Professional Documents
Culture Documents
MaynoothSlides JH AF
MaynoothSlides JH AF
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 1/31
Outline of the presentation
PART 1
PART 2
Major conclusions
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 2/31
PART1 - Mutriku power plant test case
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 3/31
Hydrodynamic model
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 5/31
Biradial turbine to be installed at Mutriku within OPERA H2020 Project
p 1.4
Ψ= Π
ρin Ω2 d 2 1.2
η
1.0 Φ
ṁturb 0.8
Φ=
ρin Ω d 3 0.6
Pturb 0.4
[− ]
Π=
ρin Ω3 d 5 0.2
0.0
Π
η= 0.2
ΨΦ 0.4
0.6
0.8
8 6 4 2 0 2 4 6 8
d - rotor diameter Ψ [− ]
For Re > 106 and Ma < 0.3 the dimensionless curves Φ, Π and η are independent
of Re and Ma
⇒ ṁturb (p, Ω, ρin ) = ρin Ω d 3 Φ(Ψ), Pturb (p, Ω, ρin ) = ρin Ω3 d 5 Π(Ψ)
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 7/31
Mathematical model
(m2 + A∞
22 ) ẍ2 = −ρw gS2 x2 −R +F −p S p ∗
| {z } | {z22} | {zd2} | atm{z 2 }
buoyancy radiation diffraction air chamber
Dimensionless variables
p − patm
p∗ = and Ω∗ = Ω/Ωmax
patm
O [x2 ] ≈ O [p ∗ ] ≈ O [Ω∗ ] ≈ 1 ⇒ Similar orders of magnitude for errors
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 8/31
Generator feedback control law
!
rated
b−1
Pgen
Tgen = min aΩ ,
Ω
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 9/31
Generator feedback control law
Constants “a” and “b” are functions of the WEC hydrodynamics and turbine
geometry but not of sea-state
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 10/31
Generator feedback control law
30 30
20 20
10 10
0 0
1000 2000 3000 4000 5000 6000 7000 1000 2000 3000 4000 5000 6000 7000
Ω [rpm] Ω [rpm]
0.8 0.8
Π Π ×100
η η
0.6 0.6
Φ Φ
0.4 0.4
0.2 0.2
0.0 0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 11/31
Biradial turbine power output sensitivity to “a” and “b”
4.0
sea state 3 sea state 7 6.7
0.8 7.5 2k
0.9 5 1k W
3.8 4 k kW 8.3 W
1.0 W 0k
4k W
3.6 W
3.4 07 2k
19.1 kW 3.1 3.50 3.91 W
17.1 2 kW 0 kW kW kW
11.313.3 15.2 8 k 1.87 2.22.69
3.2 6 k 0 kW4 kWW
5.54 7.48 9.4W 1.06 1.47 kW 8 kWkW
kW kW2 kW kW kW
3.0 0.1 0.2 0.3 0.5 1.0 2.0 4.0 0.1 0.2 0.3 0.5 1.0 2.0 4.0
×10
4 ×10
4
a a
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 12/31
Wells turbine power output sensitivity to “a” and “b”
3.6
sea state 3 0 sea state 7
0 000..1.003
3.5
0 . 0.40.3.259.2k12W6kkWkWWkW
4 2
0.61 0.54 k8WkWkWW
k 9.280.2765433.21023..3.37044369258
3.4 kW kWkWkkWkkWWW
3.3 9.20
kW
b
0.61 8.23
3.2 0.54 kW 7 kW
kW 6.3 .27 kW
3.1
0.
0.42 48 kW 5.33 0 kW
kW 4.36 kW
kW
3.0
sea state 11 Annual average
3.5
4.23 1 0 0.22
11186846240...6.789 kW 33..32724.28.4k9.01550.k6.16kk17 kkWkW
3.4 123456503928176 k 8 kWWWWW W
kkkWWWW
3.3 4.23
b
kW
3.2 18.1 3.78
6 kW 3.34 kW
16.2 2. kW
3.1 14.3
3 kW W
5k 2.45 89 kW
kW
3.0 1 2 3 4 5 6 7 1 2 3 4 5 6 7
×10
5 ×10
5
a a
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 13/31
Comparison time-averaged turbine output power and rotational speed
20
13 14 13 14
.51
kW
3.0 25.97 kW 12 3.0 12
W
W W
Gen. law clipping
11 62
3
9
5
1
.2
.0
.9 4kkk
8 11
W
23.24 kW 80
10
2
4
62
3.9
.7
.6
.5
.3
k
Turbine power
2.5 2.5
8.3
.80
6k
15
W
15 10 10 17.66 kW
.0 17.78 kW 13
6.8
2.0 4 2.0 .9
12
H s [m]
H s [m]
kW
4.6
9 4 9
.0
5k
kW
8
5k
kW
W
W
8 12.31 kW 8
1.5 1.5
7 9.58 kW 7 10.22 kW
6 6
2 3 5 2 3 5
1 4 4
0.
.3 6.51 kW
93
1.0 1 9 1.0 1
kW
kW
4.12 kW
2.79 kW
0.5 0.5
6 8 10 12 14 16 6 8 10 12 14 16
Tp [s] Tp [s]
14 14
237
13 13
d /s
479
ra
462 rad/s
rad
998 rad/s
7
3.0 12 3.0 12
68
rad
417 rad/s
/s
894 rad/s
/s
/s
32 372 rad/s 11 58
2 raddd/s/s
ra
9ra 11
Rotational speed
7 66
1
2
3
4 776
ra
2.5 d/ 2.5 791 rad/s
s
282 rad
/s 10 10
2.0 2.0 583 rad/s
H s [m]
9
H s [m] 9
19
37
1
6
ra
ra
8 8
d/
d/
1.5 1.5
s
s
10
16
7 7
1r
8r
6 6
ad
ad
2 3 5 2 3 5
4 4
/s
/s
1.0 15 1.0 1
6
ra
d/ 146 rad/s 272 rad/s
s
0.5 0.5
6 8 10 12 14 16 6 8 10 12 14 16
Tp [s] Tp [s]
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 14/31
PART 2 - Optimal control of the turbine High-Speed Stop Valve (Latching)
HSSV
rotor
The HSSV valve of a biradial turbine with 0.5 m rotor diameter HSSV movie
closing the HSSV does not stop the relative motion between the OWC
and buoy
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 16/31
Changes to the mathematical model of Mutriku
Buoy: (m1 + A∞ ∞
11 ) ẍ1 + A12 ẍ2 = −ρw gS1 x1 − R11 − R12 + Fd1 +patm S2 p
∗
OWC: A∞ ∞
−R21 − R22 +Fd2 −patm S2 p ∗
21 ẍ1 + (m2 + A22 ) ẍ2 = −ρw gS2 x2 |
| {z } {z } | {z } | {z }
buoyancy radiation diffraction air chamber
ẋ = f(t,x,u) (1)
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 17/31
Optimal control and the Pontriagyn Maximum Principle
Optimal control
Pturb (t,x,u)
Example: L (x,u) =
ρatm Ω3max d 5
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 18/31
Solution of the Pontriagyn Maximum Principle
for t ∈ [ 0,Tf ]
For complete details see the two classical books of Luenberger [5] and
Bryson & Ho [6]
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 19/31
Receding Horizon control in a real-time framework
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 20/31
Hardware-in-the-loop tests at Tecnalia test rig
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 21/31
Hardware-in-the-loop tests at Tecnalia test rig
Hs = 2 m Hs = 4 m
0.35
∆TRH = 24 s
0.30 ∆TRH = 16 s
∆TRH = 10 s
0.25
NL (No latching)
0.20
L∗c [−]
0.15
0.10
0.05
0.00
Te [s]
ε∗c = L∗c − L∗c |NL / L∗c |NL [%]
120
100
80
60
40
20
0
8 9 10 11 12 13 14 15 16 8 9 10 11 12 13 14 15 16
Te [s] Te [s]
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 22/31
Hardware-in-the-loop tests at Tecnalia test rig
3.5
3.0 ∆ TRH = 24 s
2.5
∆ TRH = 10 s
[− ]
2.0
No latching
Pturb
1.5
∗
1.0
0.5
0.0
0.8
0.6
0.4
ηt [− ]
0.2
0.0
-0.2
-0.4
Time [s]
0.2
p∗ [− ]
0.0
-0.2
-0.4
open Time [s]
u [− ]
close
Time [s]
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 23/31
Lessons learned from the Tecnalia hardware-in-the-loop tests
To improve stability, all state variables must have the same order of magnitude
The discrete nature of the numerical solution of the PMP problem implies a
2nd-order accuracy
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 24/31
Discontinuous Galerkin Method
where
State, adjoint and control
P variables approximated by Legendre polynomials,
pj (t) such that x = j pj (t)x̃j
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 25/31
Discontinuous Galerkin Method
ẋ − f(t, x, u) = 0
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 26/31
Discontinuous Galerkin Method
0.0120 0.0120
0.0115 0.0115
0.0110 0.0110
0.0105 0.0105
0.0100 0.0100
∆t = 0.001 s ∆t = 0.001 s
0.0095 ∆t = 0.005 s 0.0095 ∆t = 0.005 s
0.0090 ∆t = 0.010 s 0.0090 ∆t = 0.010 s
0.0085
∆t = 0.050 s 0.0085
∆t = 0.050 s
Power without optimal control ∆t = 0.100 s Power without optimal control ∆t = 0.100 s
0.0080 0.0080
∆t = 0.200 s ∆t = 0.200 s
0.0075 0.0075
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
iteration, n iteration, n
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 27/31
Discontinuous Galerkin Method
open
u [− ]
close
0.4 ε = 0.0001
ε = 0.05
0.2 no control
p ∗ [− ]
0.0
0.2
0.4
t ime, t [s]
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 28/31
Major conclusions
Receding horizon optimal latching algorithm can be used to improve the OWC
spar-buoy capture width
Receding horizon time interval between 10 to 24 s
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 29/31
Acknowledgements
The work was funded by the Portuguese Foundation for Science and Technology
(FCT) through IDMEC, under LAETA Pest-OE/EME/LA0022, by WAVEBUOY
project, Wave-powered oceanographic buoy for long term
deployment, PTDC/MAR-TEC/0914/2014, and also by European Union’s
Horizon 2020 research and innovation programme under grant agreement No
654444 (OPERA Project).
The first author was supported by FCT researcher grant No. IF/01457/2014.
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 30/31
References
[5] D. G. Luenberger, Introduction to dynamic systems : theory, models, and applications, J. Wiley
& Sons, New York, Chichester, Brisbane, 1979.
[6] A. E. Bryson, Y.-C. Ho, Applied optimal control: optimization, estimation, and control, Blaisdell
Publishing Company, Waltham, 1969.
2017 Maynooth University Wave Energy Workshop / Control of the PTO system of OWCs: feedback vs model predictive control 31/31