A Modelling Workflow for Predictive Control in Residential Buildings


O’Dwyer E., ATAM E., Falugi P., Kerrigan E., Zagorowska M., Shah N.

Green Energy and Technology, Springer International Publishing Ag, ss.99-128, 2022

  • Yayın Türü: Kitapta Bölüm / Araştırma Kitabı
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1007/978-3-030-79742-3_5
  • Yayınevi: Springer International Publishing Ag
  • Sayfa Sayıları: ss.99-128
  • Boğaziçi Üniversitesi Adresli: Evet

Özet

Despite a large body of research, the widespread application of Model Predictive Control (MPC) to residential buildings has yet to be realised. The modelling challenge is often cited as a significant obstacle. This chapter establishes a systematic workflow, from detailed simulation model development to control-oriented model generation to act as a guide for practitioners in the residential sector. The workflow begins with physics-based modelling methods for analysis and evaluation. Following this, model-based and data-driven techniques for developing low-complexity, control-oriented models are outlined. Through sections detailing these different stages, a case study is constructed, concluding with a final section in which MPC strategies based on the proposed methods are evaluated, with a price-aware formulation producing a reduction in operational space-heating cost of 11%. The combination of simulation model development, control design and analysis in a single workflow can encourage a more rapid uptake of MPC in the sector.