| Summary: | 碩士 === 國立中正大學 === 會計與資訊碩士在職專班 === 107 === In the semiconductor industry, yield is essential for controlling and promoting the equipment’s stability. Poor quality in the final cut crystalline grain may influence the final product if the stability during wafer production is inadequate. In this study, to build a monitoring mechanism to detect abnormal machine quality, we use analysis of variance, principal component analysis, and cluster grouping. For each batch of operations, we analyze the reasons for the differences and quickly identify the possible issues by analyzing data rather than relying on human judgment. This approach leads to a shorter analysis time and increased control over key projects, while ensuring adjustment of on-site personnel, increase in the stability of machines, and ensuring better production quality such that the yield can be increased. Currently, the requirements for equipment quality are mostly set in a limited conditional range; however, based on this study, we will be able to initiate equipment-related improvements and control the equipment quality in a more precise range.
|
|---|
