A Comparison of Ocean Wave Height Forecasting Methods for Ocean Wave Energy Conversion Systems

Abstract

Ocean wave height plays an important role in the operation status of ocean wave energy conversion systems. In this paper, the future continuous ocean wave height within 2~3 s is forecasted by three methods, the autoregressive moving average model (ARMA) method, backpropagation (BP) neural network method, and radial basis function (RBF) neural network method. Then, the error between suggested forecast results and corresponding measured results are compared by the root mean square error (RMSE), mean absolute error (MAE), and correlation coefficient R. The comparison result indicates that the RBF neural network method is preferred to the other two methods, having the advantage of high accuracy. Lastly, the reasons for the errors of the three forecasting methods are analyzed. This study signifies that the RBF neural network method is beneficial to the operation control and efficiency improvement of ocean wave energy conversion systems.