Estimation of extreme structural response distributions for mean recurrence intervals based on short-term monitoring

Performance assessment of existing bridges with reliability theories is an important research topic for civil infrastructure systems. The key step to calculating the bridge reliability in its lifetime is to convert two random processes, the structural capacity R and the load effect Q, into variables following certain distribution types. This study develops a framework to estimate the extreme structural response due to live loads in a mean recurrence interval based on short-term monitoring. The extreme structural response is expressed with a Gumbel distribution based on the extreme value in a specific interval, and the accuracy is evaluated by the convergence of the distribution parameters for different intervals. The Gumbel distribution is derived from the extreme value theory and validated by using the Monte Carlo Simulation. The distribution parameters are estimated using the maximum likelihood parameter estimation method. Two example bridges are studied to demonstrate the application of the developed methodology. The predicted extreme structural response distribution in terms of different mean reoccurrence intervals can be used in corresponding reliability assessment of existing bridges, which will provide useful information for bridge management.