LAUSR

Abstract This paper proposed a visual servo framework for the tower-crane-like Gravity Compensation System that is widely used in various aerospace ground tests. In this framework, a real-time detection algorithm is proposed to measure the position of the target. Using this algorithm, the target's original position is marked firstly in a well-designed calibration step, and then compared to its real-time position to calculate the offset for tracking. To realize the high precision tracking, a 2-level servo mechanism is investigated, in which a large-inertia crane servo system is utilized for the rough tracking and a 2-DOF (2 Direction Of Freedom) platform is used for the precise localization. The proposed method leads to a high responding speed and low time-drift in wide range tracking, and greatly decreases the compensation error in the microgravity simulation for space manipulator and moon vehicle.