Lärande och optimering
Simulation and optimisation for future industrial applications (SOFIA)
DIAGNOSIS – Adaptive performance simulation tools for field and fleet diagnostics and robust decision-making
Continuous measurements are taken during the normal operation of powerplants for power and propulsion. Operators and manufacturers have a strong need for simulating and monitoring the condition of their individual unit or fleets.
Project manager at MDH
Description of the project
Continuous measurements are taken during the normal operation of powerplants for power and propulsion. Operators and manufacturers have a strong need for simulating and monitoring the condition of their individual unit or fleets. Analysis of field and fleet data is best performed using detailed dynamic aero-thermodynamic models coupled with adaptive performance algorithms and “big data” analysis techniques. Together with decision-making techniques, such as Bayesian networks, one can improve product understanding including normal and start-up operation, as well as diagnose early upcoming technical problems. Significant cost savings can be achieved by scheduling for example maintenance actions at the least disruptive time. Field and fleet operations can also be improved if one has knowledge of the effective operating life and tactical capabilities remaining in each individual unit.
In the DIAGNOSIS project, we, a team of researchers from Mälardalen University, Saab and Siemens propose an advanced approach for powerplant performance simulation, diagnostic analysis and decision-making. Our approach shall focus on adaptive physics-based models tuned with field and fleet operational data. We will demonstrate the predictive capabilities of the proposed method for field and fleet operation focusing on both aero and land-based applications. The developed tools and approach will be generic, which opens the opportunity for application in a variety of other fields where gas turbines are prevalent (e.g. marine propulsion, oil and gas operations etc.); the suggested approach can also have applications within the process industry.