蜂群算法在太阳电池寿命预测参数辨识中的应用  

APPLICATION OF ARTIFICIAL BEE COLONY ALGORITHM IN PARAMETER IDENTIFICATION OF SOLAR CELL LIFE PREDICTION

作　　者：简献忠[1] 武杰[1] 郭强[2] Jian Xianzhong;Wu Jie;Guo Qiang(Ministry of Education and Shanghai Municipal Key Lab of Modern Optical System,School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;National Satellite Meteorological Center,Beijing 100081,China)

机构地区：[1]上海理工大学光电与计算机工程学院教育部及上海市现代光学系统重点实验室,上海200093 [2]国家卫星气象中心,北京100081

出　　处：《太阳能学报》2018年第12期3392-3398,共7页Acta Energiae Solaris Sinica

基　　金：国家自然科学基金(41075019)

摘　　要：为解决太阳电池寿命预测模型参数辨识中参数辨识精度低的问题,提出采用人工蜂群算法进行太阳电池寿命预测模型参数辨识的方法。利用人工蜂群算法的局部快速搜索能力和高效全局收敛性能,对太阳电池寿命预测的电流衰减模型的五参数进行辨识,给定失效阈值利用电流衰减模型进行最大寿命预测。运用人工蜂群算法和最小二乘法辨识的均方根误差RMSE分别为2.858×10^-4和1.337×10^-3,R^2分别为0.9228和0.8666,实验分析表明:人工蜂群算法求得的均方根误差、误差平方和与R^2明显优于最小二乘法,为太阳电池寿命预测的电流衰减模型参数辨识提供一种新的思路。In order to solve the problem of low parameter identification accuracy in the solar cell life prediction model,a method using artificial bee colony algorithm for parameter identification of solar cell life prediction model is proposed. Using the local fast search capability and efficient global convergence performance of the artificial bee colony algorithm, the five parameters of the current decay model for solar cell life prediction are identified,and the maximum lifetime prediction is performed using the current decay model for a given failure threshold.The root mean square error RMSE identified by the artificial bee colony algorithm and the least squares method are 2.858×10^-4 and 1.337×10^-3, respectively,and R^2 is 0.9228 and 0.8666,respectively.The experimental analysis shows that the root mean square error,squared error sum and R2 obtained by the artificial bee colony algorithm are obviously better than the least squares method,which provides a new idea for the parameter identification of the current decay model for solar cell life prediction.

关 键 词：太阳电池 参数辨识 预测分析 电池寿命 人工蜂群算法 

分 类 号：TM914.4[电气工程—电力电子与电力传动] TP18[自动化与计算机技术—控制理论与控制工程]