Clogging Prediction of Permeable Pavement

This study considers the clogging progression prediction on the permeable pavement by using artificial neural networks (ANNs). Clogging, which is caused primarily by sediment deposition, may result in performance failure of permeable pavement. Measuring the volumetric water content (VWC) by time domain reflectometers (TDRs) is an automated method to track the speed of clogging. Monitoring peak VWC during rain events has been used as an indication of clogging progression over the permeable pavement. New nonlinear solutions are developed to estimate the peak VWC using a multilayer perceptron (MLP) structure. The rain event variables and the maintenance treatment were formulated as the basic site characteristic parameters that affect the clogging progression. Five ANN models are constructed from the recorded VWC to compute the peak VWC from the rainfall parameters and maintenance treatment. A comprehensive set of data, including various rain event characteristics obtained from the rain gauge and the conducted maintenance on the permeable pavement, are used for training and testing the neural network models. The performances of the ANN models are assessed and the results demonstrate the satisfactory accuracy of the models as compared with the measured values. A parametric study is completed to determine the relative importance of peak VWC resulting from the variation of the study parameters. The results indicate that the models are effectively capable of estimating the peak VWC by the permeable pavements for different locations along the permeable pavement. The MLP models consider all known contribution factors and provide more precise prediction value than the linear model. Peak 5-min intensity, the previous rainfall depth, and the cumulative rainfall depth from the installation are the most effective parameters on the hydrologic performance of the permeable pavement. Designing permeable pavement based on the important parameters can lead to more efficient future design.