Automated Analysis of Meshing Performance of Harmonic Drive Gears Under Various Operating Conditions

Accurate and efficient analysis of the meshing performance of harmonic drives (HDs) is the core and challenging problem for the improvement of the positioning accuracy of robots. This paper presents an integrated hardware and software system for an automated analysis of the meshing performance of the HD under various operating conditions. First, an optical measurement method based on the principle of periscope is proposed to capture the meshing images at a different driving speed and load torque. Next, a position tracking algorithm based on image processing and pattern recognition is designed to obtain the positions of the flexspline and circular spline teeth in the meshing process. Furthermore, a mathematical model based on the geometric relationship between the flexspline and the circular spline teeth is established to calculate the meshing backlash and depth. Last, the performance and advantages of the proposed method are demonstrated through uncertainty analysis and comparing with the conventional method. The results show that the error of the calculated meshing parameters is close to 0, and 1.59 s are required to track the FS teeth in each video frame with a laptop of 2.6-GHz CPU and 12-GB RAM. Moreover, the total rotation angle of the wave generator and the backlash and depth in the meshing process are seriously affected by the driving speed and load torque. Our method offers an effective tool for the HD manufacturers to examine the quality of products, and the obtained results can help to optimize the tooth profile of the HD.