on the design and placement of a supplementary damping ... · omar kotb. elkraft 2017. chalmers...
TRANSCRIPT
On the Design and Placement of a Supplementary Damping Controller in an Embedded VSC-MTDC
NetworkOmar Kotb*, Mehrdad Ghandhari*, Robert Eriksson**, Javier Renedo***, Luis Rouco***,
*Department of Electric Power and Energy Systems, KTH, Stockholm, Sweden**Svenska Kraftnät (Swedish National Grid), Sundbyberg, Sweden
***Institute for Research in Technology, Comillas Pontifical University, Madrid, Spain
Presented by:Omar Kotb
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
2
Presentation Outline
• Introduction• Motivation• Power System Model• Controller Design• Conclusions
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
3
Introduction: Embedded VSC-MTDC Network
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
Embedded VSC-MTDC: locatedwithin one synchronous AC grid.
Segmented AC system: VSC-MTDC connecting more thanone asynchronous AC grids.
4
Introduction: Embedded VSC-MTDC Network
AC/DC Power System Model:
• One slack bus
• VSC representedby average model
• Master-slave DC voltage control
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
MTDC Control Imperatives in AC/DC Power System
• Primary control: active&reactive power, DC voltage, AC voltage
• Supplementary control: Power Oscillation Damping (POD), frequency support
• Control coordination: PSS, other FACTS POD
5
Introduction: Embedded VSC-MTDC Network
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
6
Motivation
Supplementary (POD) control not sufficiently investigated as comparedto primary control in VSC-MTDC
Multiple damping controllers: adverse control interactions
Control coordination problem through nonlinear optimization:- Complex objective functions- Time-consuming solutions
Strategic placement of supplementary controller(s)
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
7
MLQG-MISO control allows damping of targeted modes of interest, other modes are not affected
Assumptions:
• Negligible time delays
• PMUs readily available
Motivation
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
SISO control structure
MISO control structure
8
Power System Model
( , , )0 ( , , )x f x y u
g x y u==
- Nonlinear DAE system model:
- Linearized system model:
Linearization conducted around the equilibrium point of the DAE system.
0x y u
x y u
x f x f y f ug x g y g u
∆ = ∆ + ∆ + ∆
= ∆ + ∆ + ∆
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
9
Eigenvalue sensitivity to controller parameter (q):
( )2
( , ) 1|1 ( , )i
Tii s i i
i
F s qR w B Cvq q F qλλ
λ=
∂ ∂= =
∂ ∂ −
Power System Model
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
::
i
i
wv
Mode’s left eigenvector
Mode’s right eigenvector
10
x Ax Bu wy Cx υ= + + Γ= +
- Linearized state-space system model:
Process noise measurement noise:w :υ
- Modal variables:
( ) ( )z t Mx t=
Power System Model
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
11
- Cost function:
( )( )0
limT
T T Tk mT
J E x M Q M x u Ru dt→∞
= +
∫
:mQ:R
Weighting matrix for modal variables
Weighting matrix for controller output
:M Mapping matrix
Feedback control law given by: ˆ( ) (t)u t Kx= −
MLQG Control Design
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
:K MLQG controller gain
12
Constant estimation error feedback matrix, obtained by the solution of Algebraic Ricccati Equation (ARE):
ARE solution is according to cost function for Kalman filter:
( )0
limT
T TL T
J E x Wx u Vu dt→∞
= +
∫
MLQG Control Design
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
obtained using Kalman filter:ˆ(t)x ( )ˆ ˆ ˆ(t)x Ax Bu L y Cx Lυ= + + − +
:L
( )ˆ ˆ ˆ(t)x Ax Bu L y Cx Lυ= + + − +
13
, , , ,,
A B C MW V
Γ
Model
LQG
,mQ R
Results
MLQG Control Design
MLQG POD supplementary controller design merits:
- Enhanced robustness- Targeted damping of specific oscillatory modes
(weighting matrices)- Linear control strategy
14
Results
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
Inter-area modes’ observabilities (normalized).
15
Results
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
Inter-area modes’ locations in complex plane.
16
Results
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
Disturbance: 10% load increase for 100 ms.
Power flow over inter-area AC tie line.
17
Conclusions & Future Work
Research theme: AC/DC power system stability enhancement & controlthrough MTDC supplementary controls
Strategic positioning of a single damping controller within an embedded VSC-MTDC network
Investigation of supplementary (POD) control in case of droop controlused for DC voltage regulation.
Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18
Thank you! Questions?
18Elkraft 2017Chalmers University of TechnologyGothenburg, Sweden2017-05-18