Data Center Energy Consumption Simulation and Prediction from the Perspective of Digital Twin

In recent years, the Chinese data center industry has experienced rapid growth, yet it faces challenges of high energy consumption and low resource utilization. To achieve intelligent management, the digital twin technology has received widespread attention. This paper utilizes digital twin technology and leverages Unity3D to construct a simulation platform for the data center. This platform is capable of receiving temperature field information, displaying three-dimensional temperature cloud maps visually to evaluate the energy efficiency of data centers. Additionally, we design a deep neural network to predict the device energy consumption and calculate energy efficiency. Our research method helps management personnel gain comprehensive insights into the real-time status of data centers and make effective adjustments, thereby promoting the green and high-quality development of data centers.