NEURAL NETWORKS AND AASHO ROAD TEST

The American Association of State Highway Officials (AASHO) road test is a factorial test of pavement durability that was conducted from 1958 through 1960. The primary variables of the road test were layer depths, axle load, and number of load applications. These data were processed using conventional statistical methods. The AASHO formula is the resulting databased model of road-test data. This paper reexamines the AASHO road-test data using the Monte Carlo Hierarchical Adaptive Random Partitioning (MC-HARP) neural-network model developed by Banan and Hjelmstad (1995). The authors demonstrate that data representation in the MC-HARP model is superior to data representation in the AASHO formula. In the future, the authors speculate that the MC-HARP neural network model will be an appropriate tool for the development of databased models of pavement performance.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00726127
  • Record Type: Publication
  • Files: TRIS, ATRI
  • Created Date: Sep 28 1996 12:00AM