Evaluation of Fuzzy Genetic Algorithm and Neural Network In optimization of Bridge Structural Health Monitoring System

Structural damage detection is based on that the dynamic response of structure will change because of damage. Hence, it is possible to estimate the location and severity of damage before and after the damage. In this study, damage detection issue based on modal parameters for an optimization problem using neural network and fuzzy genetic system offered and the effectiveness of these two methods in detecting the location and also the severity of the damage is assessed. For studying damage detection, the numerical model of the Crowchild bridge is made by its dynamic characteristics and has been used for various damage scenario detection. In the first method, the natural frequency, and in the second method, modal strain energy is selected as a damage indicator. Using simplified models to study the behavior of bridges due to their simplicity and acceptable accuracy is very common. The results show that the Genetic Fuzzy System can be more successful when a simplified model is used. Comparing the results of two failure detection methods shows that the fuzzy system is less sensitive to existing uncertainties.