Evolution Prediction of Longitudinal Crack of Asphalt Pavement Based on Artificial Neural Network

目的：预测沥青路面道路在长期服役条件下的纵向裂缝长度。方法：本研究从长期路面性能数据库中选择和处理与纵向裂缝发展相关的影响因素，建立人工神经网络预测模型，研究模型参数对精度的影响。结果：建立的模型R2为0.804，预测效果较好。结论：人工神经网络能较好预测沥青路面纵向裂缝的发展，能为预防性养护提供支持。Purpose: This paper aims to predict the longitudinal crack length of asphalt pavement under long-term service conditions. Method: In this paper, the influencing factors related to the development of longitudinal cracks were selected and processed from the long-term pavement performance database, an artificial neural network prediction model was established, and the effect of model parameters on accuracy was studied. Result: The R2 of the established model was 0.804, and the prediction effect was good. Conclusion: Artificial neural network can better predict the development of longitudinal cracks in asphalt pavement, and can provide support for preventive maintenance.