太陽光電發電系統發電量預測模式之建構–以臺灣地區為例

Abstract

近年來國際能源情勢動盪不安，造成全球經濟嚴重衰退，許多國家逐漸意識到開發再生能源之必要性，目前已有越來越多的國家積極投入再生能源的開發工作，其中又以太陽能之相關研究最受重視。然而歷經多年研究後，太陽光電發電系統 (photovoltaic power system, 以下簡稱PV系統) 至今仍無法普遍推廣於日常生活之用，其主要癥結在於太陽能發電之成本高昂，卻不能發揮理想的供電績效。 過去有文獻針對真實環境下之PV系統進行運轉效率之評估，發現許多PV系統在複雜多變的外在環境影響下，轉換效率常無法達到規格標準。此外，過去文獻在進行PV系統發電量預測之研究時發現，由於相關研究採用電腦模擬方式預測PV系統之發電量，因此無法充分考量PV系統運作之實際外在環境因素，導致PV系統之光電轉換效率常被高估。因此本研究針對台灣地區之太陽光電發電系統之實際發電量資料進行分析，以過去歷史資料為基礎，並且利用逐步迴歸 (stepwise regression) 分析法與自組性演算法 (group method of data handling, GMDH) 建構發電量預測模式，並且找出影響系統發電量之重要因素，以有效預測太陽光電發電系統之發電量，以供相關人員改善及推廣太陽光電發電系統之用。

Recently years, most countries in the world suffer from economic recession that arose from the extremely unstable global energy situation. Nowadays, more and more countries go into development of renewable energy, especially solar energy. However, during the past year, photovoltaic power system is still not able to popularize in our daily life because of its high setup cost and low electric power generation efficiency. Some related researches evaluated the electric power generation efficiency of PV systems under the real world conditions. They found that most PV systems didn’t reach the specifications under the complicated and various external environments. Besides, some studies found that the efficiency of PV systems was often overestimated when using computer simulation to predict the electric power generation of PV systems. Therefore, many PV system users don’t confidently rely on PV systems as their main electric power supply. For this reason, this research focuses on the performance of PV systems under the real world conditions in Taiwan, aims to construct a power generation prediction model for the PV systems by using Stepwise Regression and Group Method of Data Handling. Furthermore, this research finds out what factors influence the performance of PV systems crucially to assist engineers and users to more easily understand PV systems in practice.