A Hybrid ANFIS- PSO Model for Scour Depth Prediction

In recent years, newly-developed data mining and machine learning techniques have been applied in various fields to build intelligent information systems. However, few of these approaches offer online support or are flexibleto be adapted to large and complex datasets. Therefore, the present research work adopts Particle Swarm Optimization (PSO) techniques to obtain appropriate parameter settings for membership function and integrates the Adaptive-Network-based Fuzzy Inference System (ANFIS) model to make the model fit for predicting scour depth. A dataset of 188 scour depths for single piers presented by the USGS was used. Results of the model prediction show that the derived model is best fitted to the field data. The proposed one-order momentum method is able to learn quickly through one-pass training and provides high-accuracy short-term predictions. Moreover, this method is suitable for online learning but the two-order momentum method is appropriate for incremental learning. The PSO-ANFIS approach could provide better results in predicting scour depths compared with other models.