Dynamic indoor thermal comfort model identification based on neural computing PMV index

Research output: Contribution to journalConference article

5 Citations (Scopus)


This paper focuses on modelling and simulation of building dynamic thermal comfort control for non-linear HVAC system. Thermal comfort in general refers to temperature and also humidity. However in reality, temperature or humidity is just one of the factors affecting the thermal comfort but not the main measures. Besides, as HVAC control system has the characteristic of time delay, large inertia, and highly nonlinear behaviour, it is difficult to determine the thermal comfort sensation accurately if we use traditional Fanger's PMV index. Hence, Artificial Neural Network (ANN) has been introduced due to its ability to approximate any nonlinear mapping. Using ANN to train, we can get the input-output mapping of HVAC control system or in other word; we can propose a practical approach to identify thermal comfort of a building. Simulations were carried out to validate and verify the proposed method. Results show that the proposed ANN method can track down the desired thermal sensation for a specified condition space.

Original languageEnglish
Article number012113
JournalIOP Conference Series: Earth and Environmental Science
Issue number1
Publication statusPublished - 01 Jan 2013
Event26th IAHR Symposium on Hydraulic Machinery and Systems - Beijing, China
Duration: 19 Aug 201223 Aug 2012


All Science Journal Classification (ASJC) codes

  • Environmental Science(all)
  • Earth and Planetary Sciences(all)

Cite this