ABSTRACT
The digital twin, as a digital counterpart of a physical entity, has shown great potential in data center prototyping and predictive thermal management. In this regard, Computational Fluid Dynamics/Heat Transfer (CFD/HT) models have been widely adopted. However, the computing time of the CFD/HT simulation is prohibitively long in practice. The Proper Orthogonal Decomposition (POD) has been explored to approximate the CFD/HT simulation by a linear combination of POD modes and coefficients. Existing approaches to calculating the POD coefficients use either the black-box interpolation or the simplified physical model, leading to unsatisfactory generalization ability. To advance existing approaches, this paper proposes Reducio, a physics-guided model reduction approach based on the POD to predict the temperature field by following two key phases of i) offline POD modes calculation and coefficients interpolation and ii) online coefficients extrapolation supervised by the principle of energy balance. To extrapolate the coefficients with limited training data, we adopt the Gaussian Process (GP) model to learn a nonlinear map between the boundary conditions and POD coefficients. We conduct two case studies in two data centers with different scales. Evaluation results in the edge data center show that Reducio achieves sub-1°C mean absolute error (MAE) in temperature field prediction compared with the CFD/HT simulation result, outperforming the existing method based on the simplified physical model by 1.5 °C. When evaluating in the industry-grade hyper-scale data center with the sensor measurements, around 1°C temperature prediction MAE is observed. Furthermore, Reducio can predict the full-fledged temperature field in real-time, making it a strong candidate for building data center predictive digital twins.
- Jayati Athavale, Minami Yoda, and Yogendra Joshi. 2019. Comparison of data driven modeling approaches for temperature prediction in data centers. International Journal of Heat and Mass Transfer 135 (2019), 1039--1052.Google ScholarCross Ref
- Zhiwei Cao, Xin Zhou, Han Hu, Zhi Wang, and Yonggang Wen. 2022. Toward a Systematic Survey for Carbon Neutral Data Centers. IEEE Communications Surveys & Tutorials 24, 2 (2022), 895--936. Google ScholarCross Ref
- Drury B Crawley, Linda K Lawrie, Frederick C Winkelmann, Walter F Buhl, Y Joe Huang, Curtis O Pedersen, Richard K Strand, Richard J Liesen, Daniel E Fisher, Michael J Witte, et al. 2001. EnergyPlus: creating a new-generation building energy simulation program. Energy and buildings 33, 4 (2001), 319--331.Google Scholar
- Steven Diamond and Stephen Boyd. 2016. CVXPY: A Python-embedded modeling language for convex optimization. Journal of Machine Learning Research 17, 83 (2016), 1--5.Google ScholarDigital Library
- Rajat Ghosh and Yogendra Joshi. 2011. Dynamic reduced order thermal modeling of data center air temperatures. In International Electronic Packaging Technical Conference and Exhibition, Vol. 44625. 423--432.Google ScholarCross Ref
- GPy. since 2012. GPy: A Gaussian process framework in python. http://github.com/SheffieldML/GPy.Google Scholar
- Sandeep KS Gupta, Rose Robin Gilbert, Ayan Banerjee, Zahra Abbasi, Tridib Mukherjee, and Georgios Varsamopoulos. 2011. Gdcsim: A tool for analyzing green data center design and resource management techniques. In 2011 International Green Computing Conference and Workshops. IEEE, 1--8.Google ScholarDigital Library
- Xu Han, Wei Tian, Jim VanGilder, Wangda Zuo, and Cary Faulkner. 2021. An open source fast fluid dynamics model for data center thermal management. Energy and Buildings 230 (2021), 110599.Google ScholarCross Ref
- Hrvoje Jasak, Aleksandar Jemcov, Zeljko Tukovic, et al. 2007. OpenFOAM: A C++ library for complex physics simulations. In International workshop on coupled methods in numerical dynamics, Vol. 1000. IUC Dubrovnik Croatia, 1--20.Google Scholar
- David J Lettieri, Michael M Toulouse, Cullen E Bash, Amip J Shah, and Van P Carey. 2013. Computational and experimental validation of a vortex-superposition-based buoyancy approximation for the COMPACT code in data centers. Journal of Electronic Packaging 135, 3 (2013), 030903.Google ScholarCross Ref
- Lei Li, Chieh-Jan Mike Liang, Jie Liu, Suman Nath, Andreas Terzis, and Christos Faloutsos. 2011. Thermocast: a cyber-physical forecasting model for datacenters. In Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining. 1370--1378.Google ScholarDigital Library
- Zachary M Pardey, James W VanGilder, Christopher M Healey, and David W Plamondon. 2015. Creating a calibrated CFD model of a midsize data center. In International Electronic Packaging Technical Conference and Exhibition, Vol. 56888. American Society of Mechanical Engineers, V001T09A029.Google ScholarCross Ref
- Tu-Hoa Pham, Giovanni De Magistris, and Ryuki Tachibana. 2018. Optlayer-practical constrained optimization for deep reinforcement learning in the real world. In 2018 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 6236--6243.Google ScholarDigital Library
- Amir Radmehr, Brendan Noll, John Fitzpatrick, and Kailash Karki. 2013. CFD modeling of an existing raised-floor data center. In 29th IEEE Semiconductor Thermal Measurement and Management Symposium. IEEE, 39--44.Google ScholarCross Ref
- Emad Samadiani and Yogendra Joshi. 2010. Multi-parameter model reduction in multi-scale convective systems. International Journal of Heat and Mass Transfer 53, 9--10 (2010), 2193--2205.Google ScholarCross Ref
- Emad Samadiani and Yogendra Joshi. 2010. Proper orthogonal decomposition for reduced order thermal modeling of air cooled data centers. Journal of heat transfer 132, 7 (2010).Google ScholarCross Ref
- Emad Samadiani, Yogendra Joshi, Hendrik Hamann, Madhusudan K Iyengar, Steven Kamalsy, and James Lacey. 2012. Reduced order thermal modeling of data centers via distributed sensor data. Journal of heat transfer 134, 4 (2012).Google ScholarCross Ref
- Mike Shafto, Mike Conroy, Rich Doyle, Ed Glaessgen, Chris Kemp, Jacqueline LeMoigne, and Lui Wang. 2012. Modeling, simulation, information technology & processing roadmap. National Aeronautics and Space Administration 32, 2012 (2012), 1--38.Google Scholar
- Umesh Singh, Amarendra Singh, S Parvez, and Anand Sivasubramaniam. 2010. CFD-Based Operational Thermal Efficiency Improvement of a Production Data Center.. In SustainIT.Google Scholar
- Lawrence Sirovich. 1987. Turbulence and the dynamics of coherent structures. I. Coherent structures. Quarterly of applied mathematics 45, 3 (1987), 561--571.Google Scholar
- Kunihiko Taira, Steven L Brunton, Scott TM Dawson, Clarence W Rowley, Tim Colonius, Beverley J McKeon, Oliver T Schmidt, Stanislav Gordeyev, Vassilios Theofilis, and Lawrence S Ukeiley. 2017. Modal analysis of fluid flows: An overview. Aiaa Journal 55, 12 (2017), 4013--4041.Google ScholarCross Ref
- Wei Tian, Thomas Alonso Sevilla, Dan Li, Wangda Zuo, and Michael Wetter. 2018. Fast and self-learning indoor airflow simulation based on in situ adaptive tabulation. Journal of Building Performance Simulation 11, 1 (2018), 99--112.Google ScholarCross Ref
- Ruihang Wang, Xin Zhou, Linsen Dong, Yonggang Wen, Rui Tan, Li Chen, Guan Wang, and Feng Zeng. 2020. Kalibre: Knowledge-based Neural Surrogate Model Calibration for Data Center Digital Twins. In Proceedings of the 7th ACM International Conference on Systems for Energy-Efficient Buildings, Cities, and Transportation. 200--209.Google ScholarDigital Library
- Christopher K Williams and Carl Edward Rasmussen. 2006. Gaussian processes for machine learning. Vol. 2. MIT press Cambridge, MA.Google Scholar
- Wangda Zuo and Q Chen. 2007. Validation of fast fluid dynamics for room airflow. IBPSA Building Simulation 2007 (2007).Google Scholar
Index Terms
- Reducio: model reduction for data center predictive digital twins via physics-guided machine learning
Recommendations
Toward Data Center Digital Twins via Knowledge-based Model Calibration and Reduction
Computational fluid dynamics (CFD) models have been widely used for prototyping data centers. Evolving them into high-fidelity and real-time digital twins is desirable for the online operations of data centers. However, CFD models often have ...
Local POD Plus Galerkin Projection in the Unsteady Lid-Driven Cavity Problem
A local proper orthogonal decomposition (POD) plus Galerkin projection method is applied to the unsteady lid-driven cavity problem, namely the incompressible fluid flow in a two-dimensional box whose upper wall is moved back and forth at moderately ...
A numerical method for three-dimensional gas-liquid flow computations
A numerical method for multiphase flow computations based on a finite-difference formulation with a fixed grid is described. The method combines ideas from front tracking and the Ghost Fluid Method, with a numerical technique for velocity extrapolation ...
Comments