| Popis: |
Electrical power systems are critical to the wellbeing of our economy and society. Collapse of electrical transmission towers under hurricanes may result in significant interruptions of power systems. This research proposes a framework for the development of collapse fragility curves of transmission towers subjected to hurricanes. Incremental dynamic analysis (IDA), originally established for earthquake engineering applications, is adapted to model the hurricane induced collapse behavior. For a specific site, a set of hurricane wind speed and direction records are selected from 10,000-year synthetic hurricanes using a combination of autoencoder and k-means clustering. The autoencoder first compresses each wind record into 5 latent features, to which the k-means clustering is applied. Thus, all the collected wind records are divided into 4 clusters. Twenty wind records are picked at random from the 4 clusters and employed to run the IDA analysis, through which the collapse behavior is simulated, incorporating uncertainties in wind loading. The intensity measure of fragility curves is the storm maximum gust wind speed, and therefore the fragility curve is given as the cumulative distribution function (CDF) of the collapse capacity, which is designated as the intensity measure at the onset of collapse. The parameters of a fragility curve are estimated from the simulated data of the collapse capacity using the method of moments. The developed fragility curves are helpful in damage prediction of the electrical power systems under hurricanes. |
