Modelling and Prediction of Ocean Wave Parameters With MEA-SVR Algorithm

Aiming to overcoming the shortcomings of strong dependence on data volume and low prediction accuracy in ocean wave amplitude prediction, the mind evolutionary algorithm was combine with into the support vector regression algorithm(abbreviated as MEA-SVR algorithm) to modelling and prediction the ocean wave amplitude in this paper. In the paper we selected 87600 observation data points from China Oceanic Administration observations in 10 locations of Chinese Ocean, conducted modelling and simulation experiments with MEA-SVR algorithm, and selected GA-BP, MEA-BP, PSO-SVR algorithms for comparative analysis purpose. Simulation results show that the prediction results of our MEA-SVR algorithm are better than other models in parameters of average absolute error, relative error error, correlation coefficient, root mean square error, and consistency index. so has a broad application prospect in ocean energy harvisting.