• CN: 11-2187/TH
  • ISSN: 0577-6686

Journal of Mechanical Engineering ›› 2026, Vol. 62 ›› Issue (12): 22-32.doi: 10.3901/JME.260479

Previous Articles    

Accelerated Degradation Modeling and Reliability Estimation for Harmonic Drives in Industrial Robots Considering Periodic Shocks

WANG Jia1,2,3, YIN Huiqiang1,2, WANG Chongshuai1,2, HAN Xu1,2,3   

  1. 1. State Key Laboratory of Smart Power Distribution Equipment and System, Hebei University of Technology, Tianjin 300401;
    2. School of Mechanical Engineering, Hebei University of Technology, Tianjin 300401;
    3. School of Electrical Engineering, Hebei University of Technology, Tianjin 300401
  • Received:2025-07-25 Revised:2026-03-02 Published:2026-08-03

Abstract: The accuracy of reliability assessment for harmonic drives in industrial robots depends on the extent to which accelerated degradation models (ADMs) can reflect actual service conditions and on the validity of accelerated degradation testing methods. However, existing ADMs and test methods often fail to adequately characterize the periodic shock loads experienced by harmonic drives under complex operating conditions, such as variable speed and load, forward-reverse switching, and abrupt starts and stops. As a result, the degradation behavior of harmonic drives is difficult to characterize accurately, which undermines the credibility of reliability assessment results. To address this issue, this study develops a novel ADM for harmonic drives that explicitly incorporates periodic shock loads. Periodic step shocks are introduced into a Wiener process to capture abrupt performance changes induced by shocks during operating-condition transitions. In addition, a shock influence factor is incorporated into a power-law acceleration model to establish the relationship between shock loads and degradation rate, and the resulting ADM is formulated based on the acceleration factor constant principle. To ensure model accuracy, a hybrid parameter estimation approach combining Latin hypercube sampling and a genetic algorithm is employed for high-dimensional nonlinear estimation, and the 95% confidence intervals of the model parameters are obtained. Furthermore, a service load spectrum-driven accelerated degradation testing method is proposed based on measured load spectrum data from industrial robots. Following the principle of equivalent torque consistency, a multi-level accelerated loading spectrum is designed to simulate variable-speed, variable-load, and shock conditions. Accelerated degradation tests are then conducted on three harmonic drives to obtain transmission error degradation data for reliability assessment. The results show that the proposed model provides higher reliability assessment accuracy than existing models and enables more accurate prediction of the service life of harmonic drives.

Key words: accelerated degradation test, industrial robot, accelerated degradation model, reliability estimation, harmonic drive

CLC Number: