Parametric Sensitivity Analysis and Model Tuning Applied to Vapor Compression Systems

Abstract

This paper presents a parameter sensitivity analysis for a low-order control-oriented dynamic model of a subcritical vapor compression system. The results are used to tune immeasurable model parameters, account for the unmodeled system dynamics, and are applied to fault detection residual design. The models are validated against data taken from an experimental test stand, and the sensitivity based model tuning is shown to improve the model accuracy while providing enhanced physical insight into system dynamics.