Automated Health Management for Gas Turbine Engine Accessory Components

Impact Technologies, LLC proposes to develop embeddable software modules for autonomous health management of gas turbine engine accessory components, including actuators, valves, pumps, APUs, and other auxiliary components. As part of the proposed approach, customized physical models will be generated to determine changes in component health using only control command/response data. This proven model-based approach will be fused with innovate signal processing and classification techniques, again operating on control data, to provide a confident assessment of component health. Remaining useful life predictions will also be generated (if sufficient trending data is available) using a suite of trending algorithms and novel tracking methods. This effort will build upon ongoing development of similar technologies for flight control actuators/valves and aircraft generators, as well as previous development of health management for aircraft pumps. As part of this program, available historical data will be evaluated to 1) customize the existing suite of algorithms and models that have been developed for similar systems and 2) identify precursors to impending failures that can be used to increase detection horizon and prediction capability. The proposed algorithms/models will be developed with considerations to a number of implementations, including on-component (embedded within the component, possibly displayed through an LED indicator), on-engine (embedded with the engine controller or similar processing unit), and at-wing (as part of an at-wing automated test equipment package), in order to maximize commercialization and transition potential. False alarm mitigation techniques, such as sensor validation and feature separability analysis, will also be implemented to ensure system reliability. Although the effort will strive to utilize only existing engine sensors, a distributed system of low-overhead (low-cost, low-weight, low-power) sensors will be implemented if the existing suite does not provide sufficient fault coverage.