Details
| Original language | English |
|---|---|
| Title of host publication | 2019 IEEE Symposium Series on Computational Intelligence, SSCI 2019 |
| Subtitle of host publication | Proceedings |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 784-790 |
| Number of pages | 7 |
| ISBN (electronic) | 9781728124858 |
| ISBN (print) | 9781728124865 |
| Publication status | Published - Dec 2019 |
| Event | 2019 IEEE Symposium Series on Computational Intelligence, SSCI 2019 - Xiamen, China Duration: 6 Dec 2019 → 9 Dec 2019 |
Abstract
Modern approaches to solve dynamic problems, where random vibrations are the governing excitations, are in most cases based on the concept of the power spectrum as the core model for the representation of excitation and response processes. This is partly due to the practical applicability of spectral models for frequency domain analysis. In addition, compatible time-domain samples can easily be generated. Such samples can be used for numerical performance evaluation of systems or structures represented by complex non-linear models.The development of spectral estimation methods that use ensemble statistics to generate a single or finite number of deterministic spectra results in spectral models that can be applied directly in structural analysis. However, the properties of the measured environmental process are still lost.In order to produce reliable and realistic power spectra for the application to systems, in most cases not enough real data sets are available. To capture the epistemic uncertainties of the model by taking into account inherent statistical differences that exist across real data sets, an approach for a stochastic representation of the loads can be used. In this work, the epistemic uncertainties in the spectral density of the process are captured by using an interval approach which, in combination with the stochastic nature of the process, leads to an imprecise probability model. From all the available power spectra of the ensemble, one power spectrum is identified on which the resulting relaxed power spectrum is based. To relax the power spectrum, interval parameters are implemented, thereby forming an enveloping boundary for all estimated power spectra. In order to capture the epistemic uncertainties and to present this information effectively, imprecise probabilities are used in this newly developed load representation.The relaxed power spectrum is validated by application to a single-degree-of-freedom system and a multiple-degree-of-freedom system by determining and analysing the response spectra of the systems.
Keywords
- fuzzy set theory, imprecise probabilities, power spectrum estimation, random vibrations, relaxed power spectrum, stochastic process, uncertainty quantification
ASJC Scopus subject areas
- Computer Science(all)
- Artificial Intelligence
- Computer Science(all)
- Computer Science Applications
- Mathematics(all)
- Modelling and Simulation
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
2019 IEEE Symposium Series on Computational Intelligence, SSCI 2019: Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. p. 784-790 9002899.
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Development of a Relaxed Stationary Power Spectrum using Imprecise Probabilities with Application to High-rise Buildings
AU - Behrendt, Marco
AU - Comerford, Liam
AU - Beer, Michael
N1 - Funding information: ACKNOWLEDGEMENT This work was funded by the Deutsche Forschungsgemein-schaft (German Research Foundation) grants BE 2570/4-1 and CO 1849/1-1 as part of the project ’Uncertainty modelling in power spectrum estimation of environmental processes with applications in high-rise building performance evaluation’.
PY - 2019/12
Y1 - 2019/12
N2 - Modern approaches to solve dynamic problems, where random vibrations are the governing excitations, are in most cases based on the concept of the power spectrum as the core model for the representation of excitation and response processes. This is partly due to the practical applicability of spectral models for frequency domain analysis. In addition, compatible time-domain samples can easily be generated. Such samples can be used for numerical performance evaluation of systems or structures represented by complex non-linear models.The development of spectral estimation methods that use ensemble statistics to generate a single or finite number of deterministic spectra results in spectral models that can be applied directly in structural analysis. However, the properties of the measured environmental process are still lost.In order to produce reliable and realistic power spectra for the application to systems, in most cases not enough real data sets are available. To capture the epistemic uncertainties of the model by taking into account inherent statistical differences that exist across real data sets, an approach for a stochastic representation of the loads can be used. In this work, the epistemic uncertainties in the spectral density of the process are captured by using an interval approach which, in combination with the stochastic nature of the process, leads to an imprecise probability model. From all the available power spectra of the ensemble, one power spectrum is identified on which the resulting relaxed power spectrum is based. To relax the power spectrum, interval parameters are implemented, thereby forming an enveloping boundary for all estimated power spectra. In order to capture the epistemic uncertainties and to present this information effectively, imprecise probabilities are used in this newly developed load representation.The relaxed power spectrum is validated by application to a single-degree-of-freedom system and a multiple-degree-of-freedom system by determining and analysing the response spectra of the systems.
AB - Modern approaches to solve dynamic problems, where random vibrations are the governing excitations, are in most cases based on the concept of the power spectrum as the core model for the representation of excitation and response processes. This is partly due to the practical applicability of spectral models for frequency domain analysis. In addition, compatible time-domain samples can easily be generated. Such samples can be used for numerical performance evaluation of systems or structures represented by complex non-linear models.The development of spectral estimation methods that use ensemble statistics to generate a single or finite number of deterministic spectra results in spectral models that can be applied directly in structural analysis. However, the properties of the measured environmental process are still lost.In order to produce reliable and realistic power spectra for the application to systems, in most cases not enough real data sets are available. To capture the epistemic uncertainties of the model by taking into account inherent statistical differences that exist across real data sets, an approach for a stochastic representation of the loads can be used. In this work, the epistemic uncertainties in the spectral density of the process are captured by using an interval approach which, in combination with the stochastic nature of the process, leads to an imprecise probability model. From all the available power spectra of the ensemble, one power spectrum is identified on which the resulting relaxed power spectrum is based. To relax the power spectrum, interval parameters are implemented, thereby forming an enveloping boundary for all estimated power spectra. In order to capture the epistemic uncertainties and to present this information effectively, imprecise probabilities are used in this newly developed load representation.The relaxed power spectrum is validated by application to a single-degree-of-freedom system and a multiple-degree-of-freedom system by determining and analysing the response spectra of the systems.
KW - fuzzy set theory
KW - imprecise probabilities
KW - power spectrum estimation
KW - random vibrations
KW - relaxed power spectrum
KW - stochastic process
KW - uncertainty quantification
UR - http://www.scopus.com/inward/record.url?scp=85080901785&partnerID=8YFLogxK
U2 - 10.1109/SSCI44817.2019.9002899
DO - 10.1109/SSCI44817.2019.9002899
M3 - Conference contribution
AN - SCOPUS:85080901785
SN - 9781728124865
SP - 784
EP - 790
BT - 2019 IEEE Symposium Series on Computational Intelligence, SSCI 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2019 IEEE Symposium Series on Computational Intelligence, SSCI 2019
Y2 - 6 December 2019 through 9 December 2019
ER -