|
|
|
| Data filtering methods of FWD pavement deflection |
| 1.School of Highway， Chang′an University， Xi′an， Shaanxi 710064， China； 2.Highway Research Institute， Xinjiang Production and Construction Corps， Urumqi， Xinjiang 830002， China； 3.Architectural Engineering Research Institute， Xinjiang Production and Construction Corps， Urumqi， Xinjiang 830002， China； 4.Beijing Urban Construction and Bridge Construction Group， Beijing 100124， China |
|
|
|
|
Abstract Based on the FWD deflection data measured from an asphalt road in Xinjiang, using Pauta method, Chauvet method, Grubbs method and residual culling method to filter the data, the linear regression relationship with Benkelman Beam deflection LB was established. The results show that the linear correlation coefficient between FWD deflection data LFWD filtered by improved 1S+residual culling method and LB reaches 0.957, which is better than those of other methods under the requirements of JTG D50—2006 of Specifications for Design of Highway Asphalt Pavement. Using the improved 1S+residual culling method to filter FWD deflection data of asphalt surface, the linear regression relationship with LB was established. The linear correlation coefficient between LFWD and LB reaches 0.975, which meets normative requirements. The 1S+residual culling is an integrated method with Pauta method and residual culling, which can cull linearly independent discrete date more efficiently and can be used for filtering large amount of data.
|
|
|
|
|
|
|
|
|
2015, Vol. 36 
