摘要
为进一步提高太阳能电池可靠性筛选的准确性, 改进以往1/f噪声筛选方法存在的分析频率单一、偶然因 素大和抗干扰能力差等弊端,提出了一种利用整个低频段（0.25～10 kHz）噪声进行可靠性筛选的方法——基于 Gower相似系数的可靠性筛选方法。利用建立的低频噪声测试系统,测量 同一批次130 片太阳能电池噪声功率谱并按照本方法将这130片太阳能电池分为Ⅰ（可靠性最高）、Ⅱ、Ⅲ（可靠性最低） 3类。将本方法测量结果和原有1/f噪声筛选方法结果进行对比, 通过4片典型太阳能电池样本说 明了本方法更为准确合理。实验结果表明,对相同的130片太阳能电池,1/f噪 声筛选方法分类结果为：Ⅰ 类44片,Ⅱ类50片,Ⅲ类36片；本方法分类结果为：Ⅰ类的有34片,Ⅱ类42片,Ⅲ类54片。相对于 1/f噪声筛选方法,表明本方法提高了太阳能电池可靠性筛选的准确性,适用 于对可靠性要求高的应用领域。
To further improve the accuracy in the reliability screening for solar cells,the previous 1/f noise method is improved whose analysis fre quency is single,accidental factor is big and anti-interference ability is poor.A new method making use of the entire low fr equency noise （0.25-10kHz） in the reliability screening is put forward：the Gower similarity coefficient reliability screening （GSCRS） method.First,we build a system to measure low fre quency noise power spectra of 130solar cells in the same batch.Then,through the low-freq uency measurement system,we divide these 130solar cells into three groups according to GSCRS method：GroupⅠ （the highest reliability）,GroupⅡ and GroupⅢ （the lowest rel iability）.Last,we compare the 1/f noise method′s result with the GSCRS me t hod′s result and illustrate the reason why GSCRS method is more accurate and reasonable through four typical solar cell samples. Experimental results indicate that for the same 130solar cells,the 1/f noise method′s result is：GroupⅠ contains 44solar cells,GroupⅡ contains 50solar c ells,GroupⅢ contains 36solar cells and the GSCRS method′s result is：GroupⅠ contains 34solar cells,Group Ⅱ contains 42solar cells,and GroupⅢ contains 54solar cells.Compared with the 1/ f noise method,the GSCRS method improves the accuracy of the solar cell reliability screening and it is s uitable for high reliability applications.
出处
《光电子．激光》
EI
CAS
CSCD
北大核心
2014年第2期207-212,共6页
Journal of Optoelectronics·Laser
基金
国家自然科学基金(60906034)
吉林省自然科学基金(201115029)资助项目
作者简介
E-mail：zhouqz@jlu．edu.cn周求湛（1974-），男，博士，吉林大学通信工程学院副教授，硕士生导师，主要从事微弱信号检测和半导体器件可靠性方面的研究．
