| Summary |
Zong-Xing Wu，2007.06
Feng Chia University
Porous Asphalt (PA) is widely used in Taiwan freeway’s friction course, but its thermal behavior is different from dense grade. Accurate Temperature forecast is one of the important factors in pavement design, construction and maintenance. The literatures about PA’s thermal behavior are also not many in Taiwan. The study is intended to develop linear temperature forecast model with Stepwise Multi-Regression model ( SMR ) meanwhile, conventional forecast methods including SHRP and LTPP were also adopted to comparison. The one way analysis of variance ( ANOVA ) is used to recognized the material property difference form dense grade, PC, PA and Red color PA pavements.
The independent variables of SMR’s model are field weather station data, material’s heat conduction, and illumination. Dependent variable is surface temperature. According to the ANOVA result, the PA has significant difference in heat conduction. Compare results from SMR model and conventional forecast approach, the SMR forecast model in PA has less Mean Absolute Error ( MAE ) than others. The MAE of conventional approaches is even more than eight deg C. It may be resulted from latitudinal and weather difference between Taiwan and North America. Above all, the SMR forecast model in this study is best in PA. It is worth to do more research in future. |
