Performance Evaluation of Asphalt Pavement Based on BP Neural Network

Abstract

Asphalt pavement is widely used in road pavement because of its high temperature stability, low temperature crack resistance, water stability, fatigue resistance and many other advantages. However, with the increase of the service time, and the long-term effects of the surrounding environment and traffic loads, it is easy to cause cracks, subsidence, and water damage on the asphalt pavement, which brings enormous financial pressure to maintenance. For this reason, this study proposes an evaluation and prediction model of pavement performance based on BP artificial neural network through the analysis of the characteristics of pavement performance evaluation and use of the multi-factor fuzzy computing capability of artificial neural network algorithm, and adopts this model to predict the rut depth and flatness of asphalt pavements. The results show that the prediction data is in good agreement with the actual measured data, with high accuracy and small errors, which proves that the model has extremely high operability. This provides a theoretical basis for the prediction of asphalt pavement performance, which is of great practical significance to reduce the cost of asphalt pavement maintenance and improve the performance of asphalt pavement.