Spectral power density models and Spectral Representation Method
Stochastic process models are responsible for characterising ground motions, representing the stochastic excitations applied upon engineering structures1. To this end, a series of power spectral density models are developed and employed in stochastic dynamic analysis2. Notably, Kanai Tajimi model plays a foundational role3. Beyond which, nonstationary model attract more attention in recent years.
- Define a base Kanai Tajimi model;
- Define both separable and non-separable evolutionary power spectral density models;
- Generate sample realizations via the Spectral Representation Method;
- A general framework enabling easy addition of more nonstationary models via subclassing.
- Kanai Tajimi PSD model
where
- separable EPSD
Define an evolutionary spectrum in the form
with an example of modulating function:
- non-separable EPSD
An evolutionary spectrum with fully coupled time and frequency nonstationarity. Define an example EPS:
- Spectral Representation Method
Verification of the Spectral Representation Method with a stationary process with known PSD.
ktnsrm
was created by Y. Chen. It is licensed under the terms of the MIT license.