Data Center Cooling Prediction Using Artificial Neural Network

2007 ◽  
Author(s):  
Saurabh K. Shrivastava ◽  
James W. VanGilder ◽  
Bahgat G. Sammakia
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Data Center ◽  
Cooling Performance ◽  
The Real ◽  
Time Prediction ◽  
Linear Relationships ◽  
The Neural Network ◽  
Data Center Cooling ◽  
Computational Fluid Dynamics Cfd

An analytical approach using artificial intelligence has been developed for assessing the cooling performance of data centers. This paper discusses the use of a Neural Network (NN) model in the real-time prediction of the cooling performance of a cluster of equipment in a data center environment. The NN model is used to predict the Capture Index (CI) [1] as a function of rack power, cooler airflow and physical/geometric arrangement for a cluster located in a simple room environment. The Neural Network is “trained” on thousands of hypothetical but realistic cluster variations for which CI values have been computed using either PDA [2] or full Computational Fluid Dynamics (CFD). The great value of the NN approach lies in its ability to capture the non-linear relationships between input parameters and corresponding capture indices. The accuracy of the NN approach is 3.8% (Root Mean Square Error) for a set of example scenarios discussed here. Because of the real-time nature of the calculations, the NN approach readily facilitates optimization studies. Example cases are discussed which show the integration of the NN approach and a genetic algorithm used for optimization.

Neural Network-Based Prediction and Control of Air Flow in a Data Center

2012 ◽  
Vol 4 (2) ◽  
Author(s):  
Flavio de Lorenzi ◽  
Christof Vömel
Keyword(s):  
Neural Network ◽  
Energy Consumption ◽  
Real Time ◽  
Data Center ◽  
Cooling System ◽  
Air Flow ◽  
Data Center Cooling ◽  
Prediction And Control ◽  
Reduce Energy Consumption ◽  
And Control

As modern data centers continue to grow in power, size, and numbers, there is an urgent need to reduce energy consumption by optimized cooling strategies. In this paper, we present a neural network-based prediction of air flow in a data center that is cooled through perforated floor tiles. With a significantly smaller execution time than computational fluid dynamics, it predicts in real-time server inlet temperatures and can detect whether prevalent air flow cools the servers sufficiently to guarantee safe operation. Combined with a cooling system model, we obtain a temperature and air flow control algorithm that is fast and accurate enough to find an optimal operating point of the data center cooling system in real-time. We also demonstrate the performance of our algorithm on a reference data center and show that energy consumption can be reduced by up to 30%.

Neural-network-based real-time trajectory replanning for Mars entry guidance

2016 ◽  
Vol 231 (14) ◽  
pp. 2634-2645
Author(s):  
X Zheng ◽  
H Huang ◽  
W Li
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Trajectory Optimization ◽  
Reference Trajectory ◽  
The Real ◽  
The Neural Network ◽  
Entry Guidance ◽  
Computational Speed ◽  
Uncertain Factor ◽  
Predictor Corrector

The real-time trajectory replanning method which is used for the guidance of Mars entry is investigated in this paper. Comparing with the traditional Mars entry guidance methods, such as the reference-trajectory tracking guidance and predictor–corrector guidance, the real-time trajectory replanning method can increase the reliability of the mission remarkably. When faults occur during the Mars entry phase, a replacement trajectory will be planned quickly. Due to the limited onboard computing capacity, replanning the trajectory onboard is a challenging task. Corresponding to this problem, the neural network is trained to approximate the dynamics of the atmospheric entry. The uncertain factor of the atmospheric density is also included in the neural network. Then, by using the characters of the neural network, the analytical expressions of the Jacobian which are needed in trajectory optimization are derived. Finally, an estimation-replanning guidance procedure is introduced. The numerical simulation shows that the proposed guidance strategy can decrease the error of final states effectively, and the neural network approximation improves the computational speed of the nonlinear programming solver remarkably, which makes the method more suitable for use onboard.

Study of CFD-Based Raised-Floor Data Center Cooling With Parametric CRAC Turbofan Blower Airflow Patterns

2021 ◽  
Author(s):  
Zhihang Song ◽  
Wan Chen
Keyword(s):  
Thermal Management ◽  
Data Center ◽  
Data Centers ◽  
Air Cooling ◽  
Flow Conditions ◽  
Cooling Effectiveness ◽  
Cooling Performance ◽  
Data Center Cooling ◽  
Computational Fluid Dynamics Cfd ◽  
Tile Flow

Abstract Commonly encountered thermal management challenges of today’s rapidly changing power density, raised-floor hot/cold aisle data centers include typically uncontrollable tile flow non-uniformity along the above-floor cold aisle. For example, the operational cooling provision intensity near the Computer Room Airflow Conditioner (CRAC) unit can be far less than that on the other side (far away from the CRAC unit). This undesired trend leads to an unbalanced aisle-level air cooling and subsequent inefficient power consumption. In this study, the CRAC turbofan blower flow boundary conditions were thoroughly investigated. Computational Fluid Dynamics (CFD) based simulations were employed to describe and evaluate the differently configured CRAC turbofan blower flow conditions (i.e., normal, angled, and sheared CRAC flow patterns) as well as their impacts upon the air cooling performance. This work indicates that the considered turbofan blower boundary condition, together with their underlying transportation mechanism within the plenum, might contribute an essential influence to the flow structure adjacent to the tile perforations. In particular, it was found that the sheared CRAC turbofan blower airflow pattern is capable of giving rise to favorable tile flow straightening manners. This finding further promotes an improvement of the consequently obtained aisle-level air cooling effectiveness and efficiencies, contributing to more advanced data center thermal management in the future.

scholarly journals Augmented Reality Maintenance Assistant Using YOLOv5

2021 ◽  
Vol 11 (11) ◽  
pp. 4758
Author(s):  
Ana Malta ◽  
Mateus Mendes ◽  
Torres Farinha
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Object Recognition ◽  
Augmented Reality ◽  
Real Time ◽  
Recognition System ◽  
High Accuracy ◽  
Video Streams ◽  
The Neural Network ◽  
Deep Learning Neural Network

Maintenance professionals and other technical staff regularly need to learn to identify new parts in car engines and other equipment. The present work proposes a model of a task assistant based on a deep learning neural network. A YOLOv5 network is used for recognizing some of the constituent parts of an automobile. A dataset of car engine images was created and eight car parts were marked in the images. Then, the neural network was trained to detect each part. The results show that YOLOv5s is able to successfully detect the parts in real time video streams, with high accuracy, thus being useful as an aid to train professionals learning to deal with new equipment using augmented reality. The architecture of an object recognition system using augmented reality glasses is also designed.

scholarly journals Real-time prediction of 1H and 13C chemical shifts with DFT accuracy using a 3D graph neural network

2021 ◽  
Author(s):  
Yanfei Guan ◽  
S. V. Shree Sowndarya ◽  
Liliana C. Gallegos ◽  
Peter C. St. John ◽  
Robert S. Paton
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Quantum Chemical ◽  
Organic Molecules ◽  
Chemical Shifts ◽  
13C Chemical Shifts ◽  
Time Prediction ◽  
Nmr Data ◽  
Proton Chemical Shifts

From quantum chemical and experimental NMR data, a 3D graph neural network, CASCADE, has been developed to predict carbon and proton chemical shifts. Stereoisomers and conformers of organic molecules can be correctly distinguished.

The Design of a Wavelet-Based Neural Network Adaptive Filter

2013 ◽  
Vol 860-863 ◽  
pp. 2791-2795
Author(s):  
Qian Xiao ◽  
Yu Shan Jiang ◽  
Ru Zheng Cui
Keyword(s):  
Neural Network ◽  
Wavelet Transform ◽  
Real Time ◽  
Adaptive Filter ◽  
Adaptive Algorithm ◽  
Distributed Storage ◽  
Hopfield Neural Network ◽  
Lms Algorithm ◽  
The Neural Network ◽  
Simulation Results

Aiming at the large calculation workload of adaptive algorithm in adaptive filter based on wavelet transform, affecting the filtering speed, a wavelet-based neural network adaptive filter is constructed in this paper. Since the neural network has the ability of distributed storage and fast self-evolution, use Hopfield neural network to implement adaptive filter LMS algorithm in this filter so as to improve the speed of operation. The simulation results prove that, the new filter can achieve rapid real-time denoising.

Sign in / Sign up

Export Citation Format

Share Document