Build a Real-Time Flight Control Algorithm for the Lighter than Air Indoor Robot Using PX4 Autopilots Support from UAV Toolbox

摘要

With the widespread use of robots, there is a need to develop new designs that are suitable for special applications. These designs can differ mechanically introducing new dynamic characteristics not found in common designs. For these designs, developing adequate control algorithms that run in real-time is considered to be challenging, especially for those who do not have sufficient experience dealing with low-level embedded software. We seek to show the flexibility offered by using MathWorks provided PX4 autopilot support to build customized real-time control flight and vehicle management algorithms. We show the pipeline of autopilot's algorithm design to stably fly an in-house developed lighter than air indoor robot (LAIDR). In our study, we deal with highly nonlinear problems to fly LAIDR due to the properties of the physical body. However, the “PX4 autopilot support for UAV toolbox” allows dealing with these challenges smoothly through developing the algorithm in a modular but integrated form. We evaluated LAIDR by broadcasting live clips of an indoor (theater) event. The results show how efficient LAIDR performs to achieve different shooting angles (e.g., sky view) while following the desired flying path.