Energy Saving Research on Temperature Control in Data Center Server Rooms

Abstract

To tackle the challenge of inefficient traditional control methods and energy wastage caused by temperature disparities within the data center’s server room cooling system, this research initiative initially installed temperature and humidity sensors in each cabinet to gather comprehensive data. Subsequently, a detailed mathematical model of the room’s temperature system was formulated to assess the influence of air supply airflow distribution on the temperature distribution. Following this, through the integration of fuzzy theory with the PID control strategy, a sophisticated fuzzy PID controller was developed. Utilizing Simulink, an in-depth simulation analysis of both PID and fuzzy PID temperature control strategies was executed. The outcomes of the simulations underscored that fuzzy PID control effectively addresses the time delay and stability concerns present in conventional PID temperature control systems. Lastly, empirical validation confirmed that the fuzzy PID control algorithm can proficiently manage the server room’s temperature while concurrently decreasing the intake of cold air, thus promoting energy conservation. This control approach showcases commendable adaptability and robust performance characteristics.