Research on Hysteresis of Piezoceramic Actuator Based on the Duhem Model

Miaolei Zhou, Jingyuan WangThe Scientific World JOURNAL 2013 · DOI

TL;DR

This paper tackles the challenge of accurately modeling the hysteresis behavior of piezoceramic actuators, crucial for precise positioning systems. They propose using the Duhem model, approximating its complex functions with polynomials, and compare two parameter identification algorithms: Recursive Least Squares (RLS) and Gradient Correction (GC). The Gradient Correction method emerges as superior, achieving higher modeling accuracy.

What Problem Does This Paper Address?

Piezoceramic actuators are fantastic for micro-displacement tasks due to their high accuracy, strong driving force, and fast response. However, they suffer from inherent hysteresis, nonlinearity, and creep, which degrade repeatability and accuracy. Accurately modeling this hysteresis is key to improving the performance of precision positioning systems that rely on these actuators. While various models exist, the authors focus on the Duhem model, a differential hysteresis model, and aim to improve its practical application by developing robust parameter identification methods.

Methods

The core of the paper involves:

  • Duhem Model Formulation: The authors start with the differential form of the Duhem model: $dw/dt = \alpha [f(v) - w] + dg(v)/dt$.
  • Polynomial Approximation: To handle the piecewise continuous functions $f(v)$ and $g(v)$ within the Duhem model, they approximate them using polynomial functions. This is justified by the Weierstrass approximation theorem, which states that any continuous function on a closed interval can be uniformly approximated by polynomials.
  • Parameter Identification: Two algorithms are employed to identify the model parameters ($\alpha$, and the coefficients of the polynomials for $f(v)$ and $g(v)$):
  • Recursive Least Squares (RLS): A standard adaptive filtering technique for estimating parameters in linear models.
  • Gradient Correction (GC): An iterative optimization method that adjusts parameters based on the gradient of an error function.
  • Model Validation: The identified parameters are used to build a Duhem model of the piezoceramic actuator, and its performance is evaluated by comparing its output to actual experimental data.

Datasets

The paper uses experimental input-output data from a piezoceramic actuator. This data consists of 21 sets of voltage input signals and corresponding displacement output signals, which are used for both training (parameter identification) and evaluation of the models. No new benchmark datasets are introduced.

Key Insights and Ideas

  • Polynomials as a Practical Tool: Approximating the complex $f(v)$ and $g(v)$ functions of the Duhem model with polynomials makes the model more amenable to standard parameter identification techniques.
  • Gradient Correction Outperforms RLS: The Gradient Correction algorithm demonstrates superior modeling accuracy compared to the Recursive Least Squares algorithm for this specific application. This suggests that the iterative nature and error-driven updates of GC are better suited for capturing the nuances of the Duhem model in piezoceramic actuators.
  • Duhem Model's Suitability: The study validates the Duhem model's ability to represent the hysteresis behavior of piezoceramic actuators, especially when combined with effective parameter identification.

Key Results

  • Modeling Accuracy:
    • Recursive Least Squares (RLS) algorithm achieved a modeling accuracy of 0.24%.
  • Gradient Correction (GC) algorithm achieved a modeling accuracy of 0.11%.
  • Error Metrics:
    • Under RLS, the maximum error was 0.066 µm, with a mean square deviation of 0.0263.
  • Under GC, the maximum error was 0.048 µm, with a mean square deviation of 0.0222.
  • Algorithm Comparison: The Gradient Correction algorithm is highlighted as being simpler, more adaptable, and easier to implement, while also providing better modeling accuracy.

Limitations and Future Work

  • Limitations:
    • The paper focuses on a specific type of piezoceramic actuator and its hysteresis behavior. The generalizability of the findings to other actuator types or under different operating conditions might be limited.
  • While the polynomial approximation is effective, the choice of polynomial order could impact accuracy and computational cost.
  • Future Work:
    • The authors suggest that the precise modeling achieved with the Gradient Correction algorithm lays a foundation for further control research of piezoceramic actuators. This implies future work could involve developing advanced control strategies that leverage this accurate hysteresis model.
  • Exploring other advanced modeling techniques or parameter identification algorithms could potentially yield even better results.

Summary updated

Abstract

To improve the modeling accuracy of piezoceramic actuator in the precision positioning system, the Duhem hysteretic model of the piezoceramic actuator was proposed. The paper used the polynomial function to approach the piecewise continuous function and f(v) and g(v) in the Duhem model, adopted recursive least squares algorithm and gradient correction algorithm to identify parameter α , polynomial coefficients of f and g in the Duhem model, and established the nonlinear parametric model of the piezoceramic actuator. Contrasting the simulation results of recursive least squares algorithm and gradient correction algorithm, the modeling accuracy is 0.24% when adopting the recursive least squares algorithm, and the modeling accuracy is 0.11% when adopting the gradient correction method. The result showed that the gradient correction algorithm could meet the modeling accuracy better, and the structure of the algorithm is simple, adaptable, and easy to implement.
BibTeX
@article{zhou2013research,
  author = {Miaolei Zhou and Jingyuan Wang},
  title = {{Research on Hysteresis of Piezoceramic Actuator Based on the Duhem Model}},
  journal = {The Scientific World JOURNAL},
  year = {2013},
  doi = {10.1155/2013/814919},
}
Figure