基于灰色预测的光伏组件阴影类型诊断研究

A Gray Prediction Based Diagnosis on Shadow Types Over Photovoltaic Modules

针对光伏组件阴影类型难以判别的情况,提出了基于灰色预测的光伏组件软硬性阴影区别方法。首先通过小波理论分析灰色预测应该达到的前提条件。然后对灰色预测GM(1,1)进行改进,提出了使用新陈代谢GM(1,1)对光伏组件功率进行预测。最后对预测值与实测值进行误差分析,根据得出模型精度差异来判别阴影性质。通过仿真和实验证明了软性阴影和硬性阴影在灰色模型预测精度等级有明显的差异,可以通过精度等级判断阴影类型。该方法能有效判定阴影性质,为积灰程度判定与光伏热斑检测提供了有力的依据。

In allusion to the fact that it is hard to distinguish the shadow types over photovoltaic （Pv） modules, a gray prediction based method to distinguish the soft shadow from the hard shadow over PV modules is proposed. Firstly, the precondition that the gray prediction has to achieve is analyzed by wavelet theory; then the gray prediction GM（1,1） is improved and it is proposed to predict the power output of PV modules by metabolic GM（1,1）; finally, the error analysis on both predicted results and measured results is performed, and based on the obtained model accuracy difference the nature of the shadow can be judged. It is proved by simulation and experiments that as for the soft shadow and the hard shadow there are obvious differences in the accuracy classes predicted by gray model, thus it is possible to judge the types of the shadow by accuracy classes. Using the proposed method the nature of the shadow can be effectively judged, thus a solid foundation is provided for the judgment of the dust accumulation and the detection of hot Spots.