Online System Identification for the Real Time Control of the Plasma Parameters

Abstract In recent years, the development of AI has been accelerated so that the semiconductor manufacturing process has also been regarded as a very important issue. Many research results related to the plasma etch process which is one of the most important processes in the semiconductor manufacturing process have been presented in various forms. Recently, our research group has also presented a case of successful real-time multivariable control of plasma parameters based on optical emission spectroscopy. Although the above results are expected to bring about positive effects in various fields, they do not consider the variability of the system. Plasma related systems are so sensitive that online system identification is inevitable. In this paper, we performed the online system identification of the system composed of the plasma parameter through recursive parameter estimation method. Through the recursive algorithm using forgetting factor and input and output information which was gathered in real-time, we detected that the system was changing in real-time and updated the model parameters. We expect the results introduced in this paper contributes not only to the semiconductor manufacturing process but also to all the processes using plasma.