Electromechanical coupling modeling and motor current signature analysis of bolt loosening of industrial robot joint

Since loosening of the bolt connection reduce the stability and working accuracy of industrial robot joint, developing and applying new detection methods to provide effective and reliable diagnosis is still a challenging task. In this study, a detection method for bolt loosening of in-dustrial robot joint based on electromechanical modeling and motor current signature analysis (MCSA) is proposed. Firstly, a dynamics model is established based on the electromechanical coupling characteristics of the robot joint servo system, and the bolt loosening factor in the model is equivalent to the variations of support stiffness. The vibration performance of the system due to bolt loosening is analyzed, and the dynamical equations show the coupling relationship between the motor current and bolt loosening. Then according to the traits of motor current variation, this study proposes to utilize the time-frequency features of the motor current to detect loosening. The energy concentrated time-frequency representation (TFR) is obtained by the syn-chrosqueezing transform (SST) method, and then time-frequency ridges are extracted. The fea-tures of the time-frequency ridges are adopted as indicators of the bolt loosening. The support vector machine (SVM) classifier is utilized to identify the loosening of bolts. The simulated motor current signal analysis shows that the SST can highlight the current transients caused by bolt loosening. The bolt loosening experiment is conducted on the single-degree-of-freedom servo joint test bench, and the results prove the effectiveness of the proposed method.

Automated summary

In this study, a detection method for bolt loosening of industrial robot joint based on electromechanical modeling and motor current signature analysis (MCSA) is proposed. The method utilizes the time-frequency features of the motor current to detect loosening and achieves effective results.