Adaptive neuro-PMV model for thermal comfort control in office air-conditioning systems in a tropical climate

Sowmya Ravichandran

Adaptive neuro-PMV model for thermal comfort control in office air-conditioning systems in a tropical climate

Sowmya Ravichandran

Nanyang Technological University, Singapore

: J Nucl Ene Sci Power Generat Technol

Abstract

An adaptive Neuro-Predictive Mean Vote (Neuro-PMV) model is developed as a part of the Nanyang Technological University- Jurong Town Council Industrial Infrastructure Innovation Centre (NTU-JTC I3C) research initiative to realize a comfortcentric and energy-efficient air-conditioning system in commercial buildings. A learning algorithm based on the proposed model is implemented using intelligent concepts of Artificial Neural Networks (ANN) and fuzzy logic, to control and optimize the thermal comfort of the indoor occupants in an air-conditioned environment. In order to predicate this algorithm, a test bed facilitating realtime monitoring and control has been established at NTU Singapore using a multi-agent system supported by state-of-the art Z-wave technology. The proposed model is validated in the test bed using research intensive experimental analysis and questionnaire-based survey. The mathematical model, computational design and experimental results achieved using the proposed model of thermal comfort control are discussed in this paper.