本論文致力於針對YAG-LED擬合出的精確白光光譜模型，透過參數的最佳化，進而定義出新的數學模型來建立新的標準光源。透過CRI、TM30之???? (Color Fidelity Index)及??g (Color Gamut Score)指標為依據，優化出YAG-LED各個色溫下最佳白光光譜。透過實驗室所收集到的數據的統計結果，訂定出各個參數的範圍區間，在2,500K到25,000K的範圍內，在每一各別特定色溫下，先利用全域的最佳化演算法，快速搜尋接近系統的最佳化參數，再利用局部最佳化演算法進一步精確到最佳化的結果。同時本文也藉由不同的物理與參數範圍限制，以期使最佳化的結果能夠由實際產品之性能調整，得到理想性的效能呈現。;In this thesis, YAG-LED spectrums with optimal chromatic parameters has been obtained for various correlated color temperature by using three indices were CRI, TM30-????????, and TM30-???????? respectively. We found that the YAG-LED can be represented by eight parameters, by optimizing these parameters in using Genetic Algorithm and Levenberg-Marquardt Algorithm, the best YAG-LED spectrum for various CCT is obtained by the global and local optimization algorithms.
