China
Africa
Explore chapters and articles related to this topic
Composite pavement roughness modeling for LTPP wet freeze climate region using machine learning
Published in Inge Hoff, Helge Mork, Rabbira Garba Saba, Eleventh International Conference on the Bearing Capacity of Roads, Railways and Airfields, Volume 3, 2022
R. Barros, H. Yasarer, W. Uddin, S. Sultana, Y. Najjar
Pavement performance modeling is a challenging task due to the complexity of the pavement structure and its responses under traffic loading, dynamic weather and climate changes, variability in construction activities, and the interaction among all these elements (Gupta, Ankit; Kumar, Praveen; Rastogi 2012). Advanced modeling techniques such as artificial neural networks (ANNs) have been used successfully in several studies offering significant improvements over traditional techniques (i.e., linear regression) by processing large volumes of data with excellent accuracy. However, several performance models in the literature use the Long-Term Performance Pavement (LTPP) database to develop models without considering specific conditions of local climate and geography, which makes the model less accurate.
Multiobjective optimization of asphalt pavement design and maintenance decisions based on sustainability principles and mechanistic-empirical pavement analysis
Published in International Journal of Sustainable Transportation, 2018
Dan Chong, Yuhong Wang, Zhifeng Dai, Xiaojun Chen, Dawei Wang, Markus Oeser
As shown in Figure 1, a key component that bridges the decision variables with the economic and environmental performance is pavement performance modeling. Recently, pavement performance modeling has been advanced from empirical/statistical approaches (e.g., AASHTO, 1993, Chen, Dong, Zhu, Huang, & Burdette, 2017b) to ME approach. The ME-PDG developed in the United States is an example of the ME approach (Hallin, 2004). In ME-PDG, stresses and strains caused by individual traffic loads and their cumulative effects are computed mechanistically, while the resultant pavement distresses (cracks and permanent deformation) and functional performance (roughness) are predicted based on mechanistic analysis results and calibrated empirical equations using data from the Long-Term Pavement Performance Program (FHWA, 2017). Compared with the traditional empirical/statistical approaches, ME-PDG is capable of considering more performance influencing factors and their interactions. However, unlike the traditional approaches, ME-PDG approach does not provide a single equation that connects the decision variables to pavement performance indicators. This creates a challenge for quantitatively assessing the effects of decision variables.