Comparison of different techniques for short term prediction of tidal current speeds

Numerous techniques have been suggested for extracting energy from the sea. Tidal current turbines are a convenient method for extracting power from oceanic currents. Regarding the fact that they share many similarities to conventional wind power plants, due to the higher density of sea water they can produce 800 times more power when compared with an equivalent wind turbine of the same size operating with a similar wind speed. Apart from having remarkable power densities, tidal currents have a reputation for being very predictable and predictability is crucial to successful integration of renewable resources into the electrical grid. It is thus in this context that the need for powerful predicting algorithms for the operation and control of such resources can be perceived. This paper compares several predicting algorithms by analyzing their results using actual recorded tidal current speeds of a site in North America as input. It is found that if a tidal current flow regime at a particular location has been properly studied over time its variation can be predicted with considerable accuracy (up to %99) for operational purposes.