Asphalt pavement friction coefficient prediction method based on genetic-algorithm-improved neural network (GAI-NN) model

To overcome the limitations of pavement skid resistance prediction using friction coefficient, a genetic-algorithm-improved neural network (GAI-NN) was developed in this study. First, three-dimensional (3D) point-cloud data an asphalt surface were obtained smart sensor (Gocator 3110). The coefficient then pendulum tester. 3D dataset analyzed to recover missing and perform denoising. In particular, these filled cubic spline interpolation. Parameters for texture characterization defined, methods computing parameters developed. Finally, GAI-NN model by modifying weights thresholds. test results indicated that data, capable predicting with sufficient accuracy, error 12.1%.