DSP subsystem for knowledge based health monitoring of gas turbine engines

The design for a DSP (digital signal processing) subsystem for a health monitoring system for gas turbine engines is described. Knowledge-based techniques are emerging as useful tools for health and condition monitoring of high value engineering systems. A demonstrator system that uses these techniques for monitoring the health and performance of a marine gas turbine engine is described, with particular emphasis on the DSP subsystem which interfaces directly with the raw sensor data coming from the monitored system. The DSP subsystem is a coupled system using both numerical and symbolic methods for signal interpretation; i.e., the DSP subsystem provides a description of the sensors signals in meaningful symbolic terms that reflect the state of the monitored system. The signal abstractions are put in a form suitable for symbolic processing by the knowledge-based diagnostic subsystem of the monitoring system to determine the health of the monitored system