太陽能光電系統發電評估：故障偵測與診斷

Abstract

自21世紀開始，隨著環境保護及永續發展的意識逐漸提升，為減少碳排放量，世界各國都在積極研發對環境衝擊較低的替代能源，其中以再生能源中的太陽能最受重視。太陽能光電系統將太陽照射至地球的光能轉換成電能，擁有穩定且持續的能量來源，無二氧化碳排放且安裝限制因素較少，是個容易推廣至民間的再生能源系統。 太陽能光電系統研究發展至今已超過50年，提高產能方面的研究日趨成熟，除了研發出不同的半導體材料以提高轉換效率外，發電系統中追日機構的設計、逆變器的最大功率點追蹤演算法及太陽能模組的降溫系統都有相當數量的研究論文產出。但欲發展出一個完整且成熟的發電系統，除了產能的提升之外，擁有良好的監測系統並控制其運轉狀況和故障診斷系統也是不可或缺的一環。 在現有的太陽能電廠監測系統中，大多以實際佈線收集基本電力資訊和輻射照度等參數。本研究除了收集上述的基本資訊外，更增加了環境參數感測器，結合多種感測資料來完善對太陽能電廠的監控。本研究所提出的故障診斷機制，便是依靠分析多種感測參數及發電結果，以達成即時並準確地鎖定發生故障的變流器位置及其故障原因。電廠業者能依此機制迅速地通知工程師前來維修，大幅降低維修所需的作業時間及電廠的營業損失。

Since the 21st century, with the rising attention on environmental protection and sustainable development, most governments have actively participated in finding alternative energy, which generates fewer environmental impacts and carbon emissions. With the advantages of stable and sustainable energy sources, zero carbon dioxide emissions and low geographical limitation, Photovoltaic (PV) systems have become the most valuable asset in all renewable energy sources. The studies on photovoltaic systems have been done over 50 years, so the techniques that improve solar energy generation have become mature. Except for studies on different semiconductor materials, sun lotus tracking devices, maximum power point tracking (MPPT) methods used by power inverters, and cooling devices for PV modules are particularly popular research issues. However, to develop a complete and mature solar power generation system, it is necessary not only to focus on power generation method improvement but also to develop a monitoring device which can collect important PV system data while performing fault diagnosis tasks. Most of the recent PV monitoring systems only collect basic electrical and irradiation parameters. In this research, in addition to the parameters mentioned above, environmental sensors are used to collect environmental parameters. By analyzing electrical, irradiation and environmental parameters, this research proposes a reliable fault diagnosis method for PV systems. This method can accurately target the location of a faulty inverter and the reason causing the fault immediately. By using the proposed method, the industry can rapidly inform engineers to repair their devices so the time required on finding and excluding the fault can be reduced. The economic losses associated with the fault can also be decreased.