Autonomous Control of a Hybrid Rolling and Flying Caged Drone for Leak Detection in HVAC Ducts

In this article, a hybrid, rolling, and flying multirotor caged drone is developed for inspection and leak detection of heating, ventilation, and air conditioning (HVAC) ducts. The developed passive-cage drone is designed, manufactured, and controlled autonomously using direct model reference adaptive control (DMRAC). The platform is equipped with a 2-D LIDAR sensor that provides a scan of the surroundings in a plane. Possible leaks are detected using a thermal camera. DMRAC is used for higher level roll, pitch, and yaw control, while lower level motor speed control is achieved through Proportional-Integral-Derivative (PID) controllers. This article compares, for the first time, PID-based controllers for roll, pitch, and yaw control, with standard DMRAC and DMRAC with gain modifications, in a cascaded setting on a hybrid passive-caged drone. Several hardware tests are used to highlight the differences in the control performance when PID controllers/DMRAC with certain gain modifications are used to control the developed hybrid passive-caged drone.