The utilization of renewable energy sources, including wind energy became of in-creasing importance during the last decades in the world’s power system. The European Union has aimed to produce 20% of its primary energy consumption by 2020 this way. Notwithstanding to the doubtless advantages, the users of “green energy” have to face several problems to resolve.
Critics of wind power argue with the intermittent energy production of the turbines which has a negative effect on the security of the power system. The limited ability to forecast wind means that the delivered power of the wind power plants cannot be pres-aged with sufficient accuracy. As a consequence, to ensure secure system operation, i.e. to be able to control the wind power caused uncertainties, the energy supplied by wind should have a limited proportion in the total electricity production. The current regula-tory environment in Hungary does not allow installing new wind power plants, since by 2011 the 330 MW of total wind turbine performance defined by the operative regulation was built in. For capacity expansion a more precise predictability of wind power would be vital. To ensure this, several methods can be suggested, e.g. energy storage.
This work focuses on making production forecasts more precise, utilizing several methods. Meteorological atmospheric forecasts give scope for the determination of the turbines’ supply data, on which further statistical methods can be applied to refine the results even more. On the basis of the achieved calculations it is proven that beside the wind velocity as main influencing factor, air density cannot be neglected, either. Fur-thermore, by autoregressive filtering the forecast accuracy can also be improved.