Advanced Controls Research

Advanced Controls ResearchAdvanced Controls Research

Alan D. WrightAlan D. WrightLee FingershLee Fingersh

Maureen HandMaureen HandJason JonkmanJason Jonkman

Gunjit BirGunjit Bir

2006 Wind Program Peer Review2006 Wind Program Peer Review

May 10, 2006May 10, 2006

http://www.sandia.gov/index.html

22006 Wind Program Peer Review

Create design methodology for advanced controls: – regulate rotor-speed and/or maximize energy

extraction.– stabilize important flexible modes of the turbine to

reduce dynamic loads and response.

Develop control design and modeling tools.

Develop controls field testing capability.

Objectives


Commercial Turbine Control

Generator Torque

Nacelle Yaw

Blade Pitch

Control Actions

Generator Speed

Generator Torque

Rotor Collective Pitch

Region 2

Region 3

Wind Disturbances

PID Pitch Controller

Drive-train Damper

T = kw^2

Nonlinear Turbine


What else can we do?

Improve energy capture– Active rather than

passive rotor control• Negative inertia - Use

of shaft torque to cancel rotor inertia

– Adaptive control– Active pitch following– Optimal torque

control

Reduce loads– Load feedback– Independent pitch

control– Periodic gains– Active tower / blade /

drive-train damping– Advanced sensors– Look-ahead controls


Adaptive control0.3% - 5% energy capture increase

0 50 100 150 200 250 300 3500.5

1.0

1.5

2.0

No

rma

lize

d M

(M

/M+)

0 50 100 150 200 250 300 350

0.3

0.4

0.5

Fra

cti

on

al A

ve

rag

e P

ow

er

Time (hours)5 10 15 20

0

100

200

300

400

500

600

Gri

d P

ow

er (

kW)

Mean Equivalent Wind Speed (m/s)

Standard ControlAdaptive Control

Region 3Region 2


Control of Flexible Modes

Blade-2 Lag

Tower Side-Side

Rotor Rotation

(b) Frontview

Blade-1 Lag

Blade-1 Flap

Tower Fore-Aft

Rotor Teeter

(a) Sideview

Blade-2 Flap

Generator Rotation

Drive-train Torsion


Regulate rotor-speed in the presence of wind-speed disturbances and stabilize turbine modes.

– Stabilize flexible modes through full state feedback.

– Use state estimation to provide the controller with needed states (including wind-speed).

– Represent wind disturbance with extra states. Controller accounts for fluctuating wind speeds and shears.

Multiple input/multiple output, single control loop

State Feedback Control


State-Space Linear Model

Linear time-periodic model

xtCy

utBxtAx

)(

)()(

w

q

q

x

Structural DOFs & rates

Pitch actuator statesWind disturbance

com

gTu

Up to 17 states

y Turbine measurements

)( ),( ),( tCtBtA

State matrices periodic over each rotation

Up to 9 measurements

Tower & blade strain gauges

LSS torque

Yaw, teeter, LSS angles

Pitch angles

where


Process/Tools

Design SimulateLinear Model FAST

DAC ADAMSLQR Simulink

Field testCART

CART-3Industry

ModifyAnalyze data

Make changes

Iterate

OutData

Yaw Controller

Torque Controller

Pitch Controller

f(u)

Gen. Torque

Yaw Position

Blade Pitch

FAST Nonlinear Wind Turbine


Field Tested Collective Pitch Controller15% - 50% reduction in Shaft Torque fatigue loads

Measured Shaft Torque

60

80

100

120

140

160

180

200

0 5 10 15 20 25 30

Time (sec)

Lo

w-S

pee

d S

haf

t To

rqu

e

PI ControlState-space (FAST) controller


CART Test Results – Pitch Control

0.0

1.0

Meanshaft

pow er

Tow erfore-aftFDEL

Tow erside-side

FDEL

LSStorsionFDEL

Blade flapFDEL

Coll. Pitch

Norm

aliz

ed t

o

Base

line C

ontr

olle

r Perf

orm

ance

Significant reduction in most measured loads (30-70%)

Region 2 Region 3

0.0

1.0

RMSspeederror

Tow erfore-aftFDEL

Tow erside-side

FDEL

LSStorsionFDEL

Bladeflap FDEL

Coll. Pitch

Ind. Pitch


Disturbance Model

h

z

hubV

hub( ) V (1 / )mV z z h

hotsh

mrV

h

rmmVVrV hubhubhub

cos

4

)1(),(

2

2

h

z

hubV hub( ) V (1 / )mV z z h

hotsh

mrV

h

rmmVVrV hubhubhub

cos

4

)1(),(

2

2

Uniform wind component

Fluctuating wind component


Simulated Control With New Independent Pitch/DAC

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

43 44 45 46 47 48 49 50

Time (sec)

Bla

de

1 F

lap

De

fle

cti

on

(m

.)

Old DAC New DAC

Must measure either tip-deflection or flap-bending moments on each blade


Use of Lidar Measured Wind-speed

Simulated single lidar hub-mounted on CART

Conical scan achieved through rotor rotation


Simulated Results

Steady 18 m/s wind, power law exponent =

0.17

Steady 18 m/s wind, power law exponent =

0.3

DAC DAC + LIDAR DAC DAC + LIDAR

Blade 1 RFB moment DEL (kNm) 53 51 83

64

Turbulent 18 m/s wind,power law exponent =

0.17

Turbulent 18 m/s wind,power law exponent =

0.3

DAC DAC + LIDAR DAC DAC + LIDAR

Blade 1 RFB moment DEL (kNm) 295 263 310 278


No control Control

Measured Tower Load Reduction Using Generator Torque Control (CART)


Conclusions

17

Must move away from using old control schemes with multiple loops

Advanced Controls show great potential for meeting multiple control objectives– Stabilizing turbine structure– Enhancing energy capture– Mitigating dynamic loads

Will be critical for large flexible machines as well as offshore turbines with many flexible modes


Plans - Future Work

Complete development of control design tools for industry.

Continue advanced controls development and testing.

Develop new field testing capability on a large flexible turbine – partner with industry.

Implement and test controls on commercial turbines.


CART-3 – 3-bladed hub testing

Supplement to the 2-bladed CART

Advanced controls integration

Designed for testing modern controls– Loads– Deflection– Advanced sensors


Opportunities

Partner with industry to implement controls on large turbines.

Develop controls test-bed/floating platform simulator.


Questions and comments

Advanced Controls Research

Documents