Design a GA-based PID Controller for a Pressure Control of a Process Chamber using a Time-domain System Identification

Among numerous process conditions, e.g. pressure, temperature, and RF, the pressure is one of the essential control parameters to influence on semiconductor manufacturing because its change affects the etch rate, selectivity, aspect ratio, and so on. Therefore, it is important to delicately and consistently control several reference pressures in process recipe. In this paper, we model a system of a process chamber as a transfer function using a time-domain system identification method. After then, a scheduled proportional-integral-differential (PID) controller is designed to regulate a series of reference pressure. To this end, we employ gain sets optimized by Genetic Algorithm (GA) considering control performance indices, i.e. overshoot, rise time, and settling time. Through our experimental results, we confirm that our control scheme shows appreciable improvement compared with a current pressure controller applied in mass production line.