Conference
Semi-Active Control of a Shear Building based on Reinforcement Learning: Robustness to measurement noise and model error
| العنوان: | Semi-Active Control of a Shear Building based on Reinforcement Learning: Robustness to measurement noise and model error |
|---|---|
| المؤلفون: | Jedlińska, Aleksandra, Pisarski, Dominik, Mikułowski, Grzegorz, Błachowski, Bartłomiej, Jankowski, Łukasz |
| المساهمون: | Institute of Fundamental Technological Research (IPPT), Polska Akademia Nauk = Polish Academy of Sciences = Académie polonaise des sciences (PAN), The authors gratefully acknowledge the support of the National Science Centre, Poland, granted under the grant agreement 2020/39/B/ST8/02615., Maria Ganzha, Leszek Maciaszek, Marcin Paprzycki, Dominik Ślęzak |
| المصدر: | Annals of Computer Science and Information Systems ; 18th Conference on Computer Science and Intelligence Systems ; https://hal.science/hal-04282960 ; 18th Conference on Computer Science and Intelligence Systems, Sep 2023, Warsaw, Poland. pp.1007-1010, ⟨10.15439/2023F8946⟩ |
| بيانات النشر: | HAL CCSD |
| سنة النشر: | 2023 |
| المجموعة: | Archive ouverte HAL (Hyper Article en Ligne, CCSD - Centre pour la Communication Scientifique Directe) |
| مصطلحات موضوعية: | structural control, semi-active control, reinforcement learning, tuned mass damper (TMD), [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], [SPI.MECA]Engineering Sciences [physics]/Mechanics [physics.med-ph] |
| جغرافية الموضوع: | Warsaw, Poland |
| الوصف: | This paper considers structural control by reinforcement learning. The aim is to mitigate vibrations of a shear building subjected to an earthquake-like excitation and fitted with a semi-active tuned mass damper (TMD). The control force is coupled with the structural response, making the problem intrinsically nonlinear and challenging to solve using classical methods. Structural control by reinforcement learning has not been extensively explored yet. Here, Deep-Q-Learning is used, which appriximates the Q-function with a neural network and optimizes initially random control sequences through interaction with the controlled system. For safety reasons, training must be performed using an inevitably inexact numerical model instead of the real structure. It is thus crucial to assess the robustness of the control with respect to measurement noise and model errors. It is verified to significantly outperform an optimally tuned conventional TMD, and the key outcome is the high robustness to measurement noise and model error. |
| نوع الوثيقة: | conference object |
| اللغة: | English |
| Relation: | hal-04282960; https://hal.science/hal-04282960; https://hal.science/hal-04282960/document; https://hal.science/hal-04282960/file/8946.pdf |
| DOI: | 10.15439/2023F8946 |
| الاتاحة: | https://hal.science/hal-04282960 https://hal.science/hal-04282960/document https://hal.science/hal-04282960/file/8946.pdf https://doi.org/10.15439/2023F8946 |
| Rights: | http://creativecommons.org/licenses/by/ ; info:eu-repo/semantics/OpenAccess |
| رقم الانضمام: | edsbas.64B98761 |
| قاعدة البيانات: | BASE |
| DOI: | 10.15439/2023F8946 |
|---|