Development of an accurate gray-box model of ubiquitous residential HVAC system for precise performance prediction during summer and winter seasons

Abstract

In this paper, gray-box models of a residential HVAC system commonly found in Canadian households are developed. The system comprises of four mjor subsystems called heat recovery ventilator (HRV), air handling unit (AHU), air source heat pump (ASHP) and zone. Models were developed by writing energy balance equations of each subsystem. Measured data from residential HVAC system of Archetype Sustainable House located in Vaughan, Canada was used for parameter estimation and validation of each subystem model. The house is outfitted with temperature, flow rate and power consumption measurement sensors for all subsystems. Two distinct sets of parameeters were found using winter and summer datasets. Parameter estimation was carried out using pattern search with Latin hypercube method and Active-Set algorithm. Developed models were evaluated against measured data using several performance comparison metrics and showed a very high value of goodness of fit in the range of 70% and 88% during both winter and summer seasons.