A data center thermodynamic dynamic modeling method based on gray box modeling

Abstract

The present application belongs to the technical field of data center energy saving optimization operation, and relates to a data center thermodynamic dynamic modeling method based on gray box modeling, which comprises the following four parts: (1) data center heat capacity thermal resistance model establishment; (2) adaptive noise introduction; (3) data center thermodynamic state optimal estimation; (4) heat capacity thermal resistance parameter identification. The present application establishes a data center thermodynamic dynamic model by identifying the data center heat capacity thermal resistance parameters and introducing adaptive noise into the heat capacity thermal resistance model describing the heat flow of the data center, which is helpful for quantitatively analyzing the heat flow between various components of the data center and can provide a basis for the optimized operation of the refrigeration system.

Images