Transfer Learning-Based Coupling of Smoothed Finite Element Method and Physics-Informed Neural Network for Solving Elastoplastic Inverse Problems
نویسندگان
چکیده
In practical engineering applications, there is a high demand for inverting parameters various materials, and obtaining monitoring data can be costly. Traditional inverse methods often involve tedious computational processes, require significant effort, exhibit slow convergence speeds. The recently proposed Physics-Informed Neural Network (PINN) has shown great potential in solving problems. Therefore, this paper, we propose transfer learning-based coupling of the Smoothed Finite Element Method (S-FEM) PINN inversion elastic-plasticity aim to improve accuracy efficiency parameter different elastic-plastic materials with limited data. High-quality small datasets were synthesized using S-FEM subsequently combined pre-training purposes. pre-trained model saved used as initial state new material parameters. performance compared conventional (FEM) on set. Additionally, both non-transfer results show that: (1) our method performs well datasets, an error essentially less than 2%; (2) approach outperforms FEM terms efficiency; (3) at least twice efficient without learning, while still maintaining accuracy. Our well-suited only datasets. use learning greatly improves efficiency, making accurate solution reducing cost complexity applications.
منابع مشابه
Evaluation of Fracture Parameters by Coupling the Edge-Based Smoothed Finite Element Method and the Scaled Boundary Finite Element Method
This paper presents a technique to evaluate the fracture parameters by combining the edge based smoothed finite element method (ESFEM) and the scaled boundary finite element method (SBFEM). A semi-analytical solution is sought in the region close to the vicinity of the crack tip using the SBFEM, whilst, the ESFEM is used for the rest of the domain. As both methods satisfy the partition of unity...
متن کاملB-Spline Finite Element Method for Solving Linear System of Second-Order Boundary Value Problems
In this paper, we solve a linear system of second-order boundary value problems by using the quadratic B-spline nite el- ement method (FEM). The performance of the method is tested on one model problem. Comparisons are made with both the analyti- cal solution and some recent results.The obtained numerical results show that the method is ecient.
متن کاملSelective Smoothed Finite Element Method
The paper examines three selective schemes for the smoothed finite element method (SFEM) which was formulated by incorporating a cell-wise strain smoothing operation into the standard compatible finite element method (FEM). These selective SFEM schemes were formulated based on three selective integration FEM schemes with similar properties found between the number of smoothing cells in the SFEM...
متن کاملA Neural Network Method Based on Mittag-Leffler Function for Solving a Class of Fractional Optimal Control Problems
In this paper, a computational intelligence method is used for the solution of fractional optimal control problems (FOCP)'s with equality and inequality constraints. According to the Ponteryagin minimum principle (PMP) for FOCP with fractional derivative in the Riemann- Liouville sense and by constructing a suitable error function, we define an unconstrained minimization problem. In the optimiz...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Mathematics
سال: 2023
ISSN: ['2227-7390']
DOI: https://doi.org/10.3390/math11112529