Study on a Test Data Validation Method for Asphalt Pavements

Abstract

To establish a test data validation methodology for asphalt pavements as well as to explore and monitor the data calculation process for false data identification, data validation methods, such as abnormal data recognition, statistical tests, and variance analysis were integrated. The general process for the data validation of an asphalt pavement test was established considering the logical link, normality, and variance decomposition of the test data. The application results indicated that a logical test with abnormal data recognition could provide an improved data environment for a statistical test. Moreover, the statistical test requires the development of a three-tiered statistical test approach comprising the Kolmogorov-Smirnov test,