遗传算法在LED光源光谱匹配技术中的应用

The Genetic Algorithm in the Application of the LED Light Source Spectral Matching Technology

为利用单色LED实现CIE-D65标准光源、AM1.5标准太阳光目标光谱的匹配,提出简单遗传算法作为光谱匹配算法,通过求解超定方程组的非负最小二乘解,优化不同峰值波长、峰值波长间隔、半高宽的单色LED匹配光源组合,达到光源光谱匹配的目的.仿真实验表明:该算法拟合的相关指数达0.99以上,模拟光谱与目标光谱基本吻合;匹配光源的峰值波长间隔越小,光谱拟合性越好,但结合工程应用需要,峰值波长要选取适中;所提遗传算法运行速度快、效率高、拟合误差小,可广泛应用于植物照明、医疗照明等特殊照明领域的光谱匹配仿真试验和工程实践.

To realize the target spectrum matching of CIE-D65 standard illuminant and AM1.5 stanaara solar spectrum by employing monochromatic LED, a simple genetic algorithm as the spectral matching algorithm was proposed, by solving non-negative least square solution of overdetermined equations, the combination of monochromatic LED light source with different peak wavelength, peak wavelength interval and full width at half maximum was optimized, and the light source spectral matching was achieved. Simulation results show that the fitting correlation index is above 0.99, the fitting spectrum is consistent basically with the target spectrum. The smaller the interval of spectral matching peak wavelength is, the better the spectral fitting effect is, to meet the needs of engineering application, the peak wavelength needs to be selected moderately. The proposed algorithm runs fast, has high efficiency and small fitting error, and it can be widely used in the spectral matching simulation test and engineering practice for special lighting fields such as plants lighting, medical lighting and so on.