Locating Screw Grooves in Non-Visual Environments Based on Force Feedback Control

To address the challenge of quickly and accurately locating screw grooves in visionless environments using a robotic arm, this paper proposes a low-cost improved scanning-based search method. The method eliminates the influence of force sensor bias and load to obtain precise environmental force data. Subsequently, Fourier Transform is employed to convert time-domain signals into the frequency domain, identifying and recording specific frequency changes corresponding to contact with the screw groove. The torque variation signals generated when the screwdriver contacts the groove are primarily determined by the physical characteristics of the groove, such as its depth and width. This specific frequency can be used as a reference standard to rapidly locate grooves of screws with identical specifications. To validate the effectiveness of the proposed scanning-based search method, repeated simulations were conducted. The simulation results demonstrate that once the dominant frequency is stabilized through multiple scans, it is no longer necessary to make incremental adjustments or validations, enabling rapid detection and precise localization by directly utilizing the dominant frequency.