Takeo Kasajima

Development of a cooling control system for data centers utilizing indirect fresh air based on Model Predictive Control

In this paper, a cooling control system based on Model Predictive Control is developed in order to reduce energy consumption of the facility equipment for the modular data center utilizing indirect fresh air, while considering the relationship between energy savings and temperature management of servers. Furthermore, the developed cooling control system is applied to the actual modular data center utilizing indirect fresh air, and its practical effectiveness is verified by conducting a demonstration experiment. To develop the proposed control system, a dynamic mathematical model for the actual modular data center is designed, and the proposed control system is mounted on an existing management server. The control system is able to operate without using a separate hardware controller. As a result, it is revealed that the proposed control system reduces facilitys total power consumption by about 49% compared to a conventional system under condition where the supply air flow rate is 20°C.

Modeling of HVAC system in data center with super-multipoint temperature sensing technology

In this paper, we consider a server cooling system in a data center with fans and a super-multipoint temperature sensing technology using optimal time-domain reflectometry of optical fiber. The sensing system was developed to visualize the temperature distribution of the room in real time, and it can also be a key technology to control the distribution in order to reduce total power consumption in data centers. In this paper, we first present a concept of the fan control system. The objective of the system is to uniformize the rack air inlet temperature distribution in the presence of heat from each server. Since the control scheme requires a dynamical model of the temperature variations in the room, this paper mainly addresses the modeling. The challenge of the modeling stems from a large volume of data provided by the sensing system. It is computationally hard to directly apply standard identification techniques to the data. We thus present a two-stage reduction scheme of the output dimension using the concept of so-called mutual information. The effectiveness of the proposed scheme is finally demonstrated using real data.

Development of cooling control technology by just-in-time modeling for power reduction in large-scale data center

This study involves the development of a cooling control system utilizing the Just-In-Time (JIT) modeling for the more efficient operation of large-scale data center facilities. In addition, we propose how to control multiple air-conditioning units in order to minimize total power consumption in data center utilizing these predicted data. Data on temperature, humidity and electric power in large-scale data centers were used to generate predicted temperature and humidity for an hour later by JIT modeling. We also estimated the power consumption for air-conditioning units in data center over two months. As a result, it is clear the trend of predicted data is very similar to that of measured one. The accuracy of predicted data utilizing JIT modeling is sufficiently high from the results of calculated correlation coefficient and root mean squared error. The total power consumption for air-conditioning units utilizing proposed control system is found in 6.0% reduction compared with the actually measured data. Since these results were obtained under summer conditions that were the most severe in terms of use of fresh air, further benefits would be expected if a similar study were carried out in other seasons.