| Summary: | 碩士 === 國立高雄應用科技大學 === 應用工程科學研究所 === 99 === Genetic algorithms use the biological evolution principle for global optimization search. In recent years, genetic algorithms have extensively applied to solve various industry problems, especially for large and complex discrete problem can be easily and effectively to the global optimum.
The high bright LED (Light Emitting Diode) is applied to traffic signals such as red, and green light as traffic indicating light, the lighting device uses a LED array in intensive arrangement. The produced heat is not dispensed easily. Besides, the heat is usually gather in the middle area of the system, resulting in overheat and furthermore affecting the light efficiency of LED. This study uses of genetic algorithm to arrange 198 pieces of LED in parameter optimization and it is simulated in a natural convection circumstance with the same LED quantity and fixed lamp surface area. By changing the different arrangements, it explores the temperature distribution situation of the whole system. The result of optimized analysis indicates it achieves an optimial LED distribution position and it reduces the average temperature of the whole system by 6.8℃. Furthermore compare the 500th generation and its heat analysis and take 20 mesh temperatures of the same area. After optimized analysis, the temperatures are all lower than the uniform arrangement by space.
|
|---|
