摘要
为了有效降低胆红素脑病的风险和减少换血的需要,以实验测试的标准胆红素溶液最有效的吸收光谱作为光疗的目标光谱,根据光谱构造理论,采用不同峰值波长和半高宽的单色LED作为匹配光源,首次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,得到最优LED组合比例,进而推得所需各种峰值波长的LED的数量。仿真实验结果表明:拟合的光谱曲线与目标光谱曲线十分近似,光谱匹配度达到了98.39%,获得了一种新型的LED黄疸治疗仪光源光谱功率分布。该算法简单,效率高、拟合误差小,可应用基于LED光源的光谱功率分布匹配技术的研究。
To decrease the risk of bilirubin encephalopathy and minimize the need for exchange transfusion ef -fectively,the most efficient absorption spectrum of standard bilirubin tested by the experiment was regarded as the target spectrum .On the basis of spectral constructing theory , monochromatic light-emitting diodes （ LEDs） with different peak wavelength and full width at half maximum （ FWHM） were used as matching sources .The simple genetic algorithm（SGA） was first proposed as the spectral matching method in this study .The optimal combination ratios of LEDs were obtained by calculating the non-negative least square of the overdetermined e-quations .The required LED number at each peak wavelength was then calculated .Simulation experimental re-sults show that the fitting spectrum is very comparable to the target spectrum with 98.39%matching degree , and a new spectral power distribution of LED-based jaundice therapeutic device was obtained finally .With simplicity, high efficiency , and small fitting error , this algorithm is very suitable for the study of LED-based spectral power distribution matching technology .
出处
《中国光学》
EI
CAS
2014年第5期794-800,共7页
Chinese Optics
基金
福建省科技计划重点资助项目(No.2013I0004)
关键词
LED
黄疸治疗仪
简单遗传算法
光谱功率分布
光谱匹配
light-emitting diode jaundice therapeutic device simple genetic algorithm spectral power distri-bution spectral matching
