Implementation of the hybrid technology for quadcopter motion control in a complex non-deterministic environment
Dmitry Devitt, Roman Morozov, Mikhail Medvedev, Igor Shapovalov, G. Konovalov · International Conference on Control, Automation and Systems · 2018
This paper represents the results of development and research of the autonomous control system for the UAV. Existing control systems do not provide flight performance necessary for successful operations in non-deterministic environments. So, we propose hybrid intelligent technology based on unstable modes and the conception ofa virtual goal. The main feature of our approach is the efficient integration of global and local path planning providing smooth and quick correction of trajectories for the avoidance of obstacles detected by the quadrotor vision system. Moreover, our local planning approach gives the tool for passing areas with complex obstacles where other local path planning algorithms fail. The hardware-software architecture of the control system based on client-server model with the loose coupling in ROS is presented. The results of simulation and real test flights confirming the efficiency of the proposed technology are shown. We used the commercially available quadcopter and flight controller. That proves absence of necessity in special hardware tools for autonomous flight in environments with uncertainties.