Tidal currents forecasting using a hybrid of ANN and least squares model

Abstract

Forecasting is the first step for dealing with the future generation of the tidal current power. A neural network is one of the most commonly used models for forecasting. It is generally constructed from input, output and hidden layers. The least squares method is used to determine the approximate solution of over-determined system in which the number of equations is greater than the number of unknowns. In this study tidal currents models based on combining an artificial neural network (ANN) and the least squares method (LSM) were developed and evaluated for forecasting currents over a future month. The results of the least squares model are compared with those of the artificial neural networks. A hybrid model of ANN and least squares is proposed and this model gives good results compared to either the ANN or LSM alone. This study was done using data collected from the Bay of Fundy in 2008.