Simulation of fault impacts for vapor compression systems by inverse modeling. Part I: Component modeling and validation

Abstract

Fault detection and diagnostic systems for vapor compression systems are emerging within the marketplace but with limited information and no standards relating to how well they work. As a result, there is an ongoing effort to develop methods to evaluate fault detection and diagnostic system performance that rely on a large array of performance data of different types of faulted vapor compression systems at arbitrary fault levels. In order to generate the needed data accurately and robustly, a gray-box modeling technique that relies on parameter estimation from measured performance data for systems with and without faults has been developed. The approach involves the use of component-based models that are integrated within a system model where fault models can be introduced to predict system performance at various fault levels. This article presents the component models, parameter estimation approaches, and component prediction accuracy for eight different types of cooling systems that have various types of compressors, expansion devices, and rated capacities and where faults have been introduced within a laboratory setting. A companion article discusses the system level and fault modeling.