Data analytic approach for calibration of online photovoltaic cell metrology

光伏（PV）电池用于利用太阳能并将其转化为电能，晶体硅光伏电池是目前使用的主要技术。由于这些电池是在高通量过程中生产的，因此实时完成光伏电池的质量控制是具有挑战性的。然而，有一些方法可用于电池的离线测试，用四点探头测量薄片电阻是最常见的测量技术。Aurora Solar开发了一种基于红外反射的在线（即实时）传感器，用于监控光伏电池在生产过程中的质量。然而，处理条件的变化会影响测量结果及其与离线测量结果的比较。因此，本项目的目的是提高内联传感器的鲁棒性，并对其进行校准，使其能够在所有可能的工艺条件下获得等效的离线薄片电阻测量。先进的数据分析技术将应用于工业获得的数据，以开发一个强大的校准算法。

Photovoltaic (PV) cells are used to harness solar energy and convert it to electricity, and crystalline silicon PV cells are the major technology being used. Quality control of the PV cells being produced is challenging to accomplish in real time, since these cells are produced in a high-throughput process. However, there are some methods available for offline testing of the cells, with measurement of the sheet resistance with the four point probe being the most common measurement technique.Aurora Solar has developed an inline (i.e. real time) sensor based on infrared reflectometry for monitoring quality of the PV cells during production. However, variations in processing conditions can impact the measurements and their comparison with offline measurements. Therefore, the aim of this project is to improve robustness of the inline sensor and calibrate it to be able to obtain the equivalent offline sheet resistance measurement under all possible process conditions. Advanced data analysis techniques will be applied on industrially obtained data in order to develop a robust calibration algorithm.