Operational Failure Detection Applied to Heat Exchange Systems Using State Observer Method

Abstract

The role of computational modeling is becoming predominant in the exploration and evaluation of new technologies. Numerical simulations have solved the most complex phenomena immersed in thermal sciences by the implementation of sophisticated mathematical algorithms. This study incorporates a stationary thermodynamic model in order to evaluate the main operational parameters and the overall performance of heat exchange systems based on energy and exergy viewpoints while implementing unknown inputs. Moreover, the study introduces a UIO state observer technology in order to facilitate the early prediction of failure events in heat exchange systems using a detect, isolate and identify (FDI) methodology. Therefore, the proposed model has induced different fault conditions in order to examine the overall performance while integrating error state of signals criteria from the relative difference between the real and modeled estimates. The outcomes have demonstrated a good agreement for the estimated values of the mathematical algorithm considering the imminent presence of disturbance and interference that promotes modeling error. Finally, the results have outlined that the estimations of real operation conditions are not affected by the unknown inputs.