Search

In this study, two hybrid biogeography based and particle swarm optimization-radial basis function (BBO-RBF, and PSO-RBF) were proposed to predict the California-bearing capacity of pond ash stabilized with lime (LI) and lime sludge (LS). The model had five inputs and CBR as output variable. Comparing the measured with those predicted by BBO-RBF and PSO-RBF models, the proposed models have R2 larger than 0.9932. It means that the correlation between measured and predicted values from models is mostly the same to show the highest accuracy. In the training phase, PSO-RBF has better results, while in the testing phase, the best value of statistical indices is owned by BBO-RBF. The OBJ value for PSO-RBF is lower than those of BBO-RBF. Therefore, the PSO-RBF model outperforms the BBO-RBF model for predicting the CBR value of pond ash stabilized with LI and LS, consequently being recognized as the proposed model.