應用類神經網路技術在金門風場之長時間風力發電預測

Abstract

由於風力發電機輸出不穩定主要原因在於風速發生變化，風力發電機輸出也跟著變化，而影響風速的因素包括地形、溫度、濕度等。因此，風力發電預測是一種非常複雜、多維度且高度非線性的系統。類神經網路技術的優點在於能學習輸入與輸出間的關係而不需要提供轉換的數學函式，可完成複雜的非線性映射、聯想、資料分類、知識處理等工作，且類神經網路具有並行處理能力可減少運算時間，故適合應用於數量眾多風機組之發電量預測。 本論文的目的在於應用類神經網路技術於預測長時間之風力發電量，並分析金門風場之風力發電量預測效果。最後，利用MATLAB實現一類神經網路結合氣象資料之風力發電預測系統，具有領前時間1~48小時之風力發電預測能力。

The wind turbine systems are unstable power sources. The main reason is that their power outputs heavily depend on wind speed. However, terrain, temperature, humidity and other factors affect wind speed. Therefore, wind power forecasting is a very complex, multi-dimensional, and highly nonlinear system. Neural network has the ability to learn the relationship between system input and output, and performs complex non-linear mapping, data classification, knowledge processing, and so forth. In addition, because neural network has the ability of parallel processing, computation time can be reduced. Therefore, it is very suitable for wind power forecasting. The purpose of this thesis is to use neural network technology to design a wind power forecasting system. Moreover, the efficiency of the proposed wind power forecasting system in Kinmen farm is described. Finally, we use MATLAB to implement the proposed wind power forecasting system with 48-hours ahead in Kinmen farm.