Accurate physical modeling and synchronization control of dual-linear-motor-driven gantry with dynamic load

Chao Li, Zheng Chen, Can Yang, Bin Yao, Shiliang PuResearch 2020 · DOI

TL;DR

This paper introduces a high-precision control strategy for dual-linear-motor-driven (DLMD) gantries by treating the crossbeam's parasitic rotation as a controllable state. By developing a 3-degree-of-freedom (3DOF) model that accounts for shifting payloads, the researchers developed an adaptive robust controller that significantly reduces internal mechanical stress and improves tracking accuracy compared to industry-standard methods.

What Problem Does This Paper Address?

In industrial manufacturing, H-type gantries are workhorses. They typically use two parallel linear motors to drive a massive crossbeam (the Y-axis) and a third motor to move a payload along that beam (the X-axis). The "holy grail" here is perfect synchronization: both Y-axis motors must move in unison.

If they don't, the crossbeam twists. This rotation, even if microscopic, creates massive internal forces due to the high stiffness of the guide rails. This leads to accelerated wear, vibration, and poor accuracy. The problem is exacerbated by "dynamic loads"—as the payload moves along the X-axis, the center of mass and the inertia of the Y-axis system shift constantly. Most existing controllers either ignore this rotation or use "thrust allocation" schemes that are too slow to handle these fast-changing dynamics.

Methods

The authors tackle this with a two-pronged approach:

  1. Rigid-Flexible Modeling: They developed a 3DOF mathematical model using Lagrangian mechanics. Unlike traditional models that only look at X and Y positions, this one includes the rotational angle ($\alpha$) of the crossbeam. Crucially, they modeled the bearing balls in the guide rails as high-stiffness horizontal springs, which allows the model to "see" the internal forces generated by misalignment.
  2. TITO Synchronization Control: They proposed a Three-Input Three-Output (TITO) control structure. Instead of just commanding X and Y positions, the controller actively commands the suppression of the rotation angle $\alpha$.
  3. DCDIARC Algorithm: The core "brain" is a Desired Compensation Integrated Direct/Indirect Adaptive Robust Control (DCDIARC) law.
  4. Adaptive: It learns unknown system parameters (like friction and exact mass) online.
  5. Robust: It uses high-gain feedback to handle "unlearnable" disturbances and modeling errors.
  6. Desired Compensation: It uses reference trajectories rather than noisy sensor data for model compensation, leading to smoother control signals.

Datasets

The study did not use public datasets. Instead, the researchers evaluated their algorithms on a custom-built DLMD gantry at Zhejiang University. The setup featured:

  • Linear encoders with 0.5 $\mu$m resolution.
  • A dSPACE controller board for real-time execution (0.1 ms sampling period).
  • Two experimental sets: Set1 (normal conditions) and Set2 (adding a 40kg payload to test robustness).

Key Insights and Ideas

  • Rotation is a State, Not Just Noise: The most significant "aha moment" is that the crossbeam's rotation shouldn't just be minimized via mechanical stiffness; it should be actively controlled as a dynamic state. By "actively suppressing" the rotation, you naturally regulate the internal forces.
  • The Coupling is Complex: The paper mathematically demonstrates how the X-axis position directly alters the Y-axis inertia and the "active length" of the moment arm. This explains why static synchronization often fails during complex XY maneuvers.
  • Parameter Convergence: The use of a recursive least squares (RLS) estimator with "exponential forgetting" allows the controller to quickly adapt when a heavy payload is added, effectively "re-learning" the machine's physics on the fly.

Key Results

The researchers compared their proposed controller (C1) against static thrust allocation (C2) and dynamic thrust allocation (C3).

  • Synchronization Accuracy: Under dynamic load, the proposed C1 controller limited rotational oscillation to just 2.76 $\mu$rad, compared to a massive 34.05 $\mu$rad for the static C2 method.
  • Tracking Performance: C1 achieved a final tracking accuracy (RMS error) of 7.46 $\mu$m, significantly outperforming the standard methods.
  • Efficiency: The proposed method required the smallest control effort ($L_2$ norm of 2.102V), proving that active control is more efficient than "fighting" the hardware's natural stiffness.
  • Robustness: When a 40kg payload was added (Set2), the adaptive law converged quickly, maintaining high performance despite the drastic change in system mass.

Limitations and Future Work

While the results are impressive, the paper focuses on a specific H-gantry geometry. The authors note that:

  • Model Complexity: The 3DOF model assumes the crossbeam is a rigid body. In extremely large-scale gantries, structural flexibility (bending) of the beam itself might become a factor that this model doesn't yet account for.
  • Sensor Requirements: The method relies on high-resolution linear encoders on both sides of the gantry to derive the rotation angle. Systems with fewer sensors would need a different estimation strategy.
  • Future Directions: The authors suggest that the DCDIARC framework could be extended to handle actuator saturation or more complex multi-axis configurations.

Summary updated

Abstract

Abstract To achieve high-accuracy tracking of dual-linear-motor-driven (DLMD) gantry, high-level synchronization between redundant actuators is a nonnegligible factor and also a difficult issue to be solved prior. Especially, when both XY axes are simultaneously operating to accomplish complex tasks efficiently, additional coupling effects will be generated by the dynamic load presented on the crossbeam, which makes the synchronization issue more complicated compared to the case with static load. However, due to the absence of an accurate model to fully reveal the complete coupling characteristics, existing approaches to this issue still have inherent limitations. Therefore, this paper focuses on the systematic physical modeling and synchronization control of DLMD gantry with a dynamic load presented on the crossbeam. A complete coupling mathematical model is established firstly, by fully considering two linear motions (X-axis and Y-axis) and also including the additional rotational motion of the crossbeam. Built upon the effective model information, corresponding solutions by compensating the dynamic load effects and actively controlling the rotational dynamic to regulate the internal forces have been proposed, leading to a novel adaptive robust synchronization control method. Comparative experiments are carried out, and the results show the effectiveness and superiority of the proposed method in dealing with synchronization issue subjected to dynamic load effects.
BibTeX
@inproceedings{li2020accurate,
  author = {Chao Li and Zheng Chen and Can Yang and Bin Yao and Shiliang Pu},
  title = {{Accurate physical modeling and synchronization control of dual-linear-motor-driven gantry with dynamic load}},
  booktitle = {Research},
  year = {2020},
  doi = {10.21203/rs.3.rs-23857/v1},
}

Loading related content...

Figure