Neural network-based predictive control system for energy optimization in sports facilities: a case study
2021
Given the increased global energy demand and its associated environmental impacts, the
management and optimization of sports facilities is becoming imperative as they are
characterized by high energy demand and occupancy profiles. In this work, the theory of model
predictive control ȋMPCȌ is combined with neural networks for temperature setpoint selection to
achieve energy and performance optimization of sports facilities. It is demonstrated using the
building information model ȋBIMȌ of a sports hall in the sports complex of Qatar University. MPC
systems are powerful as they allow integrated dynamic optimization that accounts for the future
system behavior in the decision-making process, while neural networks are advantageous for
their ability to represent complex interdependencies with high accuracy. The proposed approach
was able to achieve a total energy savings of around ͵͵Ψ. Considerations about the network
performance, MPC settings tuning, and optimization sub-optimality or failure are essential during
the design and implementation phases of the proposed system.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
0
Citations
NaN
KQI