Development of virtual metrology algorithm for the high mixed manufacturing process

With the development of manufacturing technology, production resembles an automated assembly line in which many similar products with different specifications are manufactured. This production mode is named as a high-mix manufacturing process. The traditional virtual metrology method can not satisfy the demand of the manufacturing process. In order to get the real time measurement for the high-mix manufacturing process, a new virtual metrology algorithm which combines the advantage of both standard statistical methods and time serial analysis is proposed. Firstly, key variables are selected, which contain important information regarding the source of variation of the current process condition. Then, the statistical MANCOVA technique is adopted to build conjecture model for products with different specifications considering the product effects. Thirdly, the time serial, i.e. the IMA(1,1), is integrated into virtual metrology approach, which reflects the current unmeasured information of the process, and thus the accuracy of the virtual metrology model is substantially improved. Applications on the wet etching process evaluate the proposed method.