Sensor fusion for a blade surface-mount icing detector for wind turbines Michael J. Moser, Markus Brandner and Hubert Zangl, Graz University of Technology
Sensor fusion for a blade surface-mount icing detector for wind turbines Michael J. Moser, Markus Brandner and Hubert Zangl, Graz University of Technology
Le web ouvert avec Firefox OS et Firefox - Linux Meetup Montréal - 2014-08-05Frédéric Harper
1.7K views•64 slides
More Related Content
Similar to Sensor fusion for a blade surface-mount icing detector for wind turbines Michael J. Moser, Markus Brandner and Hubert Zangl, Graz University of Technology
Similar to Sensor fusion for a blade surface-mount icing detector for wind turbines Michael J. Moser, Markus Brandner and Hubert Zangl, Graz University of Technology (20)
Sensor fusion for a blade surface-mount icing detector for wind turbines Michael J. Moser, Markus Brandner and Hubert Zangl, Graz University of Technology
1. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
1
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Sensor fusion
for a blade surface-mount icing detector
for wind turbines
Michael J. Moser, Markus Brandner and Hubert Zangl
Institute of Electrical Measurement
and Measurement Signal Processing
Graz University of Technology, Graz, Austria
2. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
2
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Motivation
• Why fusion?
3. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
3
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Motivation
• Why fusion?
- Both capacitive and optical icing detectors have shown to work but
might have drawbacks.
4. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
4
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Motivation
• Why fusion?
- Both capacitive and optical icing detectors have shown to work but
might have drawbacks.
• Why blade-mount?
5. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
5
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Motivation
• Why fusion?
- Both capacitive and optical icing detectors have shown to work but
might have drawbacks.
• Why blade-mount?
- Detect icing where it is relevant.
6. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
6
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Motivation
• Why fusion?
- Both capacitive and optical icing detectors have shown to work but
might have drawbacks.
• Why blade-mount?
- Detect icing where it is relevant.
• Why measure, not estimate from weather conditions?
7. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
7
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Motivation
• Why fusion?
- Both capacitive and optical icing detectors have shown to work but
might have drawbacks.
• Why blade-mount?
- Detect icing where it is relevant.
• Why measure, not estimate from weather conditions?
- Get reliable information e.g. to trigger heating.
8. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
8
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
State of the Art
• Weather monitoring and estimation
- Detection of icing e.g. by means of standard ice detectors
9. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
9
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
State of the Art
• Weather monitoring and estimation
- Detection of icing e.g. by means of standard ice detectors
• Ultrasonic and microwave ice thickness measurement
- Deterioration due to pollution is possible, power demand is high
10. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
10
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
State of the Art
• Weather monitoring and estimation
- Detection of icing e.g. by means of standard ice detectors
• Ultrasonic and microwave thickness measurement
- Deterioration due to pollution is possible, power demand is high
• Indirect Detection of aerodynamical changes
- Loss of power generation is an indicator, detection of vibrations
11. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
11
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
State of the Art
• Weather monitoring and estimation
- Detection of icing e.g. by means of standard ice detectors
• Ultrasonic and microwave thickness measurement
- Deterioration due to pollution is possible, power demand is high
• Indirect Detection of aerodynamical changes
- Loss of power generation is an indicator, detection of vibrations
• Measurement of rotor blade weight
- Averages over entire blade surface
12. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
12
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
New Approach
• Aim #1: „Direct“ measurement
13. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
13
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
New Approach
• Aim #1: „Direct“ measurement
– Measurement of icing on multiple points on the rotor blade surface
14. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
14
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
New Approach
• Aim #1: „Direct“ measurement
– Measurement of icing on multiple points on the rotor blade surface
• Aim #2: Scalability
15. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
15
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
New Approach
• Aim #1: „Direct“ measurement
– Measurement of icing on multiple points on the rotor blade surface
• Aim #2: Scalability
– In principle, applicable to any surface
16. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
16
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
New Approach
• Aim #1: „Direct“ measurement
– Measurement of icing on multiple points on the rotor blade surface
• Aim #2: Scalability
– In principle, applicable to any surface
• Aim #3: Low power consumption
17. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
17
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
New Approach
• Aim #1: „Direct“ measurement
– Measurement of icing on multiple points on the rotor blade surface
• Aim #2: Scalability
– In principle, applicable to any surface
• Aim #3: Low power consumption
– In particular: energy harvesting is feasible -> no wires
18. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
18
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
New Approach
• Aim #1: „Direct“ measurement
– Measurement of icing on multiple points on the rotor blade surface
• Aim #2: Scalability
– In principle, applicable to any surface
• Aim #3: Low power consumption
– In particular: energy harvesting is feasible -> no wires
• Aim #4: No change in aerodynamics
19. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
19
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
New Approach
• Aim #1: „Direct“ measurement
– Measurement of icing on multiple points on the rotor blade surface
• Aim #2: Scalability
– In principle, applicable to any surface
• Aim #3: Low power consumption
– In particular: energy harvesting is feasible -> no wires
• Aim #4: No change in aerodynamics
– Mounted e.g. within heating foil / on or under shell coat
20. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
20
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
EMT‘s Key Competences
Which measurement principles are suitable?
21. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
21
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
EMT‘s Key Competences
Which measurement principles are suitable?
22. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
22
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
EMT‘s Key Competences
Which measurement principles are suitable?
23. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
23
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive Sensing
Evaluation of the variable capacitance of a
measurement capacitor, modification of
24. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
24
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive Sensing
Evaluation of the variable capacitance of a
measurement capacitor, modification of
- capacitor‘s geometry
25. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
25
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive Sensing
Evaluation of the variable capacitance of a
measurement capacitor, modification of
- capacitor‘s geometry
- permittivity of the dielectric medium
26. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
26
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive Sensing
Evaluation of the variable capacitance of a
measurement capacitor, modification of
- capacitor‘s geometry
- permittivity of the dielectric medium
Planar capacitor
Dielectric medium
Parallel plate
capacitor
27. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
27
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive: Pros and Cons
+ -
28. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
28
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive: Pros and Cons
High accuracy
High resolution
Temp. Range -40°C
+ -
29. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
29
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive: Pros and Cons
High accuracy
High resolution
Temp. Range -40°C
Wireless
Low Power
Thickness measurement
+ -
30. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
30
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive: Pros and Cons
x Small signal changes
x Large offsets
High accuracy
High resolution
Temp. Range -40°C
Wireless
Low Power
Thickness measurement
+ -
31. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
31
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive: Pros and Cons
x Small signal changes
x Large offsets
x Disturbers / EMC
x Ambiguities
x Model needed
High accuracy
High resolution
Temp. Range -40°C
Wireless
Low Power
Thickness measurement
+ -
32. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
32
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Method: Capacitance Tomography (ECT)
33. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
33
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive Icing Detection on OHTL
Capacitive Icing Sensor for
OHTL
34. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
34
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive Icing Detection on OHTL
Capacitive Icing Sensor for
OHTL
Field Test since 2009-2011
35. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
35
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Capacitive Icing Detection on OHTL
Capacitive Icing Sensor for
OHTL
Field Test since 2009-2011
Commercial Product since
2011
36. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
36
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Optical Icing Detection
- Make use of optical properties of ice, water and air
37. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
37
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Optical Icing Detection
- Make use of optical properties of ice, water and air
38. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
38
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Optical Icing Detection
- Make use of optical properties of ice, water and air
39. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
39
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Optical Icing Detection: Pros and Cons
+ -
40. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
40
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Optical Icing Detection: Pros and Cons
High accuracy
High resolution
+ -
41. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
41
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Optical Icing Detection: Pros and Cons
High accuracy
High resolution
Robustness against disturbers
Selectivity
+ -
42. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
42
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Optical Icing Detection: Pros and Cons
High accuracy
High resolution
Robustness against disturbers
Selectivity
+ -x more power needed
x pollution effects
x lifetime of optical components
43. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
43
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Optical Icing Detection: Pros and Cons
High accuracy
High resolution
Robustness against disturbers
Selectivity
+ -x more power needed
x pollution effects
x lifetime of optical components
x mechanical issues
x no thickness measurement
44. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
44
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Lab Measurement Setup
45. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
45
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Lab Measurement Setup
46. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
46
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Lab Measurement Setup
47. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
47
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Icing Experiment Results
0 500 1000 1500
3.5
4
4.5
5
5.5
6
6.5
x 10
4 Icing Experiment
IcingIndicator(a.u.)
48. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
48
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Summary
• A capacitive measurement principle for the detection of rotor blade icing
is feasible.
49. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
49
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Summary
• A capacitive measurement principle for the detection of rotor blade icing
is feasible.
• Capacitance Tomography methods can be applied to estimate ice layer
thickness and quality.
50. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
50
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Summary
• A capacitive measurement principle for the detection of rotor blade icing
is feasible.
• Capacitance Tomography methods can be applied to estimate ice layer
thickness and quality.
• Water (melting ice layers, rain and drying) can be clearly distinguished
from any other state.
51. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
51
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Summary
• A capacitive measurement principle for the detection of rotor blade icing
is feasible.
• Capacitance Tomography methods can be applied to estimate ice layer
thickness and quality.
• Water (melting ice layers, rain and drying) can be clearly distinguished
from any other state.
• Ice layers can be watched while growing, thickness can be determined.
52. Institute of Electrical Measurement and Measurement Signal Processing, Graz University of Technology, Austria
Professor Horst Cerjak, 19.12.2005
52
Sensor fusion for a blade surface-mount icing detector for wind turbines – Winterwind 2012, Skellefteå, Sweden – 7./8.2.2012
Summary
• A capacitive measurement principle for the detection of rotor blade icing
is feasible.
• Capacitance Tomography methods can be applied to estimate ice layer
thickness and quality.
• Water (melting ice layers, rain and drying) can be clearly distinguished
from any other state.
• Ice layers can be watched while growing, thickness can be determined.
• Sensor Fusion approach: Optical icing detection can eliminate
ambiguities in capacitive measurement (improved ice/water
discrimination)