Research On Short-term Prediction Method Of Significant Wave Height

Wave is a common and complex wave phenomenon in the ocean.Due to the influence of wind speed,direction,terrain and other factors,the size of waves varies greatly,usually ranging from a few centimeters to a few meters,but in the deep sea and extreme weather conditions up to tens of meters.The irregularity of waves brings great challenges to the safety of maritime navigation,maritime scientific research,maritime operations and the development and utilization of marine energy.Significant wave height(SWH)is an important parameter to describe wave characteristics.Prediction of significant wave height is an effective way to solve the above problems.Prediction of SWH is conducive to understanding the marine wave environment in advance,making correct decisions,and ensuring the development of marine resources and the safety of maritime crew and cargo.The availability of accurate SWH forecasts in the management routines involved in wave energy converters provides considerable support for optimizing the operational costs and improving their reliability.In the other words,by applying accurate SWH forecasting,the wave energy can be scheduled and wave power penetration will be increased.This has significant economic impact on the system operation and can substantially reduce costs.In this paper,some aspects of prediction for SWH are studied and discussed:Firstly,an SWH series prediction method based on the autoregressive integrated moving average model(ARIMA)was established after the analysis of the stability of SWH series: Time series of historical SWH are used to train the time series model so as to predict the SWH in the future.From the prediction result,the statistical errors of the time series method are all better than persistence method.Secondly,the prediction method based on physical model is established: The SWH in the north Pacific region is simulated and predicted based on the wind data from GFS and WAVEWATCH-? wave model.Through the comparison with the actual SWH data,it is proved that this method can be used for the prediction of SWH,especially for the regional wave prediction with good performance.Finally,a hybrid model based on Average Mutual Information(AMI)criterion is proposed: The features of model are confirmed by AMI criterion.Future SWH is predicted by Support Vector Regression method optimized by Genetic Algorithm(GASVR)with the consideration of driven wind and historical data.It can be seen from the statistical results that this method is superior to the time series algorithm in all statistical indicators.