A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm
Accurately capturing data on the external loads that large structural systems endure is crucial for improving the performance of energy equipment. This paper introduces a novel hybrid model-data-driven framework for the dynamic load identification of interval structures, which seamlessly combines fi...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/180299 |
| _version_ | 1826118252058640384 |
|---|---|
| author | Liu, Yaru Wang, Lei Ng, Bing Feng |
| author2 | School of Mechanical and Aerospace Engineering |
| author_facet | School of Mechanical and Aerospace Engineering Liu, Yaru Wang, Lei Ng, Bing Feng |
| author_sort | Liu, Yaru |
| collection | NTU |
| description | Accurately capturing data on the external loads that large structural systems endure is crucial for improving the performance of energy equipment. This paper introduces a novel hybrid model-data-driven framework for the dynamic load identification of interval structures, which seamlessly combines finite-element modeling with machine learning techniques. To address potential ill-posed issues in model-driven methods and the interpretability limitations of data-driven methods, we propose a physics-informed neural network. This neural network effectively inverts uncertain modal responses with low data requirements and high predictive performance high by integrating the underlying modal transformation equation into the loss function of a fully connected neural network. To identify the modal loads using predicted modal displacement/acceleration responses, we introduce a pioneering dynamics inversion method. This method modifies the traditional Kalman filter with an assumption of unknown inputs to reduce the sensitivity of load identification process to different noises. In addition, our approach incorporates a subinterval Chebyshev expansion method to adaptively determine the interval boundaries of external loads. The efficiency of the proposed method is assessed through two numerical examples and validated through comparative research against baseline methods. Our findings suggest that this approach enhances precision, robustness, and generalization in dynamic load identification, even when facing challenges such as limited training data, significant noise interference, and non-zero initial conditions. |
| first_indexed | 2025-03-09T12:07:55Z |
| format | Journal Article |
| id | ntu-10356/180299 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2025-03-09T12:07:55Z |
| publishDate | 2024 |
| record_format | dspace |
| spelling | ntu-10356/1802992024-10-01T01:43:41Z A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm Liu, Yaru Wang, Lei Ng, Bing Feng School of Mechanical and Aerospace Engineering Engineering Dynamic load identification Model-data-drive method Accurately capturing data on the external loads that large structural systems endure is crucial for improving the performance of energy equipment. This paper introduces a novel hybrid model-data-driven framework for the dynamic load identification of interval structures, which seamlessly combines finite-element modeling with machine learning techniques. To address potential ill-posed issues in model-driven methods and the interpretability limitations of data-driven methods, we propose a physics-informed neural network. This neural network effectively inverts uncertain modal responses with low data requirements and high predictive performance high by integrating the underlying modal transformation equation into the loss function of a fully connected neural network. To identify the modal loads using predicted modal displacement/acceleration responses, we introduce a pioneering dynamics inversion method. This method modifies the traditional Kalman filter with an assumption of unknown inputs to reduce the sensitivity of load identification process to different noises. In addition, our approach incorporates a subinterval Chebyshev expansion method to adaptively determine the interval boundaries of external loads. The efficiency of the proposed method is assessed through two numerical examples and validated through comparative research against baseline methods. Our findings suggest that this approach enhances precision, robustness, and generalization in dynamic load identification, even when facing challenges such as limited training data, significant noise interference, and non-zero initial conditions. The authors would like to thank the National Natural Science Foundation of China (12072007, 12132001, 52192632), the China Scholarship Council (No. 202206020119), the Academic Excellence Foundation of BUAA for PhD Students, and the Defense Industrial Technology Development Program (JCKY2019205A006, JCKY2019203A003, JCKY2021204A002) for the financial supports. 2024-10-01T01:43:41Z 2024-10-01T01:43:41Z 2024 Journal Article Liu, Y., Wang, L. & Ng, B. F. (2024). A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm. Applied Energy, 359, 122740-. https://dx.doi.org/10.1016/j.apenergy.2024.122740 0306-2619 https://hdl.handle.net/10356/180299 10.1016/j.apenergy.2024.122740 2-s2.0-85183638880 359 122740 en Applied Energy © 2024 Elsevier Ltd. All rights reserved. |
| spellingShingle | Engineering Dynamic load identification Model-data-drive method Liu, Yaru Wang, Lei Ng, Bing Feng A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm |
| title | A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm |
| title_full | A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm |
| title_fullStr | A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm |
| title_full_unstemmed | A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm |
| title_short | A hybrid model-data-driven framework for inverse load identification of interval structures based on physics-informed neural network and improved Kalman filter algorithm |
| title_sort | hybrid model data driven framework for inverse load identification of interval structures based on physics informed neural network and improved kalman filter algorithm |
| topic | Engineering Dynamic load identification Model-data-drive method |
| url | https://hdl.handle.net/10356/180299 |
| work_keys_str_mv | AT liuyaru ahybridmodeldatadrivenframeworkforinverseloadidentificationofintervalstructuresbasedonphysicsinformedneuralnetworkandimprovedkalmanfilteralgorithm AT wanglei ahybridmodeldatadrivenframeworkforinverseloadidentificationofintervalstructuresbasedonphysicsinformedneuralnetworkandimprovedkalmanfilteralgorithm AT ngbingfeng ahybridmodeldatadrivenframeworkforinverseloadidentificationofintervalstructuresbasedonphysicsinformedneuralnetworkandimprovedkalmanfilteralgorithm AT liuyaru hybridmodeldatadrivenframeworkforinverseloadidentificationofintervalstructuresbasedonphysicsinformedneuralnetworkandimprovedkalmanfilteralgorithm AT wanglei hybridmodeldatadrivenframeworkforinverseloadidentificationofintervalstructuresbasedonphysicsinformedneuralnetworkandimprovedkalmanfilteralgorithm AT ngbingfeng hybridmodeldatadrivenframeworkforinverseloadidentificationofintervalstructuresbasedonphysicsinformedneuralnetworkandimprovedkalmanfilteralgorithm |