A New Approach to Predict Dynamic Loads Considering Highway Alignment Using Data Mining Techniques

Premature damage to heavy-duty pavement has been found be significantly caused by the vehicle–highway alignment interaction, especially in mountainous regions. This phenomenon was further verified field investigations and tests. In order elucidate potential mechanism of this it is important address vehicle dynamic loads generated interaction between pavement. Based on this, paper realizes a new method load prediction using data mining techniques, namely artificial neural network (ANN) support vector machine (SVM)). The data, including highway geometric characteristics, were collected wheel force transducer (WFT) global positioning system (GPS), respectively. coefficient determination (R2) root mean square error (RMSE) used evaluate performance models. results showed that proposed model established ANN better than SVM. Moreover, implied highly correlated with curvature longitudinal grade, furthermore, have larger effect. technique provides feasible rapid approach identify under complex interactions.