Key
2110
Conference Year
2012
Keywords
Inverse modeling, Parameter estimation, Vapor compression cycle, Gray-box models, Component-based models
Abstract
This paper utilizes the component models described in a companion paper to develop a full system inverse model for normal and faulty performance of vapor compression cooling and heating equipment. Algorithms to simulate different faults, such as loss of refrigerant charge, compressor valve leakage, liquid line restriction, etc. are presented. Component model parameters were estimated from laboratory experimental data without complete knowledge about the component characteristics. The system model was tuned to offset the bias that resulted from model simplifications. The method was carried out with data from a 3-ton R410a packaged unit with fixed orifice expansion, operating with faults such as incorrect refrigerant charge and heat exchanger fouling. The system model outputs had good agreement with the experimental data. The impacts of faults on performance determined through simulation are also presented.