Bilateral haptic teleoperation of vtol aerial robots in complex cluttered environment

Abstract

This thesis focuses on investigating the haptic interfaces, obstacle avoidance algorithms and control schemes for haptic teleoperation of VTOL aerial robots in the cluttered three-dimensional environment. An admittance haptic interface is developed that provides spatial haptic cues that can better help a pilot feel the motion of a slave robot and achieve navigation tasks in the cluttered environment than a classical impedance haptic interface. Different state mapping strategies between the master and slave devices are introduced and discussed. An analysis of an operator's perception of a vehicle's motion was conducted for both admittance and impedance haptic interfaces. Experiments were carried out on an aerial robotic platform to verify the performance of the proposed approach. Two user studies were conducted to investigate the applicability of the novel approach and compare with an impedance haptic interface. The results from the user studies verified the author's claim that the proposed admittance haptic interface outperforms the classical impedance haptic interface in perceiving the slave vehicle's motion and navigating through cluttered environment. Stability analysis of the proposed approach was provided as well as a discussion on the controller parameter tuning for optimum stability and performance. A disadvantage of the admittance configured master joystick is its inability to render virtual environmental forces. A novel dynamic kinesthetic boundary approach was proposed to overcome this issue and provide the admittance haptic interface with an effective method to render the local environment to an operator and aid in obstacle avoidance. This approach converts the spatial measurements of the remote environment into a virtual boundary in the master joystick's haptic workspace, such that the human operator can perceive the environment by touching the boundary. The proposed approach also provides obstacle avoidance capability by modulating the velocity reference input from user limiting the velocity of the slave robotic vehicles. Analysis provided shows that the proposed approach offers guaranteed obstacle avoidance performance under some unrestrictive conditions. Further developments of this approach were also presented that improve the operator's perception of the environment and deal with moving obstacles in the environment. Simulation and experiments on an aerial robotic platform were conducted. The results provided demonstrate the effectiveness of the dynamic kinesthetic boundary approach. This thesis presents the novel energy consistent models of VTOL aerial robotic vehicle teleoperation systems using the port Hamiltonian modeling framework and bond graph modeling. The proposed framework was developed for both admittance and impedance master joystick configurations. In addition to the classical master joystick devices, a haptic trackball master device was built. This device allows direct position-to-position control of mobile robotic vehicles by mapping roll and pitch to planar displacement of the mobile vehicle improving the operators' perception of the vehicle's motion. Such a device can be directly integrated into the port hamiltonian modeling framework proposed in the natural manner. Experiment on the robotic platform with the interconnected admittance haptic joystick and haptic trackball was conducted. The experimental result demonstrated the stable interaction between operator and the teleoperation system under the proposed system framework.