Evaluación de curvas de deterioro para puentes vehiculares del territorio colombiano a partir de una base de datos de condición estructural

Abstract

This work is based on the collection and standardization of information from the National Institute of Roads (INVIAS) based on visual inspection reports conducted in 2012 on bridges located in Colombia. A database was created with information on approximately 900 bridges from the different territorial directorates of INVIAS. A review of secondary information on structural deterioration models was conducted, including multivariate linear regression, generalized linear models, and multivariate polynomial regression. Initially, a descriptive statistical analysis of each variable was developed to understand its behavior and its relationship with structural deterioration. Subsequently, specific deterioration curves were generated for each structural component using polynomial regression, while predictive models adjusted to the observed deterioration characteristics were constructed using multivariate and generalized linear regression. Finally, a comparative analysis of these models was conducted, and recommendations were made for their application in bridge management, highlighting their usefulness in prioritizing interventions, optimizing resources, and improving the safety and durability of road infrastructure in Colombia.