Modeling of a 2-DOF Piezoelectric Micromanipulator at High Frequency Rates through Nonlinear Black-box System Identification

Ayala, Helon Vicente Hultmann; Rakotondrahe, Micky; Coelho, Leandro dos Santos

doi:10.23919/acc.2018.8431386

Cited by 6 publications

(4 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can thus be seen that the hysteretic behavior has been adequately captured. This reinforces the great opportunity in the application of data-driven modeling tools in order to design the mathematical abstraction for systems with higher complexity as demonstrated in a recent work which investigated 2-DOF piezoelectric micromanipulators [54] with ad-hoc model structures for NN. Thus, machine learning-based feedback control such as [55] can be directly applied on the basis of the accurate models constructed with less intervention from the engineer, as we have demonstrated.…”

Section: Case Studysupporting

confidence: 59%

Nonlinear black-box system identification through coevolutionary algorithms and radial basis function artificial neural networks

Ayala

Habineza²,

Rakotondrabe

et al. 2020

Applied Soft Computing

Self Cite

View full text Add to dashboard Cite

The present work deals with the application of coevolutionary algorithms and artificial neural networks to perform input selection and related parameter estimation for nonlinear black-box models in system identification. In order to decouple the resolution of the input selection and parameter estimation, we propose a problem decomposition formulation and solve it by a coevolutionary algorithm strategy. The novel methodology is successfully applied to identify a magnetorheological damper, a continuous polymerization reactor and a piezoelectric robotic micromanipulator. The results show that the method provides valid models in terms of accuracy and statistical properties. The main advantage of the method is the joint input and parameter estimation, towards automating a tedious and error prone procedure with global optimization algorithms.

show abstract

Section: Case Studysupporting

confidence: 59%

Nonlinear black-box system identification through coevolutionary algorithms and radial basis function artificial neural networks

Ayala

Habineza²,

Rakotondrabe

et al. 2020

Applied Soft Computing

Self Cite

View full text Add to dashboard Cite

show abstract

“…Two degrees of freedom (dof) piezoelectric actuators (PAs) were modeled through Duhem hysteresis model with parameters identification made by artificial neural network (ANN) in Wang et al (Wang and Chen, 2017). Ayala et al (Ayala et al, 2018) used a shallow ANN at high frequency rates in a nonlinear autoregressive exogenous input (NARX) model for the same application. A nanoscale PA was identified through a recurrent neural network (RNN) in order to implement a model predictive control with real-time trajectory tracking in Xie et al (Xie and Ren, 2019).…”

Section: Introductionmentioning

confidence: 99%

Comparison of Deep Learning Architectures for Nonlinear System Identification of a Hysteretic Piezoelectric Precise Positioner

Ribeiro

Domingues

Reynoso-Meza

et al. 2021

Procedings Do XV Simpósio Brasileiro De Automação Inteligente

View full text Add to dashboard Cite

The characterization of hysteretic components poses a difficult nonlinear system identification problem. Several studies have addressed this by employing artificial neural networks, where deep learning (DL) has recently gained attention in system identification tasks. However, there is a lack of studies comparing different deep neural network (DNN) architectures. Therefore, this work proposes the comparison of three DNN architectures, including feedforward neural networks (FFNN), long short term memory (LSTM), and convolutional neural networks (CNN), for the characterization of a piezoelectric positioning system (positioner) typified by hysteresis. Moreover, Bayesian optimization is employed for hyperparameter tuning in all DNN architectures. Results show that all DL architectures achieved desirable values for the coefficient of determination (R 2 ) and root mean squared error (RMSE). However, LSTM obtains the best overall results, outperforming both the FFNN and CNN, being a more appropriate black-box architecture for identifying frequency-dependent hysteresis loop shapes.

show abstract

“…For this purpose, in (Noël, Esfahani, Kerschen & Schoukens, 2017) the authors uses a black-box based in Bouc-Wen equations to model the hysteresis while in (Habineza, Rakotondrabe & Le Gorrec, 2015), (Aljanaideh & Rakotondrabe, 2018) and (Rakotondrabe, 2017) the authors use models of 2-DoF with different approaches for modelling and control of a multivariable hysteresis in piezoelectric systems to get a feedback. A nonlinear black-box system identification for modelling the 2-DoF micromanipulation problems is discussed in (Ayala, Rakotondrahe & Coelho, 2018).…”

Section: Literature Reviewmentioning

confidence: 99%

“…To evaluate the quality of the model was used the free-run simulation in which the sample is used only to define the initial conditions of the model and the model itself at each step is based on the previously calculated step. The number of samples is equal the number of regressors chosen (Ayala et al, 2018):…”

Section: Model Validationmentioning

confidence: 99%

Application of the Keras Framework to Nonlinear Black-Box System Identification of Piezoelectric Micromanipulators

MEDEIROS¹

View full text Add to dashboard Cite

Agradeço a PUC-Rio, que me deu oportunidade para expandir meus horizontes e ao meu orientador Helon Ayala pela oportunidade e suporte neste projeto. Agradeço aos meus pais Rosélia e Alberto e meus avós Marlene e Adalberto por todo o suporte que me deram durante esta jornada. Agradeço a minha namorada Thallita, pelo suporte e pela compreensão nos momentos em que estive distante. E para todos aqueles que estiveram envolvidos direta ou indiretamente em minha graduação, meu muito obrigado.

show abstract

Modeling of a 2-DOF Piezoelectric Micromanipulator at High Frequency Rates through Nonlinear Black-box System Identification

Cited by 6 publications

References 39 publications

Nonlinear black-box system identification through coevolutionary algorithms and radial basis function artificial neural networks

Nonlinear black-box system identification through coevolutionary algorithms and radial basis function artificial neural networks

Comparison of Deep Learning Architectures for Nonlinear System Identification of a Hysteretic Piezoelectric Precise Positioner

Application of the Keras Framework to Nonlinear Black-Box System Identification of Piezoelectric Micromanipulators

Contact Info

Product

Resources

About