Vibration Control of an Experimental Flexible Manipulator Against Input Saturation

Zhijia Zhao, Sentao Cai, Ge Ma, F. Richard YuIEEE CAA J. Autom. Sinica 2023 · DOI

Abstract

Flexible robotic manipulators offer significant advantages in energy efficiency and weight for aerospace and medical applications, yet they present complex control challenges due to distributed elasticity and physical motor limits. This paper details a novel boundary control strategy that integrates backstepping technology with a Nussbaum function and smooth saturation approximation. The proposed approach effectively manages input saturation while suppressing unwanted vibrations and ensuring precise positioning. Experimental results on a rotary flexible link platform demonstrate that the system achieves faster stabilization and smoother control signals compared to traditional methods, successfully eliminating the hardware-damaging "chattering" effect.

Introduction

The adoption of flexible manipulators is often hindered by two primary technical hurdles: the complexity of modeling distributed parameters (where mass and elasticity are spread throughout the arm) and the physical reality of input saturation. When a controller demands torque exceeding a motor's capacity, standard anti-saturation techniques often rely on discontinuous functions that cause high-frequency "chattering." This phenomenon not only degrades precision but also risks mechanical failure over time. This research addresses these issues by developing a unified control law that maintains performance even when the system reaches its physical limits.

Methodology

The researchers developed a boundary control scheme built upon three core pillars:

  • Backstepping Technology: A recursive design process that decomposes the non-linear system into manageable sub-systems.
  • Smooth Saturation Approximation: Instead of a "hard" limit, a Hyperbolic Tangent Function (HTF) is employed to create a smooth transition at the motor's physical boundaries, preventing abrupt signal changes.
  • Nussbaum Function: This mathematical tool handles the unknown control gain near saturation limits, allowing the controller to adapt seamlessly without requiring a hard switch.

The stability of this integrated system was rigorously verified using Lyapunov stability analysis, ensuring that all system signals remain uniformly bounded during operation.

Experimental Setup

The proposed control strategy was validated using a Quanser rotary flexible link experimental platform. The hardware configuration included a lightweight stainless steel link equipped with foil strain gauges to monitor bending vibrations. Real-time control and data acquisition were facilitated via a DC motor unit, power amplifier, and a dedicated PC interface.

Results and Discussion

The performance of the proposed controller was compared against basic boundary control and an auxiliary-system-based anti-saturation method.

As illustrated in Figure 1, the proposed method achieved superior position tracking. The manipulator reached its target 30-degree angle in approximately 0.28 seconds (measured from the step onset at 0.2 seconds) and achieved full stabilization by 0.85 seconds.

Comparison of position tracking performance for different control strategies.

Vibration suppression was equally effective. As shown in Figure 2, the link's tip deflection reached an initial peak of approximately 0.054m but was suppressed to within 0.012m within just 0.4 seconds of the command onset.

Vibration suppression performance at the link tip over time.

Crucially, the control input remained smooth throughout the operation. By replacing standard sign functions with the Nussbaum-HTF combination, the researchers eliminated the high-frequency chattering typically observed in anti-saturation controllers, thereby protecting the experimental hardware from excessive wear.

Conclusion

This research demonstrates a robust solution for controlling flexible manipulators under physical constraints. By smoothing the saturation response and utilizing adaptive mathematical tools, the system achieves high-performance positioning and vibration damping without compromising hardware longevity. Future work will focus on extending this methodology to more complex multi-link robotic systems and optimizing control gains for broader industrial applications.

Summary updated

BibTeX
@article{zhao2023vibration,
  author = {Zhijia Zhao and Sentao Cai and Ge Ma and F. Richard Yu},
  title = {{Vibration Control of an Experimental Flexible Manipulator Against Input Saturation}},
  journal = {IEEE CAA J. Autom. Sinica},
  year = {2023},
  doi = {10.1109/jas.2023.123345},
}
Figure