| Summary: | 碩士 === 聖約翰科技大學 === 電機工程系碩士在職專班 === 106 === Electric energy is an important basis for the economic development of a country. It is also an
essential energy source of human life. Effective use of solar power can prevent the waste of
power resource and increase power supply quality. In micro-grids, accurate photovoltaic
prediction has many benefits, like providing proper power scheduling, reducing operating costs
even improving power supply reliability.
The traditional solar photovoltaic forecasting method designs the prediction model by
environmental information and solar photovoltaic module information. The relationship
between these two data and the photovoltaic power generation amount is nonlinear. The
traditional method is difficult to solve the nonlinear problems. This paper confirms the
relationship between input parameters and output generation amount. By the correlation
coefficients in the statistics, the relative influence of each parameter can be obtained. In the
proposed forecast method, illumination is considered as one of the forecast data. However, if
the illumination changes greatly, the fluctuation of power generation and the prediction error
will greatly increase.
|
|---|
