가변속 냉동시스템의 강인제어를 위한 입자 군집 최적화 알고리즘을 갖는 H∞ 제어기 설계

Abstract

A variable speed refrigeration system(VSRS) is widely used in fields requiring high-precision temperature control, such as precision machine tools because of its excellent ability to respond to partial load and energy-saving performance. The system uses a variable-speed compressor to control the target temperature by adjusting the refrigerant’s mass flow rate. In addition, superheat is simultaneously controlled by adjusting the opening angle of the electronic expansion valve(EEV) to prevent coefficient-of-performance(COP) decrease due to superheated vapor compression and liquid back caused by rapid variations in the refrigerant’s mass flow rate. However, accurate VSRS temperature control is difficult due to the adverse effects of disturbance, noise, and model uncertainty. The model uncertainty for VSRS is usually caused by the difference between the real model, which is a nonlinear system with dead time, and the mathematical nominal model, which is simplified by linearization and low-dimensionalization processes to design the control system. It is also caused by changes in the surrounding environment or in operating points different from modeling conditions when the nominal model is obtained. The disturbance occurs by unexpected thermal load fluctuations and noise is predominantly caused by the measuring sensor. Therefore, it is essential to design a controller that is robust to the disturbance, noise, and model uncertainty to achieve accurate temperature control of VSRS. In this study, H∞ controller is used for the robust control of VSRS because of its satisfactory feedback characteristics by loop shaping and robust stabilization. H∞ control guarantees robust performance by minimizing the H∞ norm, which is defined as the maximum singular value of the transfer function from exogenous input to the regulated output. Particularly, the H∞ controller was designed to ensure robustness to disturbance, noise, and model uncertainty by loop shaping and small gain theorem via the weighting functions. Therefore, the weighting functions are crucial design parameters for H∞ controller. However, it wastes much time since finding appropriate weighting functions that satisfy the given controller design specifications requires several trials and errors. Even the weighting functions selected through trial and error do not guarantee optimal performance. Thus, optimal weighting functions are selected using the particle swarm optimization(PSO) algorithm to address the aforementioned issues. PSO algorithm is effective in solving nonlinear optimization problems because it is simple, easy to realize, and does not require many parameters or gradient information to optimize. Therefore, in this study, H∞ controller that is robust to disturbance, noise, and model uncertainty is designed by selecting the weighting functions using the PSO algorithm. The proposed H∞ controller was validated using MATLAB /Simulink-based computer simulations and practical experiments using a VSRS-based oil cooler system(OCS) under stepwise disturbances and sensor noise, as well as model uncertainty. Moreover, the effectiveness of the proposed controller was verified by comparing it with the performance of the conventional PI controller, which was tested under the same experimental conditions. Finally, the simulations and experiments result demonstrated that the suggested controller exhibits significant robustness.