Throughout industry there are complex, dynamic systems that are controlled in real-time. In most of these, it is hard to measure accurately what the system is doing, and situation assessment is very difficult. Detecting changes or deteriorations and predicting the development of faults is vital to maintain the full availability of the equipment.
TIGER plans to deliver a generic approach and software system, developed and successfully demonstrated on demanding applications, to perform real-time situation assessment of complex, dynamic, hard-to-measure systems that support fault prediction and early detection. The project will focus on two applications: a large gas turbine at the Exxon-Shell Fife Ethylene Plant, and a smaller system at Dassault.
The system to be built includes the following aspects: real-time diagnosis using numeric and qualitative simulation; temporal reasoning mechanism, including a trend-based system to support fault prediction; model based diagnosis; and the user interface.
The project is expected to be the basis for several major advances, including the application of qualitative simulation and model based reasoning, a major improvement in the tools available for high-speed situation assessment and a major improvement in gas turbine condition monitoring systems.
KY4 8EP Mossmorran
G81 1YA Clydebank