Fault-Tolerant Damage Control of Nonlinear Structures Using Artificial Intelligence
In this paper, the artificial intelligence is employed to design a Fault-Tolerant Controller (FTC) for structural vibrations. The FTC is designed to reduce the probability of damage considering sensor fault. For this purpose, Neural Networks (NNs) are used as fault detection and accommodation and fu...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of Tehran Press
2020-12-01
|
| Series: | Civil Engineering Infrastructures Journal |
| Subjects: | |
| Online Access: | https://ceij.ut.ac.ir/article_79022_947663bacb31bc80868e4a494c7c592e.pdf |
| _version_ | 1818251487069863936 |
|---|---|
| author | Amir Baghban Abbas Karamodin H. Haji Kazemi |
| author_facet | Amir Baghban Abbas Karamodin H. Haji Kazemi |
| author_sort | Amir Baghban |
| collection | DOAJ |
| description | In this paper, the artificial intelligence is employed to design a Fault-Tolerant Controller (FTC) for structural vibrations. The FTC is designed to reduce the probability of damage considering sensor fault. For this purpose, Neural Networks (NNs) are used as fault detection and accommodation and fuzzy logic is used as a controller. This control strategy requires two groups of neural networks. The first group of neural networks finds the faulty sensor by estimating the structural responses and comparing them with the responses obtained from the sensors. The second group has the task of estimating the response of the faulty sensor using data obtained from healthy sensors. To evaluate this method, the time history analysis of a 3-story benchmark building equipped with accelerometers and active actuators has been used. This evaluation is based on determining the probability of structural damage and the generation of fragility curves under forty ground motions. To develop fragility curves, the criteria specified in the FIMA 356 (IO, LS and CP) for the moment frame based on the inter-story drift are used. This study show that in the absence of the neural networks, sensor fault reduces the performance of the fuzzy controller and it is even possible to increase the structural responses compared to the structure without the controller. In addition, results demonstrate that the proposed control strategy can rectify the deterioration of sensor faults and decrease the probability of failure. |
| first_indexed | 2024-12-12T16:09:03Z |
| format | Article |
| id | doaj.art-8e8242453845420f97f193d3bf1fd30e |
| institution | Directory Open Access Journal |
| issn | 2322-2093 2423-6691 |
| language | English |
| last_indexed | 2024-12-12T16:09:03Z |
| publishDate | 2020-12-01 |
| publisher | University of Tehran Press |
| record_format | Article |
| series | Civil Engineering Infrastructures Journal |
| spelling | doaj.art-8e8242453845420f97f193d3bf1fd30e2022-12-22T00:19:14ZengUniversity of Tehran PressCivil Engineering Infrastructures Journal2322-20932423-66912020-12-0153239540610.22059/ceij.2020.287804.160979022Fault-Tolerant Damage Control of Nonlinear Structures Using Artificial IntelligenceAmir Baghban0Abbas Karamodin1H. Haji Kazemi2Department of Civil Engineering, University of Gonabad, Gonabad, IranAssociate Professor, Department of Civil Engineering, Faculty of Engineering, Ferdowsi University Of Mashhad (FUM)Professor, Department of Civil Engineering, Ferdowsi University of Mashhad, IranIn this paper, the artificial intelligence is employed to design a Fault-Tolerant Controller (FTC) for structural vibrations. The FTC is designed to reduce the probability of damage considering sensor fault. For this purpose, Neural Networks (NNs) are used as fault detection and accommodation and fuzzy logic is used as a controller. This control strategy requires two groups of neural networks. The first group of neural networks finds the faulty sensor by estimating the structural responses and comparing them with the responses obtained from the sensors. The second group has the task of estimating the response of the faulty sensor using data obtained from healthy sensors. To evaluate this method, the time history analysis of a 3-story benchmark building equipped with accelerometers and active actuators has been used. This evaluation is based on determining the probability of structural damage and the generation of fragility curves under forty ground motions. To develop fragility curves, the criteria specified in the FIMA 356 (IO, LS and CP) for the moment frame based on the inter-story drift are used. This study show that in the absence of the neural networks, sensor fault reduces the performance of the fuzzy controller and it is even possible to increase the structural responses compared to the structure without the controller. In addition, results demonstrate that the proposed control strategy can rectify the deterioration of sensor faults and decrease the probability of failure.https://ceij.ut.ac.ir/article_79022_947663bacb31bc80868e4a494c7c592e.pdffault diagnosisfault-tolerant controlfuzzy logic controller (flc)neural networksprobability of damage |
| spellingShingle | Amir Baghban Abbas Karamodin H. Haji Kazemi Fault-Tolerant Damage Control of Nonlinear Structures Using Artificial Intelligence Civil Engineering Infrastructures Journal fault diagnosis fault-tolerant control fuzzy logic controller (flc) neural networks probability of damage |
| title | Fault-Tolerant Damage Control of Nonlinear Structures Using Artificial Intelligence |
| title_full | Fault-Tolerant Damage Control of Nonlinear Structures Using Artificial Intelligence |
| title_fullStr | Fault-Tolerant Damage Control of Nonlinear Structures Using Artificial Intelligence |
| title_full_unstemmed | Fault-Tolerant Damage Control of Nonlinear Structures Using Artificial Intelligence |
| title_short | Fault-Tolerant Damage Control of Nonlinear Structures Using Artificial Intelligence |
| title_sort | fault tolerant damage control of nonlinear structures using artificial intelligence |
| topic | fault diagnosis fault-tolerant control fuzzy logic controller (flc) neural networks probability of damage |
| url | https://ceij.ut.ac.ir/article_79022_947663bacb31bc80868e4a494c7c592e.pdf |
| work_keys_str_mv | AT amirbaghban faulttolerantdamagecontrolofnonlinearstructuresusingartificialintelligence AT abbaskaramodin faulttolerantdamagecontrolofnonlinearstructuresusingartificialintelligence AT hhajikazemi faulttolerantdamagecontrolofnonlinearstructuresusingartificialintelligence |