基於回歸模型與利用歷史數據之太陽能電力預測

The Solar Power Prediction Based On Regression Model and History Data

董俊良(Chun-Liang Tung) ； 許凱程(Kai-Cheng Hsu)

《管理學術研討會》 第十六屆 (2018/11) Pp. 1175-1184

https://doi.org/10.29893/NCUTMAN.201811.0114

摘要

在過去的幾十年裡，人口急遽成長以及工業時代的來臨，生活物質與品質的需求增加，致使石化工業的過度開發，該石化工業為壞地球環境的非再生能源製成，對地球環境造成空氣的污染，從而造成氣候的變遷，對此必須擺脫對非再生能源的依賴，從再生能源著手找尋新的能源開採方式或增加現有非再生能源開採功率（如太陽能），才能減緩對環境的損害。太陽能發電越來越受到各國政府的重視，並積極尋求提高太陽能開採功率的方式，太陽能主要取決於日照度決定發電量，隨著時間與氣候的影響，發電量會有較大波動，可能對電網、輸電和配電設備造成嚴重的負面影響，而至今仍未提出較精簡的方式預測發電量，因此，本研究提出一簡化式電量預測機制，該機制透過多元線性回歸模型的方式，僅需日照度、環境溫度以及太陽能板模組溫度組成等多元線性回歸模型，便能以較低的成本，達到未來一日發電量的精準預測。

關鍵字

太陽能發電系統 ； 電力預測 ； 多元線性回歸

並列摘要

During the past decade population has rapidly grown. The development of the industrial time and the increased quality of daily life requirements has caused the overdevelopment of the petrochemical industry. Petrochemical products are made from non-renewable resources that damage the Earth environment by causing air pollution and climate change. Therefore, it is necessary to reduce the usage of non-renewable resources, hence the need to find the new uses of renewable resources or increase the capacity of renewable resources to reduce the damage to the environment. The development of solar energy has been highly emphasized by many governments around the world who have been actively seeking for the increased usage of solar energy. The efficiency of solar power is affected by sunlight; therefore, the changes in time and climate may cause a large fluctuation in the solar energy which may damage the power grid and other power supply equipment's. There hasn't been a simplified method to measure the quantity of solar power, therefore our research is about how to measure the quantity using the linear regression model. With the measurement of the irradiance, environment temperature, and PV temperature, we can accurately measure the power production capacity in a low cost.

並列關鍵字

Photovoltaic Energy ； Power Prediction ； Multiple Linear Regression