Search and Rescue Operations using a Fixed-Wing UAV Equipped with an Automatically Controlled Gimbal
Abstract
This paper describes the development of methods to be used in search and rescue operations when using fixed wing unmanned aerial vehicles (UAVs) with a controllable gimbal, including methods for using the gimbal in an object-tracking scenario. A complete payload system to be used on a wide range of different UAVs has been developed and thoroughly tested. The level of autonomy of the system reduces the need for onsite personnel, and encourages increased effort in the rescue part of the operation. The system gives personnel at a ground station the ability to control the UAV and its features using an intuitive interface. One of the primary applications of the payload is to compute a search pattern that covers a given search area efficiently and feed waypoints along the path to the onboard autopilot. In order to increase the area of coverage and at the same time reduce the total flight path needed to cover the search area, active sweeping methods by using the gimbal have been implemented. Another major application is object tracking based on real-time information given by a machine vision module coupled with a thermal imaging camera attached to the gimbal system. Different trajectory planning methods for the gimbal are presented. These are especially useful to utilize when the configuration space of the gimbal is restricted, and replaces the need of constraint handling optimization methods. Trajectory planners based on polynomial trajectories proved to be the most efficient planner, although sample based trajectories are easier to use in a non-convex gimbal configuration space. The trajectory planning methods presented can just as easily be implemented as trajectory planners for the UAV itself, since desirable waypoints can be expressed explicitly.