| Summary: | 博士 === 國立交通大學 === 光電工程學系 === 100 === With the rapid progress in light-emitting diode techniques, all kinds of lighting purposes can be achieved by strategically manipulating the spectral power distributions of LEDs clusters. For example, in general lighting there is a fundamental tradeoff between the efficiency and color rendering quality, where an optimal boundary (the Pareto Front) will be produced. By optimizing the composite spectrum, the LEDs lighting system can be operated alongside the boundary. When the spectral-controllable technique is utilized to fields of intelligent illumination, the color temperature can be adjusted in accordance with end demands. As to the medical lighting, the spectral distortion will be predictable with respect to variant ambient temperature and drive current.
In this dissertation, a novel methodology for spectral manipulation has been proposed, including a well-defined spectral model and six optimization steps. The spectral model employing the double-Gaussian function can closely estimate the practical spectrum that is imperfectly symmetric and depends on the junction temperature. For the optimization, the concept in imaging system design has successfully been adopted to develop a composite spectral process. The proposed algorithm would be applied to the low power and high power cluster, respectively, to achieve the color-turntable systems with high efficiency, high color rendering property as well as wide operation window at ambient temperature.
|
|---|
