Design a LED Modulation Lighting System Using Multi-Objective Genetic Algorithm

• Main Authors: Yen-Chun Hsu, 許延駿
• Other Authors: Rong-Fong Fung
• Format: Others
• Language: en_US
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/53275713990250209374

碩士 === 國立高雄第一科技大學 === 光電工程研究所 === 98 === In this paper, a real-coded genetic algorithm (RGA) technique is developed for the optimal design of a supplemental lighting system for plant-growth, which is affected by the parameters: wavelength, frequency, duty ratio and amplitude. A light intensity stimulation system of functional is designed that can provide a flexible interface to adjust the parameters for users and researchers. Here, the technique of modulation and synchronizing in multi-wavelength light emitting diodes (LEDs) are employed. Through operations of the personal computer, microprocessor and related interface circuits, the system allows the light quality, light intensity, frequency (0 ~ 10 kHz), duty ratio (0 ~ 100%) and phase shift (0 ~ 360o) to be controlled. The optimal design in height and current of the plant-growth lamp by the RGA shows the improvement in the light uniformity and illumination efficiency. The innovatively designed systems are compared favorably with the typical and expert-designed lighting systems. It is concluded that the system is satisfactorily suitable to the photosynthesis applications. In multi-objective optimization, a describing genetic algorithm (GA) specifically developed for problems with multiple objectives. Control the input voltage distance of the RGB LEDs, the optimal illumination and uniformity can be found with by multi-objective genetic algorithms (MOGA). They primarily differ from traditional GA by using specialized fitness functions and introducing methods to promote solution diversity.