Evolutionary spectra-based time-varying coherence function and its engineering application

Yan Jiang, Liuliu Peng, Guoqing Huang, Xinzhong Chen

Research output: Contribution to conferencePaperpeer-review

1 Scopus citations


Priestley's evolutionary spectral theory, due to its physical meaning, is extensively employed in the characterization, modeling and simulation of multivariate nonstationary excitation processes and the structural dynamic analysis under these excitations. When the nonstation-ary processes are modeled as oscillatory processes, the evolutionary spectra-based coherence function becomes time-invariant. The sigma-oscillatory process theory and Wold-Cramer decomposition model can be used to account for the time-varying coherence function. The Wold-Cramer decomposition model is a special case of the general sigma-oscillatory process theory. Due to the simplicity of Wold-Cramer decomposition model, it is applied to model the nonstationary winds and the corresponding alongwind response analysis framework of tall buildings in engineering field. Taken the measured wind speed data as an numerical example, the time-varying coherence function models of a downburst event are developed. The influence of time-varying coherence of nonstationary winds on the alongwind tall building response is investigated. Results show that use of conventional time-invariant coherence function may underestimate the alongwind tall building response in comparison with the time-varying coherence function.

Original languageEnglish
StatePublished - 2017
Event2017 13th Americas Conference on Wind Engineering, ACWE 2017 - Gainesville, United States
Duration: May 21 2017May 24 2017


Conference2017 13th Americas Conference on Wind Engineering, ACWE 2017
Country/TerritoryUnited States


  • Alongwind response
  • Downburst
  • Evolutionary power spectral density (EPSD)
  • Multivariate nonstationary random process
  • Tall building
  • Time-varying coherence function


Dive into the research topics of 'Evolutionary spectra-based time-varying coherence function and its engineering application'. Together they form a unique fingerprint.

Cite this