多波段發光二極體之優化演算法於智慧型照明應用

Abstract

隨著節能意識抬頭，具有環保節能與可調性等優點的白色發光二極體逐漸成為市場的新型光源。本論文藉由LED混光平台的方式以啟發式演算法(模擬退火法與基因演算法)優化設計照明頻譜，多目標函數的參數包括評價演色性的CQS，計算LED效率的 Luminous Efficacy (輻射光視效能)以及Luminous Efficiency (發光效率)等，在設定的色溫以及限制條件下用演算法計算多目標函數找尋全域最佳解，最後得到擁有最佳的照明評價參數下的頻譜分布。

實驗驗證為利用實驗室擁有的LED陣列平台，並用頻譜儀等實驗儀器測量LED陣列平台的光譜並與優化的光譜做比較，未來希望引入產業上解決組合不同單色光LED達已定目標色溫的問題，例如智慧型照明應用以及業界上LED binning 的問題，期以增加LED的有效利用之外也降低LED的成本。

With the energy-saving awareness, white light LED has the advantage of energy saving and adjustable spectrum to serve as a new light source for general illumination. In the thesis, we optimized the lighting spectrum by the way of additive mixing approach via multi-color LEDs platform with heuristic algorithm, which including the simulated annealing algorithm and the genetic algorithm. The lighting parameters of metric function include Color Quality Scale, Luminous Efficacy and Luminous Efficiency. Under a designated correlated color temperature and lighting parameters condition, we searched for the global maximum solution of the metric function and found out the available spectrum distribution. The simulation was also verified by an off-the-shelf LED matrix. This platform associated with the algorithm program and LED matrix will be expected to have impact on the LED binning problem and intelligence lighting applications in the future.