Use of generated artificial road profiles in road roughness evaluation

In the evaluation of road roughness and its effects on vehicles response in terms of ride quality, loads induced on pavement, drivers＇ comfort, etc., it is very common to generate road profles based on the equation provided by ISO 8608 standard, according to which it is possible to group road surface profiles into eight different classes. However, real profiles are significantly different from the artificial ones because of the non-stationary fea- ture of the first ones and the not full capability of the ISO 8608 equation to correctly describe the frequency content of real road profiles. In this paper, the international roughness index, the frequency-weighted vertical acceleration awz according to ISO 2631, and the dynamic load index are applied both on artificial and real profiles, highlighting the different results obtained. The analysis carried out in this work has highlighted some limitation of the ISO 8608 approach in the description of performance and conditions of real pavement profiles. Furthermore, the different sensitivity of the various indices to the fitted power spectral density parameters is shown, which should be taken into account when performing analysis using artificial profiles.

Local calibration of JPCP transverse cracking and IRI models using maximum likelihood estimation

The calibration of transfer functions is essential for accurate pavement performance predictions in the PavementME design. Several studies have used the least square approach to calibrate these transfer functions. Least square is a widely used simplistic approach based on certain assumptions. Literature shows that these least square approach assumptions may not apply to the non-normal distributions. This study introduces a new methodology for calibrating the transverse cracking and international roughness index(IRI) models in rigid pavements using maximum likelihood estimation(MLE). Synthetic data for transverse cracking, with and without variability, are generated to illustrate the applicability of MLE using different known probability distributions(exponential,gamma, log-normal, and negative binomial). The approach uses measured data from the Michigan Department of Transportation's(MDOT) pavement management system(PMS) database for 70 jointed plain concrete pavement(JPCP) sections to calibrate and validate transfer functions. The MLE approach is combined with resampling techniques to improve the robustness of calibration coefficients. The results show that the MLE transverse cracking model using the gamma distribution consistently outperforms the least square for synthetic and observed data. For observed data, MLE estimates of parameters produced lower SSE and bias than least squares(e.g., for the transverse cracking model, the SSE values are 3.98 vs. 4.02, and the bias values are 0.00 and-0.41). Although negative binomial distribution is the most suitable fit for the IRI model for MLE, the least square results are slightly better than MLE. The bias values are-0.312 and 0.000 for the MLE and least square methods. Overall, the findings indicate that MLE is a robust method for calibration, especially for non-normally distributed data such as transverse cracking.

Prediction of pavement roughness using a hybrid gene expression programming-neural network technique

Effective prediction of pavement performance is essential for transportation agencies to appropriately strategize maintenance, rehabilitation, and reconstruction of roads. One of the primary performance indicators is the international roughness index （IRI） which rep- resents the pavement roughness. Correlating the pavement roughness to other perfor- mance measures has been under continuous development in the past decade. However, the drawback of existing correlations is that most of them are not practical yet reliable for prediction of roughness. In this study a novel approach was developed to predict the IRI, utilizing two data sets extracted from long term pavement performance （LTPP） database. The proposed methodology included the application of a hybrid technique which combines the gene expression programming （GEP） and artificial neural network （ANN）. The developed algorithm showed reasonable performance for prediction of IRI using traffic parameters and structural properties of pavement. Furthermore, estimation of present IRI from his- torical data was evaluated through another set of LTPP data. The second prediction model also depicted a reasonable performance power. Further extension of the proposed models including different pavement types, traffic and environmental conditions would be desir- ab]e in future studies.