基于UKF的分数阶微分融合算法在光伏检测中的应用  

Application of fractionalorder differential fusion algorithm based on UKF in photovoltaic detection

作　　者：左延红 耿国庆 周超 夏仕龙 Zuo Yanhong;Geng Guoqing;Zhou Chao;Xia Shilong(School of Mechanical&Electrical Engineering,Anhui Jianzhu University,Hefei 230031,China)

机构地区：[1]安徽建筑大学机械与电气工程学院,合肥230031

出　　处：《黑龙江科技大学学报》2025年第2期337-343,共7页Journal of Heilongjiang University of Science And Technology

基　　金：国家自然科学基金项目(51874005);安徽省教育厅重点自然科学研究项目(K201004317)。

摘　　要：在光伏电站的运行过程中,光伏电板的温度是一个非常重要的参数,为了实现对光伏电板温度的准确检测,提出了一种基于UKF的分数阶微分算子融合算法,检测光伏电板的温度。通过对多个温度传感器的数据进行无迹卡尔曼滤波处理,消除传感器的噪声误差,利用分数阶微分算子融合不同传感器的数据进行检测。结果表明,与KF算法和UKF算法进行了对比,该算法在减小偏差和提高数据融合精确度方面表现出明显的优势,融合后的传感器数据均方根误差为0.021,平均绝对误差为0.017,明显优于其他算法。During a photovoltaic(PV)power station in run,the temperature of the PV panels is a very important parameter.This paper aims to achieve the accurate temperature detection of the PV panels,and proposes a fractional-order differential operator fusion algorithm based on Unscented Kalman Filter(UKF).The study involves eliminating the inherent noise errors in the sensors by unscented Kalman filtering on the data from multiple temperature sensors;and fusing the measurement data from different sensors by fractional-order differential operator.The experimental results show that this algorithm exhibits significant advantages in reducing the deviation and improving the accuracy of data fusion compared with the Kalman Filter(KF)algorithm and the UKF algorithm.The proposed algorithm is significantly better than the other two algorithms with the root mean square error of the fused sensor data by 0.021,and the mean absolute error by 0.017.

关 键 词：光伏 UKF 分数阶微分算子 噪声误差 数据融合 

分 类 号：TM615[电气工程—电力系统及自动化]