Title :
Exhaust gas temperature data prediction by autoregressive models
Author :
Kumar, Amar ; Srivastava, Alka ; Goel, Nita ; McMaster, Jon
Author_Institution :
Tecsis Corp., Ottawa, ON, Canada
Abstract :
Gas turbine engine performance and health conditions are continuously assessed by exhaust gas temperature that indicate the thermal health condition of engine. Analysis of exhaust gas temperature (EGT) data and its prediction is very important for operational safety, reliability, life cycle cost and power output. Autoregressive (AR) and moving average (MA) techniques, either singly or in combination are used for modeling, validation and prediction of EGT in this work. Model performance is investigated by estimating percent error and mean squared error. Models are used for short and long term predictions. Autoregressive models with small indices are found to offer best performance.
Keywords :
autoregressive moving average processes; condition monitoring; engines; exhaust systems; gas turbines; life cycle costing; reliability; thermal analysis; autoregressive models; continuous assessment; engine thermal health condition; exhaust gas temperature data prediction; gas turbine engine performance; life cycle costing; mean square error method; moving average techniques; operational safety; reliability; Analytical models; Autoregressive processes; Computational modeling; Data models; Engines; Predictive models; Time series analysis; Auto regression; Exhaust gas temperature; Mean square error; Moving average;
Conference_Titel :
Electrical and Computer Engineering (CCECE), 2015 IEEE 28th Canadian Conference on
Conference_Location :
Halifax, NS
Print_ISBN :
978-1-4799-5827-6
DOI :
10.1109/CCECE.2015.7129408
