A Robust Resonance Suppression Method for Linear Tooth Belt Drive With Variable Resonance Frequency

To suppress mechanical resonance, notch filters and observer-based state feedback strategies are widely used. However, their performance is undermined by the variable resonance dynamics due to the flexible belt in linear tooth belt drive. To this end, we separate the model of the linear tooth belt drive into nonresonance and pure-resonance parts to identify the root cause of the resonance as the undamped feature of the latter part. The model separation decouples the model into deterministic and uncertain parts, which simplifies the controller design. Further, we estimate the state of the pure-resonance part by a simple reduced-order observer so as to adjust the system damping. The observer is constructed based on an equivalent transformation of the feedback channel rather than the second-order resonance model to reduce the order of the observer and make it insensitive to the varying parameters. Therefore, the proposed method can effectively improve the robustness, regardless of the varying resonance frequency. The simulation and experiment results show that the proposed method outperforms the notch filter method and the active disturbance rejection control for the linear tooth belt drive systems.