Regresar

Quantitative causality analysis for the diagnosis of energy systems

Abstract:

The aim of Thermoeconomic diagnosis is to identify malfunctioning components in energy systems and to quantify the impact on fuel consumption caused by each one. Starting from a diagnosis algorithm previously developed, a complete methodology, including the analysis of measurement errors influence, is presented. The accuracy of the approach is demonstrated by quantifying the error produced in the diagnosis of a large amount of points of a real example. Besides, the method is compared with two diagnosis approaches based on linear regression and neural networks. Results show that these alternatives provide acceptable results for the most important variables, but fail to diagnose the complete set of variables affecting the behaviour of the system, specially the nonlinear ones.