engleski

Feasibility Assessment of Wind Power Plant with Scarce Local Wind Data Using Cascade-Correlating Algorithm

Since the introduction of aboundant incentives schemes for wind power plants in most of European countries the interest of investors for wind power plants has substantially risen. This resulted in some economically unjustifiable investments because investors were rushing to qualify for the wind power plants approved quoatas. The most important factor in wind power plant feasibility assessment is the wind data from the past. This is the basis for selection of type and number of wind turbines. Nevertheless, wind data for potential wind power plant location from the past is often scarce, which means that investors do not have the basis for making valid decisions of the quality of potential wind power plant location. This paper describes a method for wind power plant output power estimation even if no wind data from the past is available for the location of the wind power plant. The method is based on wind characteristic measurements at the remote locations. Neural network method with cascade- correlating algorithm is used. Results were verified using the algorithm to predict the power output for wind power plant which is already in operation and significant correlation between actual and calculated wind data was obtained. Therefore, the proposed method can be used as a robust tool for potential wind power plant location feasibility assessment.

wind power plant ; output power ; feasibility assessment ; neural networks ; cascade-correlating algorithm

nije evidentirano