Evaluation of the Acceptance of the Hot Mix Asphalt Paving Mixture Using Backpropagation Artificial Neural Network


The asphalt content in hot mix asphalt paving mixture is a key factor in producing quality pavements. In recent years, the artificial neural networks approach has attracted wide attention and found a growing number of pavement applications. This paper explores the feasibility of using the backpropagation artificial neural network with sigmoid function as activation function by MATLAB 7.8 software to determine the acceptability of the hot mix asphalt paving mixtures based on the percent of asphalt content and aggregate gradation using their Marshall properties. Several networks architectures, using two hidden layers with different numbers of nodes, are tested to obtained the best results. The results showed that the network ( 10-20-10-3) had the best performance, and this network can be used as appropriated method for determining the asphalt content and aggregate gradation acceptability of hot mix asphalt paving mixture. This work concludes that the artificial neural network is a good method which can reduce the time consumed and can be used as a tool in evaluating the hot mix asphalt paving mixtures