解析法-Nelder Mead单纯形算法确定太阳电池参数

发布时间：2018-07-27 21:11

【摘要】：光伏发电系统的设计计算、性能评估及优化控制要求快速、准确的确定太阳电池模型参数。针对太阳电池单、双二极管模型参数辨识问题,该文提出一种基于解析法和Nelder Mead单纯形法重启策略的A-bcNM混合算法。先用实测I-V曲线上的部分关键点计算合成参数,再以单二极管模型近似解析式快速定位搜索始点,最后在边界范围内利用Nelder Mead单纯形算法的重启策略最小化实测数据与模拟结果之间的均方根误差,以提高拟合精度及参数解的质量。MATLAB环境下,利用2种典型太阳电池的实测数据对A-bcNM算法的有效性进行了测试和验证。与已有的其他算法相比,A-bcNM算法收敛速度快,计算量小,辨识精度高,可快速、准确的确定单、双二极管模型参数,为识别太阳电池的工作特性提供了一种有效方法。
[Abstract]:The design calculation, performance evaluation and optimal control of the photovoltaic power system require fast and accurate determination of the parameters of the solar cell model. Aiming at the problem of parameter identification of the solar cell single and double diode model, a A-bcNM hybrid algorithm based on the analytic method and the Nelder Mead simplex method is proposed in this paper. First, the part of the measured I-V curve is used. At the key point, the synthetic parameters are calculated, and the single diode model is used to locate the beginning of the search quickly. Finally, the root mean square error between the measured data and the simulation results is minimized by the restart strategy of the Nelder Mead simplex algorithm in the boundary range, so as to improve the fitting precision and the quality of the parameter solution in the.MATLAB environment, 2 typical types are used. The measured data of the solar cell are tested and verified for the effectiveness of the A-bcNM algorithm. Compared with other algorithms, the A-bcNM algorithm has fast convergence speed, small calculation and high identification precision. It can quickly and accurately determine the single and double diode model parameters, which provides an effective method for identifying the working characteristics of the solar cell.
【作者单位】：河南农业大学农业部农村可再生能源新材料与装备重点实验室;郑州大学信息工程学院;
【基金】：河南省重点科技攻关项目(102102210154) 教育部博士点专项科研基金项目(20104105110004) 国家自然科学基金资助项目(U1204607)
【分类号】：TM914.4

【参考文献】