We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
基于 PINN 的变截面压电半导体纤维力学特性研究.
- Authors
吴文锐; 房 凯; 李 鹏; 钱征华
- Abstract
In order to investigate the mechanical properties of non-uniform piezoelectric semiconductor (PS) fiber, the physics-informed neural network model (PINN) is proposed in this paper, and a deep learning algorithm is applied to solve the partial differential equations with variable coefficients. Taking the static extension of PS fiber with variable cross-section as an example, a deep neural network is firstly established as a trial function, and then substituted into the governing equations of PS to form a residual and used as the weighted loss function for the machine learning. Then, the numerical solution is approximated through the deep machined learning. The research results indicate that the presented method has wide applicability, and can be used to solve the linear and non-linear governing equations of PS materials with arbitrary cross-section.
- Subjects
SEMICONDUCTOR junctions; PIEZOELECTRIC materials; ELECTRIC potential; DEEP learning
- Publication
Piezoelectrics & Acoustooptics, 2023, Vol 45, Issue 5, p686
- ISSN
1004-2474
- Publication type
Article
- DOI
10.11977/j.issn.1004-2474.2023.05.007