Plant 공정 제어를 위한 Support Vector Machine을 이용한 제어기의 구현

PID controller is known as an efficient one whose operation and performance is good and widely used for long period of time in various fields. The PID control logic is indispensable in running real plant and is widely included in DCS (Distributed Control System). Even if for the system of same structure, a system performance is well known to be highly dependent on the gain of PID controller. Accordingly in fields, lots of manpower, time, money and efforts are investigated in tuning the PID gain. In the study a conventional PID controller used in general plant operation is improved in conjunction with SVR (Support Vector Regression) and is proposed for better performance and stability. The proposed controller is much improved in terms of response speed and its stability. Simulation Is conducted in regards of POS (Percent over Shoot), ST (Settling Time) and ITAE (Integral of Time-multiplied Absolute value of Error) for an operation of plant system using Simulink.