Multivariate Statistical Process Control Applications In Solar Cell Manufacturing

The traditional quality control of product mainly focuses on a single variable or one-dimensional data. It is the basis of product quality control, but whether manufacturers or consumers need multi-dimensional data to assess the product quality in practical work. So we have to study new method of quality control for multi-dimensional data, which is not only the practical need but also the new requirement of the quality control theory. However, the variables of product quality often have interaction with each other which often has impact on the product quality. Considering all factors, it is the innovation of systems engineering that we can control the product quality by small number of variables.Solar cell manufacturing industry in China has a history of more than fifty years. It is significant to improve the overall performance of solar cell by monitoring the manufacturing process.In this paper, multivariate statistical process control (MSPC) is applied on the manufacturing process of solar cell to improve the quality of the solar cell. The main research work includes as following:1. This paper elaborates the history and the present research of the statistical process control (SPC) and proposes the importance of applying SPC. What's more, some relative concepts about the solar cell and its working principle are studied.2. This paper studies the principle of the control chart in SPC and applies it in an example of the resistivity of solar cell. In addition, the limitations of traditional SPC in practical applications are illustrated.3. The main theory of MSPC is introduced. It is a technique that making high dimensional space project into low dimensional space. It includes principal component analysis, principle component regression and Partial Least Squares (PLS). Then this paper introduces the principal and application of MSPC control chart including T2 chart, contribution chart, trajectory chart, score chart and so on. They are very effective management tools for fault recognition in practical production process.4. The characteristics of bath processes are described as well as the techniques for process monitoring. In addition, the implementation steps for process monitoring are elaborated.5. The whole manufacturing process of solar cell is elaborated and one of the key processes of making P-N knot using spreading method is introduced in details. The paper applies the technique of MSPC to the key process. Specifically, the paper establishes the index system, builds TMPCA model and BMPLS model, analyzes the monitoring process and last recognizes the fault. Then the results of the two models are compared and the conclusion is consistent.6. Lastly, the whole article is concluded and the future research areas are highlighted.