Development and validation of an effective and robust chiller sequence control strategy using data-driven models

Kui Shan, Shengwei Wang, Dian Ce Gao, Fu Xiao

Research output: Journal article publicationJournal articleAcademic researchpeer-review

30 Citations (Scopus)


Chiller sequence control significantly affects the efficiency and operation stability of plants with multiple chillers. However, in real practice, the energy efficiency is commonly sacrificed to avoid uncertainties. Also, the strategies found in literature may be too complicated to be used practically. An effective and robust strategy for centrifugal chiller plants is therefore developed. The strategy innovatively utilizes chillers inlet guide vane openings as the load, and more particularly the energy efficiency indicator. A validation of the use of such an indicator is conducted using the in situ measurements from the chiller plant in a high-rise building. The strategy is compared with two other commonly used strategies through tests. In the ideal condition (no measurement errors), the proposed strategy saves 3% of the energy comparing to the original strategy. When systematic errors exist in the cooling load measurements, energy performance of the plant is not affected when controlled by the proposed strategy.
Original languageEnglish
Pages (from-to)78-85
Number of pages8
JournalAutomation in Construction
Publication statusPublished - 1 May 2016


  • Centrifugal chiller
  • Chiller sequence control
  • Data-driven models
  • Measurement error
  • Optimal chiller loading
  • Robust control

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Civil and Structural Engineering
  • Building and Construction

Cite this