Use of Hybrid Engine Modeling for On-Board Module Performance Tracking

Abstract

A practical consideration for implementing a real-time on-board Module performance tracking system is the development of a high fidelity engine model capable of providing a reference level from which performance changes can be tracked. Real-time engine models made their advent with the State Variable Model (SVM) in the mid-80’s which provided a piecewise linear model that granted a reasonable representation of the engine during steady state operation and mild transients. Increased processor speeds over the next decade allowed more complex models to be considered which were combinations of linear and non-linear physics based components. While the latter may provide greater fidelity over transient operation and flight envelope excursions, it bears the limitation of potential model obsolescence as performance improvements in the form of hardware modifications, bleed and stator vane schedules alterations, cooling flow adjustments, and the like are made during an engine’s life cycle. Over time, these models may deviate enough from the actual engine being monitored that the module performance estimations are inaccurate and misleading. This paper describes an alternate approach to engine modeling by applying a hybrid engine model architecture that incorporates both physics-based and empirical components. This methodology provides a means to tune the engine model to a particular configuration as the engine development matures and furthermore, aligns the model to the particular engine being monitored to insure accurate performance tracking while not compromising real-time operation.