本論文由分析光譜擬合特性參數並優化，得到不同色溫下最佳演色性之光譜。從實驗室以往累積的光譜資料與模型為基礎，藉由擬合光譜，找出特性參數之關係趨勢與限制範圍，並藉由基因演算法得到不同色溫下最佳演色性之光譜組成。利用此種方式，將特性參數條件增加，能找到不同LED更真實存在的最佳演色性光譜。;In this thesis, the LED spectrum with optimal chromatic characteristics has been obtained for various correlated color temperature. First, it is found that the general LED spectrum can be represented by 10 featured parameters. As fitting the spectrum to the LED spectral model, all the 10 featured parameters are easily to extract. Then, the possible range for these 10 featured parameters is discussed. Finally, the optimal LED spectra for every CCT is obtained by the global optimization algorithm.
