Teleoperation Control Of Lunar Rover Based On Augmented Reality

In this paper, we took BH2 lunar rover prototype in Beijing University as the object of study, with the control technology for lunar vehicle under the condition of large time delay as the focus. In order to weaken the impact of communication delays on the operability and stability of the control system, we proposed a comprehensive use of forecasts and bilateral passive control, and did some in-depth investigations on the development of teleoperation environment based on augmented reality and passive control technology. Key results are as follow,Firstly, a novel teleoperation framework based on forecast and passive control was designed for lunar rover. By establishing an augmented reality environment consistency with the lunar environment, a closed loop control circuit on the ground station was formed, the operator got real-time visual feedback by controlling the virtual lunar rover in the augmented reality environment, which can avoid time delays in the communication channel and enhance the operability of the control system; By setting controllers on each side of the teleoperation system, passive control achieved stability of the control system and the coordination of speed and position.Secondly, an augmented reality teleoperation environment was established. Accurate modeling for the lunar terrain was achieved with self-made laser scanning system and ODE was used to simulate the microgravity environment on the lunar surface, which ensures the consistency of the augmented reality environment with the lunar surface.Thirdly, the dynamics models of phantom omni and lunar rover were derivated according to Lagrange method, on which passive control was discussed to ensure the effective control of lunar rover.Finally, simulation system based on forecast and passive control for teleoperation control of lunar rover was established. Queue buffer was used to simulate the communication delay between earth and lunar. The performance of the framework was studied with the simulation system under 10s time delay. Experimental results demonstrate the feasibility.