以多目標演化技術實現矽薄膜太陽能電池結構設計最佳化之研究

Abstract

在本碩士論文中，吾人嘗試將多目標最佳化演算法，應用到以數值計算半導體物理方程式的半導體元件模擬器上。兩者的程式連結使用本實驗室早期發展具高度發展彈性的統合式最佳化架構 (UOF)，此UOF架構的目的旨在聯結兩者不同環境工具介面，UOF程式所使用的語法為C++語言，在問題與解法上的程式發展擁有非常高度的編碼自由度。 太陽能為一零排放再生能源，在半導體元件，特別是矽材料太陽能電池，近年來用於將太陽光能轉換為電能。隨著能源需求增加，對於半導體太陽能電池效率要求更高。因此，非晶、多晶以及結晶矽太陽能電池效率的提升研究成為重要的課題之一。 為了計算半導體太陽能電池元件的光、電特性，我們利用數值方式求解耦合的半導體物理方程式，其中包含泊松 (Poisson) 方程式、電子與電洞的電流連續性方程式以及光的傳遞模型來完成。其中太陽能電池的電特性結果包含：短路電流(short-circuit current)、開路電壓(open-circuit voltage)、以及轉換效率(efficiency)，這幾項結果能用來判斷被最佳化的矽薄膜太陽能電池的特性的優劣。利用這些電特性結果來計算多目標演算法所需要的適應性(fitness)分數，適應性分數在演化式演算法中即是對一組元件的參數設計的評分。在此論文的研究工作中，整合到UOF的最佳化程式，吾人是選擇非主控排序基因演算法：NSGA-II (non-dominating sorting genetic algorithm)。 在此架構中，半導體太陽能電池的參數設計範圍於開始前，根據實作上的技術能力加以定義，以多目標演算法產生初始的多組設計，再藉由UOF的資料傳遞方式呼叫TCAD工具進行元件特性模擬，並記錄結果回傳到演算法中，用以產生下一次多組設計的參數子代。 值得注意的是，要被最佳化的變數為矽薄膜太陽能電池元件的摻雜濃度以及各半導體薄膜層的厚度，我們對給定的矽薄膜太陽能電池規格，分別是短路電流、開路電壓，以及轉換效率，能對上述這些變數同時做最佳化。以此研究的薄膜矽太陽能電池為例，由最佳化設計結果能使效率從原始5.68%提升到8.1%，而實驗結果更達到9.6%的效率。此法亦可推廣到其他半導體太陽能電池設計的最佳化。 總之，利用UOF架構結合最佳化方法與半導體模擬器在半導體太陽能電池設計與製造上能有很大幫助。

In this thesis, we implement a numerical semiconductor device simulation-based multi-objective evolutionary algorithm (MOEA) for the characteristic optimization of amorphous silicon thin-film solar cells, based upon a unified optimization framework (UOF). The UOF program design can connect two different tool interfaces, where the UOF connects TCAD and MOEA program composed by C++.

Solar energy is zero-emission and renewable, so semiconductor devices; in particular, silicon-based solar cells, used for converting sunlight to electrical power has been more attractive in recent years. As the energy demand growing, it is necessary for us to decrease the fabrication cost of semiconductor solar cells and improve the energy conversion efficiency (η). Currently, the materials of crystalline silicon (c-Si), polycrystalline silicon (poly-Si), and amorphous silicon (a-Si) have been of great interest for manufacturing of silicon-based solar cells. Most of optimal designs of thin-film Si solar cells for pursuing the higher conversion efficiency are mainly achieved empirically.

To calculate the devices characteristic, a set of coupled solar cell transport equations consisting of the Poisson equation, the electron-hole current continuity equations, and the photo-generation model is solved numerically. Electrical characteristics, the short-circuited current, the open-circuited voltage, and the conversion efficiency are calculated to analyze the properties of the explored solar cells. The aforementioned device simulation results are used to evaluate the fitness score and access the evolutionary quality of designing parameters via the implemented multi-objective evolutionary algorithm: non-dominating sorting genetic algorithm (NSGA-II).

In this UOF structure, the parameter range of solar cell is defined at first. We can include these settings with the initial design solutions from MOEA , then with data expression in UOF, we can use TCAD to make simulation. And, the result will be returned to the algorithm to generate the new design solutions.

Notably, designing parameters including the material and structural parameters, and the doping concentrations are simultaneously optimized for the explored solar cells. In this thesis, we have optimized an armophous silicon solar cell. The optimized design results have improved the efficiency from 5.68% to 8.1%, and result of fabrication can reach 9.6% efficiency.

The simulation-based MOEA methodology presented in this thesis is useful in optimal structure design and manufacturing of semiconductor solar cells.