Novel Physics-Informed Artificial Neural Network Architectures for System and Input Identification of Structural Dynamics PDEs
Herein, two novel Physics Informed Neural Network (PINN) architectures are proposed for output-only system identification and input estimation of dynamic systems. Using merely sparse output-only measurements, the proposed PINNs architectures furnish a novel approach to input, state, and parameter es...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Buildings |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-5309/13/3/650 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839582871814144000 |
|---|---|
| author | Sarvin Moradi Burak Duran Saeed Eftekhar Azam Massood Mofid |
| author_facet | Sarvin Moradi Burak Duran Saeed Eftekhar Azam Massood Mofid |
| author_sort | Sarvin Moradi |
| collection | DOAJ |
| description | Herein, two novel Physics Informed Neural Network (PINN) architectures are proposed for output-only system identification and input estimation of dynamic systems. Using merely sparse output-only measurements, the proposed PINNs architectures furnish a novel approach to input, state, and parameter estimation of linear and nonlinear systems with multiple degrees of freedom. These architectures are comprised of parallel and sequential PINNs that act upon a set of ordinary differential equations (ODEs) obtained from spatial discretization of the partial differential equation (PDE). The performance of this framework for dynamic system identification and input estimation was ascertained by extensive numerical experiments on linear and nonlinear systems. The advantage of the proposed approach, when compared with system identification, lies in its computational efficiency. When compared with traditional Artificial Neural Networks (ANNs), this approach requires substantially smaller training data and does not suffer from generalizability issues. In this regard, the states, inputs, and parameters of dynamic state-space equations of motion were estimated using simulated experiments with “noisy” data. The proposed framework for PINN showed excellent great generalizability for various types of applications. Furthermore, it was found that the proposed architectures significantly outperformed ANNs in generalizability and estimation accuracy. |
| format | Article |
| id | doaj-art-be4a704e2fa24a05a4e3d41f2022c93c |
| institution | Matheson Library |
| issn | 2075-5309 |
| language | English |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Buildings |
| spelling | doaj-art-be4a704e2fa24a05a4e3d41f2022c93c2025-08-04T02:06:26ZengMDPI AGBuildings2075-53092023-02-0113365010.3390/buildings13030650Novel Physics-Informed Artificial Neural Network Architectures for System and Input Identification of Structural Dynamics PDEsSarvin Moradi0Burak Duran1Saeed Eftekhar Azam2Massood Mofid3Department of Civil Engineering, University of Tehran, Tehran P.O. Box 14155‑6619, IranDepartment of Civil and Environmental Engineering, University of New Hampshire, Durham, NH 03824, USADepartment of Civil and Environmental Engineering, University of New Hampshire, Durham, NH 03824, USADepartment of Civil Engineering, Sharif University of Technology, Tehran P.O. Box 11365-11155, IranHerein, two novel Physics Informed Neural Network (PINN) architectures are proposed for output-only system identification and input estimation of dynamic systems. Using merely sparse output-only measurements, the proposed PINNs architectures furnish a novel approach to input, state, and parameter estimation of linear and nonlinear systems with multiple degrees of freedom. These architectures are comprised of parallel and sequential PINNs that act upon a set of ordinary differential equations (ODEs) obtained from spatial discretization of the partial differential equation (PDE). The performance of this framework for dynamic system identification and input estimation was ascertained by extensive numerical experiments on linear and nonlinear systems. The advantage of the proposed approach, when compared with system identification, lies in its computational efficiency. When compared with traditional Artificial Neural Networks (ANNs), this approach requires substantially smaller training data and does not suffer from generalizability issues. In this regard, the states, inputs, and parameters of dynamic state-space equations of motion were estimated using simulated experiments with “noisy” data. The proposed framework for PINN showed excellent great generalizability for various types of applications. Furthermore, it was found that the proposed architectures significantly outperformed ANNs in generalizability and estimation accuracy.https://www.mdpi.com/2075-5309/13/3/650physics-informed neural networkssystem identificationparameter estimationinput estimation |
| spellingShingle | Sarvin Moradi Burak Duran Saeed Eftekhar Azam Massood Mofid Novel Physics-Informed Artificial Neural Network Architectures for System and Input Identification of Structural Dynamics PDEs Buildings physics-informed neural networks system identification parameter estimation input estimation |
| title | Novel Physics-Informed Artificial Neural Network Architectures for System and Input Identification of Structural Dynamics PDEs |
| title_full | Novel Physics-Informed Artificial Neural Network Architectures for System and Input Identification of Structural Dynamics PDEs |
| title_fullStr | Novel Physics-Informed Artificial Neural Network Architectures for System and Input Identification of Structural Dynamics PDEs |
| title_full_unstemmed | Novel Physics-Informed Artificial Neural Network Architectures for System and Input Identification of Structural Dynamics PDEs |
| title_short | Novel Physics-Informed Artificial Neural Network Architectures for System and Input Identification of Structural Dynamics PDEs |
| title_sort | novel physics informed artificial neural network architectures for system and input identification of structural dynamics pdes |
| topic | physics-informed neural networks system identification parameter estimation input estimation |
| url | https://www.mdpi.com/2075-5309/13/3/650 |
| work_keys_str_mv | AT sarvinmoradi novelphysicsinformedartificialneuralnetworkarchitecturesforsystemandinputidentificationofstructuraldynamicspdes AT burakduran novelphysicsinformedartificialneuralnetworkarchitecturesforsystemandinputidentificationofstructuraldynamicspdes AT saeedeftekharazam novelphysicsinformedartificialneuralnetworkarchitecturesforsystemandinputidentificationofstructuraldynamicspdes AT massoodmofid novelphysicsinformedartificialneuralnetworkarchitecturesforsystemandinputidentificationofstructuraldynamicspdes |