Time Series Analysis, Forecasting and Control
Time Series Analysis, Forecasting and Control
| Hi-index | 0.00 |
This paper presents physics-based models as a key component of prognostic and diagnostic algorithms of health monitoring systems. While traditionally overlooked in condition-based maintenance strategies, these models potentially offer a robust alternative to experimental or other stochastic modeling data. Such a strategy is particularly useful in aerospace applications, presented in this paper in the context of a helicopter transmission model. A lumped parameter, finite element model of a widely used helicopter transmission is presented as well as methods of fault seeding and detection. Fault detection through diagnostic vibration parameters is illustrated through the simulation of a degraded rolling-element bearing supporting the transmission's input shaft. Detection in the time domain and frequency domain is discussed. The simulation shows such modeling techniques to be useful tools in health monitoring analysis, particularly as sources of information for algorithms to compare with real-time or near real-time sensor data.