H-Shaped Multiple Linear Motor Drive Platform Control System Design Based on an Inverse System Method

Caiyan Qin, Chaoning Zhang, Haiyan LuEnergies 2017 · DOI

TL;DR

This paper presents a novel control strategy for H-shaped platforms driven by multiple linear motors, addressing challenges in synchronous control and overall positioning accuracy. By employing an inverse system method for dynamic decoupling, the authors transform the complex, coupled system into a simpler, pseudo-linear one, enabling precise trajectory tracking and high synchronization performance suitable for demanding industrial applications.

What Problem Does This Paper Address?

H-shaped platforms, widely used in precision manufacturing and semiconductor packaging, are typically driven by multiple linear motors. A key challenge in these systems is maintaining precise synchronization between motors operating along the same axis (e.g., the two motors in the Y-direction) and minimizing the combined positioning error across all axes (X and Y). Traditional control methods often treat these axes independently, leading to suboptimal overall accuracy. Furthermore, the inherent mechanical coupling and nonlinear dynamics of these multi-motor systems make achieving high-precision, real-time control difficult.

Methods

The core of this paper's contribution lies in the application of the inverse system method for dynamic decoupling. The authors first derive the mechanical dynamics of the H-shaped platform, accounting for the coupling between the motors. They then develop a mathematical model for the permanent magnet synchronous linear motors, including coordinate transformations (ABC to d-q).

The inverse system method is then used to transform the original nonlinear, coupled system into a pseudo-linear, decoupled system. This is achieved by designing an inverse system model that effectively compensates for the nonlinearities and cross-couplings. Once the system is linearized and decoupled, traditional linear control techniques, specifically PID controllers, can be effectively applied. This approach integrates the control of all three motors into a single, unified system, rather than controlling axes independently. State feedback is used to achieve this decoupling.

Datasets

Not discussed in the paper. The authors used simulations in MATLAB/Simulink and real-time testing on an NI controller.

Key Insights and Ideas

  • Unified Control for Coupled Systems: The most significant insight is the successful application of the inverse system method to treat the H-shaped platform as a single, integrated system. This contrasts with traditional approaches that handle axes independently, leading to better overall performance.
  • Pseudo-Linearization for Simplicity: Transforming a complex nonlinear, coupled system into a pseudo-linear one is a powerful technique. It allows the use of well-understood and simpler linear controllers (like PID) while achieving high performance.
  • Dynamic Decoupling is Crucial: The paper highlights that dynamic decoupling is essential for managing the inherent cross-couplings in multi-motor systems, directly impacting both synchronization and trajectory accuracy.
  • Real-Time Viability: The demonstration of the control strategy's real-time performance on an industrial controller, with computation times well within typical requirements, is a critical validation for practical application.

Key Results

  • High Synchronization Performance: The proposed control strategy significantly reduces synchronous deviation between the Y-axis motors. Under a disturbance of half the rated load force, the maximal synchronous deviation was kept within 0.1 µm with a recovery time of 0.02 s.
  • Reduced Total Trajectory Error: Simulations show that the proposed method achieves a smaller total trajectory error compared to traditional independent PID control, especially under sudden load disturbances.
  • Improved Circle Trajectory Tracking: When tracking a circular path, the proposed model exhibited better performance, attributed to less coupling between the X and Y axes.
  • Efficient Real-Time Performance: The control system's average loop computation time was measured at approximately 180-200 µs, which is significantly faster than the 1 ms requirement for many high-precision industrial applications.

Limitations and Future Work

  • Assumptions in Model Simplification: The mechanical analysis simplifies the rotation angle $\delta$ to be very small (less than 0.1°). While this is realistic for high-precision applications, it's a simplification.
  • Inverse System Condition: The paper mentions that not all systems are invertible and that certain conditions must be met for the inverse system to exist. While these conditions are stated, their practical implications for system design beyond this specific platform are not deeply explored.
  • Future Work: The paper implicitly suggests that this method can be extended to other multi-axis, coupled systems requiring high precision and real-time control. Further exploration of adaptive control within this framework could also be a natural extension.

Summary updated

Abstract

Due to its simple mechanical structure and high motion stability, the H-shaped platform has been increasingly widely used in precision measuring, numerical control machining and semiconductor packaging equipment, etc. The H-shaped platform is normally driven by multiple (three) permanent magnet synchronous linear motors. The main challenges for H-shaped platform-control include synchronous control between the two linear motors in the Y direction as well as total positioning error of the platform mover, a combination of position deviation in X and Y directions. To deal with the above challenges, this paper proposes a control strategy based on the inverse system method through state feedback and dynamic decoupling of the thrust force. First, mechanical dynamics equations have been deduced through the analysis of system coupling based on the platform structure. Second, the mathematical model of the linear motors and the relevant coordinate transformation between dq-axis currents and ABC-phase currents are analyzed. Third, after the main concept of inverse system method being explained, the inverse system model of the platform control system has been designed after defining relevant system variables. Inverse system model compensates the original nonlinear coupled system into pseudo-linear decoupled linear system, for which typical linear control methods, like PID, can be adopted to control the system. The simulation model of the control system is built in MATLAB/Simulink and the simulation result shows that the designed control system has both small synchronous deviation and small total trajectory tracking error. Furthermore, the control program has been run on NI controller for both fixed-loop-time and free-loop-time modes, and the test result shows that the average loop computation time needed is rather small, which makes it suitable for real industrial applications. Overall, it proves that the proposed new control strategy can be used in industrial applications that have high-precision and high real-time performance requirements.
BibTeX
@inproceedings{qin2017h,
  author = {Caiyan Qin and Chaoning Zhang and Haiyan Lu},
  title = {{H-Shaped Multiple Linear Motor Drive Platform Control System Design Based on an Inverse System Method}},
  booktitle = {Energies},
  year = {2017},
  doi = {10.3390/en10121990},
}
Figure