SBIR-STTR Award

Stress Technology Incorporated proposes a Probabilistic Diagnostic and Prognostic System (ProDaPS) capable of probabilistic assessment of engine sensor signals, mechnanical and performance diagnostics, and critical component prognostics. The ProDaPS system will perform real-time risk assessment of engine performance and mechanical faults so that catastrophic engine failures can be avoided. Also, costly inspection policies and premature component replacements can be averted by optimizing the maintenance activities based on this risk-cost relationship. The ProDaPS system will integrate refined stochastic modeling and risk assessment tools together with advanced engine health monitoring techniques using the latest sensor technology and advanced probabilistic fault classifiers. The proposed ProDaPS system will ultimately provide a real-time, predictive system with truly cognitive and reasoning capabilities.The ProDaPS system will be capable of real-time probabilistic risk assessment and accurate, early detection of engine performance/vibration faults while component failures can still be prevented. The implementation of state-of-the-art probabilistic, diagnostic and prognostic techniques will include; advanced stochastic map (random field) models used for pattern recognition, Bayesian belief analysis for updating statistical information from models/measurements, and the Karhunen-Loeve (KL) expansion as a stochastic fault classifier. Component prognostics will assess thermomechanical fatigue and creep life in critical engine components utilizing efficient Monte Carlo simulation procedures and algorithms taking into account the interaction of LCF/HCF/creep failure modes. Remaining life predictions and probabilistic risk assessment will be based on the results from the real-time diagnostic/prognostic modules and corresponding economic and safety consequences.The proposed program has genuine support from GE Aircraft Engines and Rolls-Royce that is based on a strong working relationship that recently resulted in the successful completion of an advanced engine health monitoring system for the NAVY F405 engine and future development for the GE F101 engine.

Keywords:
performance monitoring real-time health monitoring life prediction vibration monitoring engine