| Popis: |
An innovative sensor diagnostic and validation tool has been developed for testing and in-flight analysis of gas turbine engine data under the Air Force Small Business Innovations Research (SBIR) program. The software-based diagnostics tool fuses signal processing and data-driven modeling techniques to monitor signals for sensor anomalies and abnormal engine behavior and to provide an estimate of the healthy signal in cases of fault diagnoses. The data-driven modeling approach requires no modification to the candidate system, enabling the diagnostic tool to be used virtually anywhere data are available. Prior to online monitoring, a short training period is required to create a data-driven model which provides estimated values for failed sensors. The integrated diagnostics sensor validation tool has been installed, tested, and proven at the AEDC Engine Test Facility, successfully interfacing with real-time and stored data sources for on-line diagnostics. The integration of a MATLAB software suite enhances the diagnostic capability by enabling users to insert customized analysis algorithms without requiring specialized interface code to the data system. These tools are easily deployable, facilitating test article sensor diagnostics and enabling further focus on engine development. Background |
