Research on Gear Fault Detection of Planetary Reducers for Construction Robots Based on an Adaptive Observer

Abstract

At present, construction robots have received extensive attention in construction scenarios such as wall masonry and component handling. The operational reliability of their joint transmission systems directly affects the working accuracy and construction safety of robots. To address the problem that planetary reducers in construction robots are prone to faults under complex working conditions, this paper proposes a fault-detection method based on an adaptive observer. First, a state–space model containing unknown disturbances and actuator faults is established by combining the transmission structure and typical fault mechanism of the planetary reducer. On this basis, an adaptive observer with more degrees of freedom is designed, and the L∞ disturbance-robustness index is introduced. The augmented-output function is used to apply the H− performance index to enhance fault sensitivity. The simulation results show that when the system is fault-free, the residual remains below the threshold, and no false alarm occurs. After a time-varying crack fault is introduced, the residual can quickly exceed the threshold and trigger a complete fault alarm. Compared with the L∞ observer, the introduced H− performance index significantly improves the fault sensitivity of the observer. Compared with the traditional Luenberger observer, the proposed method has better disturbance robustness and fault sensitivity, providing an effective method for early fault detection of planetary reducers in construction robots.