SLIDE 1
industries (FUDIPO) Erik Dahlquist, FUDIPO, Malardalen University - - PowerPoint PPT Presentation
industries (FUDIPO) Erik Dahlquist, FUDIPO, Malardalen University - - PowerPoint PPT Presentation
FUture DIrections of production Planning and Optimized energy- and process industries (FUDIPO) Erik Dahlquist, FUDIPO, Malardalen University Frankfuhrt, 18 October 2018 Partner distribution 2 At a glance sorting good data used for
SLIDE 2
SLIDE 3
At a glance
3
sorting “good data” used for tuning from “bad data” used for fault detection
SLIDE 4
FUDIPO structure
4
Data preprocessing Statistical models Physical Models Machine learning Fault diagnostics Model adaptation Decision support Maintenance OD Risk of failure Order plan Production plan Costs - expences Sales - income Modified Production plan Optimization Model based control Soft sensors
Deviation (sim- meas)
Processes
SLIDE 5
- 1. Demonstrator Background: Mälarenergi AB, Block 6
High temp corrosion sensor
SLIDE 6
Results BN at CFB boiler 5
6
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 6000 12000 18000 24000 30000 36000 42000 48000 54000 60000 66000 72000 78000 84000 90000 96000 102000 108000 114000 120000 126000 132000 138000 144000 150000 156000 162000 168000 174000 180000 186000 192000 198000 204000 210000 216000 222000 228000 234000 240000 246000 252000 258000 264000 270000 276000 282000 288000 294000 300000 306000 312000 318000 324000 330000 336000 342000 Probability Time [s]
Furnace status 2011-09-10 to 2011-09-18
Unballanced right Unballanced left High combustion Normal Sensor 10 false Sensor 11 false Sensor 12 false Sensor 13 false Sensor 14 false Sensor 15 false Sensor 16 false Sensor 17 false Sensor 20 false Sensor 21 false Sensor 22 false Sensor 24 false Sensor 25 false 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 30000 60000 90000 120000 150000 180000 210000 240000 270000 300000 Probability Time [s]
Furnace status
Unballanced right Unballanced left High combustion Normal Sensor 10 false Sensor 11 false Sensor 12 false Sensor 13 false Sensor 14 false Sensor 15 false Sensor 16 false Sensor 17 false Sensor 20 false Sensor 21 false Sensor 22 false Sensor 24 false Sensor 25 false
Temp 1 Temp 2 MC in fuel Temp in Cyclone Corrosion Sensor fault 0.2 0.5 0.3 Tp-Tm Tp-Tm Moisture NIR
Diagnostics and decision support
p=predict m= measured
SLIDE 7
Learning system in a fiberline – updating models semi on-line
7
Residual alkali Kappa number
Billerud-Korsnäs
SLIDE 8
Temperature was higher than predicted. Indicate channeling.
- Temperature in the extraction flow during
channelling:
- yellow curve = measured process value
- violet line = predicted value from simulation
SLIDE 9
Oil refinery at Tupras - Connection of the Physical Models
Optimize use of feed Determine feed comp by NIR
SLIDE 10
Overall scheme of the WWTP at Mälarenergi (Sweden) BN and MPC Reduce: NO3, NH4, BOD, PO4 Minimize electricity Maximize biogas prod
SLIDE 11
Micro gas turbines, mCHP
11
Fleet management of mGT plants Physical and statistical models Measurements Data pretreatment Diagnostics Decision support Maintenance
- n demand
SLIDE 12
Conclusions
- Goal to integrate different functions from low level to
high level
- Build learning systems, that are self adapting
- Develop data structures that can make this posible
- Make supervised AI systems for process industries
12
SLIDE 13
The projects leading to this application have received funding from the European Union’s Horizon 2020 research and innovation program