Presenting a new estimation technique to model International Roughness Index and treatment improvement

International Roughness Index is one of the most important indices to assess the pavement quality, and it is generally employed in pavement maintenance planning. In this study, by virtue of differential evolutionary programming and Long Term Pavement Performance data, two prediction models are introduced to predict pavement deterioration function and treatments’ improvement. For pavement deterioration modeling, climate condition, traffic, and pavement structural features are taken into account. Besides, the most-applicable pavement treatments utilized in Iran are considered in this investigation. 520 and 336 data are applied to model pavement deterioration and treatments’ improvement. The predicted models have a very high accuracy and can be used in various fields. The pavement deterioration model and treatments’ improvement model reach the coefficient determination of 99% and 97% for both training and testing data, which indicates the high-level accuracy of the introduced models. The treatments’ improvement model provides the mean squared error of 0.2 approximately.