ADAPTIVE DYNAMIC MODELS FOR MAINTENANCE-ON_DEMAND AND PROCESS OPTIMIZATION OF COMBINED
HEAT AND POWER PLANTS (ADMADE)
Prof Erik DahlquistMalardalen University
Objectives
• The aim of this application is to build a foundation of mathematical tools for application in the future energy sector, including renewable energy as well as intelligent energy.
• Secondly we need more information on moisture and heating value of different fuels, to optimize the performance.
• Measured process data will be analysed and utilised for process optimization, and not only be collected and stored as is often the case today.
Project
• In the project we will develop the mathematical modeling foundation for doing these type of diagnostics and optimizations for later implementation in different power plant and process industries generally.
• - Physical models will be combined with statistical models in a systematic way to make it possible to adapt the models as conditions change, and to follow effect of new fuels.
• - A hierarchical structure will be introduced for • 1) measurement of fuel properties using NIR and RF together with statistical
models like PLS, • 2) process diagnostics comparing simulations to measurements in the
process combined with Bayesian Nets and • 3) production planning including when maintenance has to be done. • 4) on-line control and optimization using model based, multivariable control.
This includes both the production and district heating system.
Partners
• Mälarenergi AB• Eskilstuna Energy and Environment• ENA Energy• Vattenfall