Intelligent control method of active ventilation floor of data center

Abstract

Provided is an intelligent control method of an active ventilation floor of a data center. According to the intelligent control method, a Markov decision process model is established for the hotspot problem of a data center rack, and three model solving algorithms are provided, comprise the basic intelligent algorithm, the sample value variant intelligent algorithm and the structural variant intelligent algorithm and are used as the core of an active ventilation floor control algorithm. The model is composed of a system state, the behavior, the award and a value function. The solution of the model is that the optimal behavior is selected continuously in a series of system states, so that the system accumulated reward is maximized; the active ventilation floor control algorithm is used forcontinuously exploring and learning the complex relationship between the temperature distribution of an air inlet of the rack and the rotating speed of an active ventilation floor fan, and finally theoptimal PWM signal duty ratio can be generated according to the temperature distribution of the air inlet of the rack, so that the rotating speed of the active ventilation floor fan is adjusted, thetemperature distribution of the air inlet of the rack is uniform, and the hotspot problem of the rack is relieved. Compared with other schemes, the method is higher in universality, easier to deploy and more cost-effective.

Description

technical field [0001] The invention belongs to the technical field of automatic control, in particular to an intelligent control method for an active ventilation floor in a data center. Background technique [0002] Rack hotspots, that is, one or several positions in the data center computer room rack, where the temperature is significantly higher than that of other positions. Excessive temperature will reduce the efficiency of some servers in the data center, thereby reducing its overall power density, and also reduce its reliability, which is obviously contrary to the needs of the data center. [0003] Using global regulation to alleviate or eliminate rack hotspots, such as increasing the power of the air conditioner in the computer room to provide sufficient cooling air, will inevitably cause most rack areas to be in an over-cooled state, causing waste of cooling resources and making the data center always available. Refrigeration energy consumption, which accounts for ...

AI Technical Summary

Problems solved by technology

Therefore, it is generally difficult to describe the thermal energy efficiency and airflow model of a data center with an analytical model

[0006] Most of the existing research on actively ventilated floors is based on performance modeling and evaluation based on measurement or simulation, and there is no research literature on the problem of actively ventilated floor control.

Images