用于薄膜光伏电池测量的高光谱系统研究

薄膜光伏电池的性能与其空间物理特性分布密切关联，因此空间分辨的二维光谱成为分析电池不均匀性的重要手段。起源于遥感领域的高光谱技术，近年来被应用于薄膜电池的测量，并体现出快速、无损、及光照均匀等优势。若能按测量需求进一步改进系统，将更有利于制定提升器件性能与稳定性的优化方案。因此，本项目拟首先研究高光谱与光浴老化系统的实现与集成，以探索薄膜光伏电池的原位老化与观测方案；其次，将后处理算法应用于薄膜电池的高光谱数据，以筛选、优化适用的超分辨率算法，并明确算法对图像质量的改善程度；最后，借助计算机视觉技术，研究提取高光谱图像特征的自动化算法，以减少检测过程中的人工工作量。本研究旨在拓展薄膜光伏电池的原位老化分析手段，并为研究者提供快速而细节突出的空间分辨光谱图像。

Performance of thin film photovoltaic cells is closely related to their spatial physical distribution characteristics, so the spatially resolved two-dimensional spectrum becomes an important approach for analyzing the nonuniform characteristics. The hyperspectral technology which was originated in the field of remote sensing has been applied to the measurement of thin film batteries recently, and shows the advantages of fast, non-destructive, and uniform illumination during measurement. If the system can be further enhanced based on measurement requirements, it will be more conducive to develop optimization schemes for devices to improve their performance and stability. Therefore, the project firstly intends to study the implementation and integration of hyperspectral and light soak aging systems in order to explore in-situ aging and observation schemes for thin-film photovoltaic cells; Secondly, post-processing algorithms will be applied to hyperspectral data of the thin-film cells for selecting and optimizing corresponding super-resolution algorithms, and furtherly clarify how much image quality could be improved through algorithmic approaches; Finally, with computer vision technologies, automatic algorithms for hyperspectral image feature extraction will be studied to reduce manual workload in measurement process. The study aims to expand the in-situ aging analysis approaches for thin-film photovoltaic cells, and provide fast and spatially resolved spectral images with distinct details for researchers.