光子学报, 2016, 45 (4): 0423006, 网络出版: 2016-05-11  

基于果蝇算法优化径向基神经网络模型的白光发光二极管可靠性

Study on the Reliability of White LED Using RBF Neural Network Optimization by FOA Algorithm
作者单位
1 华南理工大学 发光材料与器件国家重点实验室,广州 510640
2 广东银禧科技股份有限公司, 广州 东莞 523000
摘要
根据白光发光二极管失效物理机制选取理想因子、结温、色坐标漂移量等参数作为输入量,利用果蝇算法自学习优化标准径向基神经网络基函数宽度,提高输出精度.研究表明,径向基神经网络模型可以成功预测白光发光二极管可靠性衰变趋势,具有较高的稳定性和鲁棒性;利用果蝇算法优化后,预测平均误差成功减少为3.1%,对未来建立以神经网络为基础的发光二极管可靠性预测模型库提供有益帮助.
Abstract
Fruit fly Optimization Algorithm(FOA) and Radial-based Function(RBF) neural network model was proposed for evaluating the reliability of white Light Emitting Diode(LED) chip. The failure factors of white LED such as junction temperature, color coordinate shift were selected to the neural network input. Using fruit fly algorithm to optimization RBF neural network in order to improve the precision of the output. Studies have shown that RBF neural network is successfully predicted the LED reliability decay trend, with high stability and robustness, using fruit fly algorithm to predict average error successfully reduced to 3.1%, benefit to set up reliability prediction model in the future.

黄伟明, 文尚胜, 傅轶. 基于果蝇算法优化径向基神经网络模型的白光发光二极管可靠性[J]. 光子学报, 2016, 45(4): 0423006. HUANG Wei-ming, WEN Shang-sheng, FU Yi. Study on the Reliability of White LED Using RBF Neural Network Optimization by FOA Algorithm[J]. ACTA PHOTONICA SINICA, 2016, 45(4): 0423006.

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!