A High-Fidelity Building Performance Simulation Test Bed for the Development and Evaluation of Advanced Controls

Thibault Marzullo, Sourav Dey, Nicholas Long, Jose Vilaplana, Gregor Henze

Research output: Contribution to journalArticlepeer-review

11 Scopus Citations

Abstract

We present an open-source building performance simulation test bed, the Advanced Controls Test Bed (ACTB), that interfaces high-fidelity Spawn of EnergyPlus building models, with advanced controllers implemented in Python. The ACTB leverages the Building Optimization Testing and Alfalfa platforms for managing simulations, providing an external clock, a representational state transfer (REST) application programming interface (API), and key performance indicators for evaluating the effectiveness of control strategies. The REST API allows the development of external controllers programmed in languages such as Python, which provides flexibility and a rich choice of scientific libraries for designing control sequences. We present three test cases based on the U.S. Department of Energy's Reference Small Office Building to demonstrate the ACTB's capabilities: (a) rule-based controls compliant with ASHRAE Guideline 36 control sequences; (b) an economic model predictive control implemented using do-mpc; and (c) a deep Q-network reinforcement learning agent implemented using OpenAI Gym. Abbreviations: ACTB: Advanced Controller Test Bed; AHU: Air Handling Unit; AI:Artificial Intelligence; API: Application Programming Interface; BEM: Building EnergyModeling; BSS: Best Subset Selection; DOE: Department of Energy; DQN: Deep-QNetwork; EKF: Extended Kalman Filter; FMI: Functional Mock-up Interface; FMU:Functional Mock-up Unit; FSS: Forward Stepwise Selection; HVAC: Heating; Ventilationand Air Conditioning; KPI: Key Performance Indicator; LTI: Linear Time-Invariant; MBL: Modelica Buildings Library; MHE: Moving Horizon Estimator; MPC: ModelPredictive Control; N4SID: Numerical Subspace State-Space System Identification; REST: Representational State Transfer; RL: Reinforcement Learning; ROM: Reducedorder model.

Original languageAmerican English
Pages (from-to)379-397
Number of pages19
JournalJournal of Building Performance Simulation
Volume15
Issue number3
DOIs
StatePublished - 2022

Bibliographical note

Publisher Copyright:
© 2022 International Building Performance Simulation Association (IBPSA).

NREL Publication Number

  • NREL/JA-5500-81683

Keywords

  • Advanced
  • controls
  • machine learning
  • model predictive control
  • simulation

Fingerprint

Dive into the research topics of 'A High-Fidelity Building Performance Simulation Test Bed for the Development and Evaluation of Advanced Controls'. Together they form a unique fingerprint.

Cite this