Gradient-like minimization methods for aeroengines diagnosis and control

Nowadays, there is an ever growing interest for gas turbine and aeroengines prognostics. The capability to assess not only the current state of an asset, but also to be able to predict its remaining useful life (RUL), and hence to perform condition-based maintenance (CBM) —if, and only when, it is needed— can represent a huge deal in the manufacturer profits. Against the plethora of data-driven methods that have arisen in the past few years, there is still some knowledge to be gained in terms of understanding the underlying phenomenology of engine degradation. In fact, it is certainly a non-trivial problem, to realize what has happened to the rotating components of an engine just by observing the pressure being measured by certain sensor rise, or some other temperature measured along the main gas-path decrease its value. In this regard, model-based approaches —and, in particular, gas path analysis (GPA)— can assist us in gaining such knowledge. In this paper, a non-linear GPA technique is revisited, introducing some novelties to the solver, and making use of current computational methods and resources, to establish a solid ‘foundation’ that will serve as the basis for further research.