Thermal-aware Flow Field Optimization for Energy Saving of Data Centers

Abstract

Improving the energy efficiency of data centers has become an essential problem. Flow field optimization has emerged in recent years as a critical part of thermal management. A novel temperature prediction model based on artificial neural network for various flow field is presented in this paper. The model approximates the solution given by Computational Fluid Dynamics methods, and it is used to formulate a constrained nonlinear optimization problem with the objective of minimizing the energy consumption of the cooling system. The particle swarm optimization algorithm is employed to optimize the flow field in the data center level. The optimizer tunes the flow rates of individual racks based on the predictions of the temperatures of the racks proactively. Simulation results demonstrate that over-provisioning of cooling air is eliminated by the flow field optimization. The efficiency of the cooling system is improved by 17% compared to the traditional temperature-based feedback controller.