菲涅爾透鏡自動化設計技術;Automated Design Technology of Fresnel Lens

NCU Institutional Repository > 理學院 > 光電科學研究所 > 博碩士論文 > Item 987654321/97714

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: https://ir.lib.ncu.edu.tw/handle/987654321/97714

题名:	菲涅爾透鏡自動化設計技術;Automated Design Technology of Fresnel Lens
作者:	江俊毅;JIANG, JUN-YI
贡献者:	光電科學與工程學系
关键词:	自動優化;Zemax;人工智能運算;太陽能;Optimization;Zemax;artificial intelligence computing;solar energy
日期:	2025-07-28
上传时间:	2025-10-17 11:49:38 (UTC+8)
出版者:	國立中央大學
摘要:	面對如今全球溫室議題與環境污染的挑戰，綠色能源的需求持續攀升，且至 2023 年止台灣再生能源發電占全國發電量約 9.47 %，其中太陽光電 4.58 %占比最高，成為台灣主要再生能源選擇。然而，太陽能模組使用期限通常不超過 25 年，退役後由於回收成本高昂，僅約 10 % 的面板被回收，其餘多數被掩埋，導致含鉛、鎘等重金屬的電子廢棄物釋出污染土壤與地下水。根據台灣環保署預估，2023 年台灣產生廢棄太陽能板約 1 萬公噸，2035 年起每年將超過 10 萬公噸，而使用特殊設計之菲涅爾光學透鏡系統結合舊有太陽能板可以有效增強發電效率。本研究針對的是開發一套多迴圈迭代的自動化光學透鏡設計流程：先以隨機取樣及模擬獲得初步參數，再結合機器學習演算法模型篩選優化組合，提高發電效率達1.65倍以及太陽能板利用率的最大；達到僅需輸入太陽能板至透鏡距離，即可透過Zemax Macro 自動運算出最佳透鏡參數，並自動執行光追跡模擬分析，降低非專業使用者門檻，本研究所建構之多迴圈自動化優化框架，不僅可以使用在此菲涅爾透鏡模組亦可使用在 ii 更為複雜的光學結構設計。;In the face of global challenges related to climate change and environmental pollution, the demand for green energy continues to rise. As of 2023, renewable energy accounts for approximately 9.47% of Taiwan′s total power generation, with solar photovoltaic (PV) contributing the largest share at 4.58%, making it the primary source of renewable energy in Taiwan. However, solar panels typically have a lifespan of no more than 25 years. Due to the high cost of recycling, only around 10% of retired panels are actually recycled, while the majority are landfilled, releasing hazardous heavy metals such as lead and cadmium into the soil and groundwater. According to estimates from Taiwan’s Environmental Protection Administration, about 10,000 tons of waste solar panels were generated in 2023, and this amount is expected to exceed 100,000 tons annually starting in 2035. Utilizing specially designed Fresnel lens systems in combination with existing solar panels offers a promising solution to enhance power generation efficiency. This study focuses on developing a multi-loop iterative automated design process for optical lenses. Initial parameters are generated through random sampling and simulation, followed by optimization using machine learning algorithms to identify the most efficient combinations. This process improves power generation efficiency by up to 1.65 times and maximizes solar panel utilization. By simply inputting the distance between the solar panel and the lens, the system can automatically calculate optimal lens parameters through Zemax macros and conduct ray-tracing simulations. This reduces the entry barrier for non-expert users. The proposed multi-loop automated optimization framework is not only applicable to the Fresnel lens module but also adaptable to more complex optical system designs.
显示于类别:	[光電科學研究所] 博碩士論文

文件中的档案:

档案	描述	大小	格式	浏览次数
index.html		0Kb	HTML	18	检视/开启

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....