The demand for chiller equipment that keeps each machine at a constant temperature to maintain the best possible condition is growing rapidly. PID (Proportional Integral Derivation) control is a popular temperature control method. The error, which is the difference between the desired target value and the current system output value, is calculated and used as an input to the system using a proportional, integrator, and differentiator. Through such a closed-loop configuration, a desired final output value of the system can be reached using only the target value and the feedback signal. Furthermore, various temperature control methods have been devised as the control performance of various high-performance equipment becomes important. Therefore, it is necessary to design for accurate data-driven temperature control for chiller equipment. In this research, support vector regression is applied to the classic PID control for accurate temperature control. Simulated annealing is applied to find optimal PID parameters. The results of the proposed control method show fast and effective control performance for chiller equipment.

1. 서 론
2. 제어기 설계를 위한 접근법
    2.1 고전적인 PID 제어 기법
    2.2 Support Vector Regression 기법
    2.3 Simulated Annealing의 최적화 기법
3. 제안된 SA, SVR이 적용된 PID 제어
    3.1 PID-SVR 제어
    3.2 SA를 이용한 PID-SVR 제어의 최적화 목적식
4. 시뮬레이션을 통한 실험
    4.1 초기 파라미터 선정 및 검증
    4.2 PID-SVR과 PID 제어 성능 비교 실험
    4.3 SA-PID-SVR 제어와 타 제어 비교 실험
5. 결론 및 향후 연구 방향
References

• Young-shin Park(공주대학교 산업시스템공학과) | 박영신
• Dongju Lee(공주대학교 산업시스템공학과) | 이동주 Corresponding author