Publication:
Modeling of nonlinear ferroelectric microactuator with Fourier Neural Operator
Date
2026
Proceedings Paper
Loading...
Author(s)
Journal
2026 27TH INTERNATIONAL CONFERENCE ON THERMAL, MECHANICAL AND MULTI-PHYSICS SIMULATION AND EXPERIMENTS IN MICROELECTRONICS AND MICROSYSTEMS, EUROSIME
Abstract
While offering excellent electromechanical coupling properties, PZT material exhibits highly nonlinear and hysteresis behaviours at large applied electric held, limiting its usage in certain applications. For example, whereas PZT-based ultrasound transducers operate at their resonance frequency within the linear regime using small signals, actuators functioning in the quasi-static regime with large signals are subject to challenges associated with hysteresis compensation. In this contribution, we present a novel approach of neural operator learning for the characterization of ferroelectric micro-actuator under applied large signal and high frequencies. Based upon the LDV measurement where the applied signals at various amplitudes and frequencies and the corresponding deformation as labelled training data, a Fourier Neural Operator (FNO) is trained. The proposed approach can predict the complex deformation of ferroelectric microactuator beyond the capacity of classical model. In addition, the success of neural-operator-based method can extend and exploit further the functionality of the ferroelectric actuator near resonance frequency.