turn, the cost of collecting data using a UAV is determined by the length of the chosen route. The problem of constructing an optimal flight path for data collection is considered. There may be a shortage of fuel (battery charge in the case of an electric motor) of the UAV and it is necessary to lay out several routes of limited length, the total length of which is minimal. This paper solves the problem of optimizing the trajectory of an individual UAV. An essential feature of the problem is the ability to simultaneously collect data from several sources within the radio visibility radius, which eliminates the need to fly up to each of the sensors. The paper experimentally compares two approaches to solving the problem: solving the traveling salesman problem using previously found points and removing from the list of unvisited points those points that are within the visibility of the next visited one. It is shown that in the latter case, the algorithm that solves the classical traveling salesman problem better may be inferior to a faster and less accurate algorithm for solving the classical problem.