ABSTRACT
The calibration of traffic simulation models for bridge design or assessment is normally based on weigh-in-motion or WIM data, this implies the estimation of a large number of parameters related to load distribution and vehicle weights. In a similar stochastic problem, the use of genetic algorithm in the optimization process can greatly simplify the work by allowing the calculation of a very large number of parameters even with a small amount of data and conditions. This method is illustrated by an application example of genetic optimization to evaluate the correlation between the gross vehicle weight and its distribution on single vehicle axles.
