Performance Prediction of Highway Asphalt Pavement Based on IFA-SVM

Abstract

To solve the problem of the low accuracy of the traditional qualitative method for highway asphalt pavement performance, a prediction model based on improved firefly algorithm (IFA)—support vector machine (SVM) is established by combining SVM theory and IFA. First, firefly field search is introduced into the prediction model to overcome the random movement of fireflies with the increase in the number of iterations in the optimization process. Second, in the subsequent optimization process, the dynamic adjustment algorithm is used to search the step size to balance the global search ability, which accelerates the optimization selection of the performance parameters of the SVM model. Finally, the example is verified and compared with the standard FA-SVM prediction method to verify the validity of the IFA-SVM model and the feasibility of prediction accuracy. The result shows the following: (1) The maximum relative error is 2.5435% and the minimum is 0.8206% when the standard FA-SVM is used to predict the pavement performance of the Baiyin section of the G6 expressway. The maximum relative error is 1.0858% and the minimum is 0.3654%, and their root mean square error is smaller than that of the standard FA-SVM method; and (2) the IFA-SVM model has a faster convergence rate and a higher accuracy than the standard FA-SVM when predicting the performance of asphalt pavement on highways. The prediction result is not only closer to the measured value but also provides effective support for the maintenance decision of asphalt pavement on highways.