Real-time control of wave energy converters requires knowledge of future incident wave elevation in order to approach optimal efficiency of wave energy extraction. We present an approach where the wave elevation is treated as a time series and it is predicted only from its past history. A comparison of a range of forecasting methodologies on real wave observations from two different locations shows how the relatively simple linear autoregressive model, which implicitly models the cyclical behavior of waves, can offer very accurate predictions of swell waves for up to two wave periods into the future.
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.
[Show abstract][Hide abstract]ABSTRACT: This paper describes an expert system designed for the analysis of an incomplete, non-stationary and non-Gaussian, long-term, time series of wave significant heights by means of specific linear parametric model. Using this system makes it possible to complete missing-value gaps, forecast wave-height short-term evolution or simulate arbitrarily long sequences of wave data preserving the key statistical properties of the observed series, including autocorrelation, persistence over threshold, non-Gaussian distribution and seasonality. The implemented improvements bear on specific key tasks of ARMA setup procedure, i.e. preliminary analysis, parameter estimation and optimal model-configuration identification. Specifically, a Seasonal Trend decomposition based on Loess robust method is applied to compute more stable and detailed seasonal trend, allowing assuming more confidently its deterministic nature. Moreover, aiming at accurately estimating the model parameters, a proficient method is taken in, which is based on the robust Whittle’s approximation of the maximum log-likelihood function as well as on the direct-search, non-linear, multi-parameter, constrained, optimization technique called complex modified. Finally, an automatic expert system is developed, able to identify, almost correctly, ARMA orders by selecting the model with the smallest residuals variance and parameter numbers. Confident applicability of the suggested procedure is tested by means of both Monte Carlo simulations and comparisons of generated series with observed one, this latter measured offshore Alghero – Italy. Analysis of results clearly indicate that the accuracy in identifying the correct ARMA model is improved; furthermore, it is shown that the simulated time series exhibit all the primary statistical properties of the observed data, including winter and summer seasonal patterns as well as sea states sequencing, persistence and severeness.
Full-text · Article · Sep 2015 · International Journal of Engineering and Technical Research
"where on the right side of the equation í µí¼ í µí± = í µí¼̂ í µí± when í µí± < í µí± In  and  the í µí°´í µí± model considered achieved exceptional forecast accuracy of surface elevation (essentially perfect for a wave cycle) in the different case studies. This was achieved by first pre-processing the data using a noncausal zero phase (forward and backward filter pass in time) low-pass filter before training and querying the í µí°´í µí± model input data. "
"The first aspect means that large amount of reactive power is required e up to ten times greater than the amount of absorbed power in average  . The second aspect means that prediction of the shortterm future of the excitation force (then of the incident wave) is necessary, which remains a research challenge [18,19]. Thus, the decision was taken not to use this kind of control for the SEAREV device. "
[Show abstract][Hide abstract]ABSTRACT: In this paper, technical and economical studies conducted on the SEAREV Wave Energy Converter (WEC) are presented. This technology was first proposed in 2002 with the aim of addressing critical challenges in wave energy conversion. It consists of a closed floating hull in which a heavy pendulum oscillates. The controlled relative motion of the pendulum is used to produce electricity.The SEAREV WEC was developed over twelve years. Through the development process, three main generations of the technology were defined and studied in detail. They are presented in the paper, showing how each new generation brought significant improvements over the previous generation with respect to production and costs.In 2013, an economical model for wave energy farms was developed to assess the viability and competitiveness of the SEAREV technology. Although results show that the SEAREV technology is a sound technical solution, the cost of energy projection is still too high to allow direct access to mass electricity market in European countries in the short term.Finally, in light of the lessons learned from the SEAREV WEC development, the need for innovative technologies with higher technology performance level (TPL) or alternative markets is discussed in the context of wave energy.