Fuzzy optimization BP neural network model for pavement performance assessment

This paper discussed the use of fuzzy optimization BP neural network as a management tool for the maintenance of expressway pavement. The algorithm was applied to model pavement performance evaluation, successful in accurately and quickly capturing the high nonlinear characteristics between objective practical data and subjective synthetic evaluations, avoided the neural network activation function´s defects of unclear physical meanings. A large training set of actual pavement performance cases was used to train the network and test the generalization ability of the system. The results show that the trained network model was able to accurately and effectively assess pavement performance.