The predicted model of international roughness index for drainage asphalt pavement

Roughness is a very important performance indicator in pavement maintenance management systems. The international rough index (IRI) of indicators was used as the core of roughness measuring of degree gradually in recent years domestically, but to predict a smooth degree of indicators the method is set up differently as there is a different prediction indicator because the theoretical foundation is different. This research bases its measurement data on the porous asphalt pavement of national highway number 3. We predicts the deterioration of IRI by different ways, including grey forecast, multiple regression, genetic programming. The result of this research found that there are better results in genetic programming method prediction than roughness index foundation method. The reason that it accords with the best prediction result is that heredity can carry on the change to plan in the parameter.