Transcript MSU_ESRDC_26May2009_Overview

MSU and ESRDC
Collaborative Efforts
Noel N. Schulz and Colleagues
1
MSU – ESRDC Team
•
•
•
•
•
•
•
•
Dr. Sherif Abdelwahed, Assistant Professor, ECE
Dr. Stan Grzybowski, Mississippi Power Endowed Professor, ECE
Dr. Herb Ginn, Associate Professor, ECE
Dr. Anurag Srivastava, Research Assistant Professor, ECE
Dr. Suresh Srivastava, Visiting Professor, IIT-Delhi
Dr. Noel Schulz, TVA Endowed Professor, ECE
Dr. Stephanie Doane, Professor, Psychology
Dr. Tomasz Haupt, Research Professor, Center for Advanced
Vehicular Systems
2
Outline
• MSU and ESRDC Collaborative Efforts
– AC-DC Converter for MVDC PHIL Demo (MSU, FSU, USC)
– System-Level Converter Control (USC, MSU-Power Electronics,
Control, and Human Systems)
– Simulation Environment for Dynamic Modeling & Simulation (MSU,
UT)
– Human-System Engineering
– High Voltage and Materials
– Power System Activities including MVDC
– Protection Research Activities for Shipboard Power Systems
• Transition Plan
3
MSU ESRDC Activities
Power
Electronics
Human
Systems
Controls
Power
Systems
High
Voltage and
Materials
Computational
Tools and
Visualization
4
MSU ESRDC Activities
Power
Electronics
Human
Systems
Controls
Power
Systems
High
Voltage and
Materials
Computational
Tools and
Visualization
5
AC-DC Converter for MVDC
PHIL Demo at FSU
•
•
The three PCM-4 phase currents during the step in the
input-current reference of PCM-1
•
•
•
Red Team (POC): MSU (Ginn)
Status: low power testing completed at 480V AC up to 850V DC
at 50kVA
Next steps: interface MSU controller to HIL
•
•
Using AMSC PEBBs
Utilizing the previously developed
multi-functional control for AC/DC
converters
Represents PCM-4 converter in the
MVDC PHIL setup
Working with USC and FSU to
develop a HIL interface for the MSU
controller
6
Systems of Multifunctional Power
Electronic Converters
A current research focus area at MSU is: Development of control methods that enable the
coordinated operation of distributed systems of multi-functional power electronic
converters.
•
This work leverages the multi-functional converter research conducted during the past two
years at MSU as well as builds upon the initial work conducted on systems of paralleled
converters.
The current efforts in this direction are:
•
Development of control strategies and
algorithms for multiple converters distributed in
a system considering the level of centralized
versus distributed control .
•
Investigate methods of control reconfiguration
of power electronic converters in order to
achieve the mission of the interconnected system
of converters.
Converter and Load Test-bed at MSU.
Lab tour available on Wednesday and Thursday!
System-level Converter Control in
Distributed Electric Networks
Challenges
• Continual balance between supply and demand under variable
compensation objectives.
• Effective and flexible management of energy flow throughout
shipboard distribution systems.
• Distributed implementation to avoid single points of failure, and
that is robust, and expandable.
Two approaches are under investigation:
– Multi-agent Systems (MAS)
This work is being performed in collaboration with USC.
– Distributed Model-Based Control
System-level Converter Control in
Distributed Electric Networks
Distributed Model-Based
Approach Objectives
Multi-Agent Approach
Objectives (with USC)
• Define appropriate multi-level
abstractions to represent the
dynamics of coordinating
multifunctional converters.
• Develop efficient system-level
control policies that can be
analyzed for stability and
convergence properties.
• Develop integrated system
modeling, analysis, and control
synthesis tools.
• Determine the agent functions
required to coordinate
multifunctional converters
• Assess different functionality
partitions among the device
controller and agent system
level device manager.
• Develop a method to share a
burden among converters with
delocalized knowledge of
absolute and current loading
conditions
Multi-Agent Approach
Approach 1: External PEBB Agent (Current)
•
The approach of agents external to the converter
has been implemented and some preliminary
testing has been conducted
Agents communicate with PEBB controllers via
Ethernet
•
System Control
Layer
System Level Control
(Mission States, etc)
Multiple Converter
Coordination Layer
Agent Based Current
Component Selection
Reference Signal
Generation
Application
Control Layer
Converter Control
Layer and Lower
Layers
External to Converter
Local to Converter
Two possibilities for Approach 1 have been investigated:
•
-
Case 1 – elaboration at PEBB’s side
The PEBB Agents, one per converter,
receive from the converter elaborated data.
The Agents receive the terms of the current
decomposition that is performed within the
PEBB control architecture. The Agents then
exchange information and make decisions
based on these quantities. Once the
decision is made, the Agent implements it
by sending the appropriate current
reference coefficients to the PEBB
•
-
Case2 – elaboration at Agent’s side
The control Agents receive voltage and
current data from the PEBB and then
proceed to further elaboration. The Agents
then exchange information and make
decisions based on the results of this
elaboration. Once the decision is made, the
Agent implements it by sending the
appropriate current reference coefficients to
the PEBB
Approach 2: Internal PEBB Agent (Future)
•
The results of the external PEBB agent study will be used to implement agents locally
within the PEBB controller
Experimental PEBB Agent Setup
•
•
The test setup incorporates four PEBB converters, two in
a back-to-back configuration to simulate energy storage
An additional load bank imposes non-linear and
unbalanced conditions on the AC system
Converters
1&2
Load Bank
Mini-dc link
Demo with
Lab Tour
MSU ESRDC Activities
Power
Electronics
Human
Systems
Controls
Power
Systems
High
Voltage and
Materials
Computational
Tools and
Visualization
12
Human-System Interface (HSI)
Design, Test, and Evaluate HSI that Optimizes Human-System Performance
•
•
•
PEBB
Agent #1
Optimize energy manager understanding of PEBB agent actions during system monitoring
Understanding predicts user ability to remain in-the-loop and intervene when required
Initial design in progress, T&E to be completed in coming year
PEBB
Agent #2
Socket Communication
HSI
PEBB
Agent #3
MSU ESRDC Activities
Power
Electronics
Human
Systems
Controls
Power
Systems
High
Voltage and
Materials
Computational
Tools and
Visualization
14
Abstract Modeling of Distributed Converter
System
•
•
New component-based
modeling for distributed
converter systems.
Modules:
– Generator
– Load (dynamic,
reconfigurable)
– Energy storage
– Converter (AC/DC, DC/AC)
– Converter control structure
•
Links:
– Power transmission lined
– Measurements
– Data/control lines
6-pulse AC/DC
Converters
AC
System 1
AC
System 2
configurable
load
u,i
u,i
Bi-directional
Voltage
Source
Converter
Bi-directional
Voltage
Source
Converter
c
Energy
Storage
u,i
c
Bi-directional
Voltage
Source
Converter
c
DC bus
15
Distributed Model Predictive Control (MPC)
•
•
•
•
MPC: at each step, a finitehorizon optimal control
problem is solved but only one
step is implemented.
In distributed MPC a global
controller manages intercomponent interaction and
enforces global requirements.
Abstract representation of the
components is used for highlevel control decisions.
Global control actions are
given as additional constraints
for local controllers.
Predicted
future states
Abstract
System
Model
Global Controller
(coordinator)
6-pulse AC/DC
Converters
System
measurements
AC
System 1
Local control
commands
(constraints)
AC
System 2
configurable
load
u,i
u,i
Bi-directional
Voltage
Source
Converter
c
Energy
Storage
Local
Control
structure
c
u,i
Bi-directional
Voltage
Source
Converter
c
Local
Control
structure
c
Local
Control
structure
Bi-directional
Voltage
Source
Converter
c
c
DC bus
16
Model-driven Engineering
Design data, constraints, operational
requirements, and platforms specs
Level of Abstraction
Model data
Design
Constaints
, requirements
Generated
Application
Code
Code
Platform
Platform
Frameworks
Active state
models
Measurement
support
Control Libraries
Execution Kernel (OS)
Hardware
Domain-specific modeling
languages
Provide support for system
modeling through a
composition of predefined
component templates
Domain-independent
modeling languages
• State Charts
• Simulink models
Provide support for
system analysis,
testing, and
simulation through
formal models
Research goal is to provide tools for transforming
design specification into formal models that can be
used for analysis, simulation, automatic control
synthesis , code generation and deployment.
MSU ESRDC Activities
Power
Electronics
Human
Systems
Controls
Power
Systems
High
Voltage and
Materials
Computational
Tools and
Visualization
18
Simulation Environment for
Dynamic Thermal Modeling and Simulation
(DTMS) Framework
•
T. Haupt, G. Henley, B. Parihar
• Mississippi State University
• in collaboration with
•
T. Kiehne and M. Pierce
• University of Texas at Austin
Simulation Environment for DTMS
The DTMS Framework is a modeling and simulation
platform designed for system-level thermal
simulations of indefinitely large shipboard systems.
DTMS supports a customizable input/output system
designed to allow it to be easily integrated with
external platforms and systems.
MSU developed a GUI (Graphical User Interface)
and a runtime environment for setting up simulations,
running DTMS and visualizing the results.
An input language was designed to serve as an interface
between the GUI and DTMS,
A parser for this language was written and included in
DTMS,
 An input file generator was written for the GUI to create the
input file based on the information in the database plus
parameters or object state information input by the user,
The runtime environment and GUI for Fire&Smoke has been
adapted for DTMS.
Architecture of the system
Thermal Model Graphical User Interface
CAD Processing
Database
dxf
CAD file
conditioning
CAD data
extraction
Geometry
and
properties
Once the CAD Processing is
complete and the geometry is
acceptable, the database
contains the necessary
information to run
simulations.
Properties
Output
(Simulation
Results)
Input
(Simulation
Setup)
DTMS
(Third-party Dynamic Thermal Modeling
and Simulation Framework).
DTMS GUI features
 3D ship view with rotate, zoom,
translate, and deck separation
feature for better interior views.
 Point and click operation for most
simulation setup
 Automatic generation of input
decks
 User-selected display for various
objects, compartments, and
decks.
 User-defined color maps and
legends.
 Real time visualizations
Poster with Demo and Lab Tour on
Thursday
Test case:
a simplified A/C plant
MSU ESRDC Activities
Power
Electronics
Human
Systems
Controls
Power
Systems
High
Voltage and
Materials
Computational
Tools and
Visualization
23
Human-System Engineering
•
Integrate Human-System Interface (HSI) and Tool Design Process to Facilitate Tool
Transition to Fleet
– Design power management tools and HSI in tandem
– Test and Evaluate in tandem
Aided interface
•Experiments performed
that asked ECE students to
redistribute power using
an unaided or an aided
interface.
•The interfaces were
similar except for DSS
advice
•User interactions with the
interface were recorded
•Scoring rules designed to
measure reconfiguration
quality applied to each
user’s aided and unaided
reconfigurations
Scores for aided and unaided interfaces of matched
problems (for 10 users)
Mean
Unaid
ed
5.5
1.4
3.5
3
2.2
5
20.6
3.4
S.D.
Aided
2.5
5.2
4.5
3.3
4.2
1.5
S.D.
Unaided
10
2.8
2.2
3.3
2.5
2.4
2.3
8
9
Mean
Aided
7
6
Total score
Matched
Mean
problem
Aided
s
1
8
2
5
3
4
4
6.5
5
8
6
9.5
Total
41
Overall
6.8
Mean
Mean
Unaided
5
4
3
2
1
0
1
2
3
4
Matched
Problems
5
Poster and Tour/Talk on Thursday
25
6
MSU ESRDC Activities
Power
Electronics
Human
Systems
Controls
Power
Systems
High
Voltage and
Materials
Computational
Tools and
Visualization
26
Accelerated Electrical Degradation of Machine Winding Insulation
•Power electronic devices cause
voltage spikes and harmonics
that reduce the machine
winding electrical insulation
strength
•Accelerated degradations of
machine winding insulation was
obtained by high frequency
pulse
voltage
and
high
temperature stresses
•Evaluation
of
electrical
properties was performed for
samples before and after 100,
500, and 1000 hours of
accelerated degradation at 20
kHz, and 40 kHz square pulse
frequency and 90% of rated
temperature
Dielectric Test System
0 to 3500 V, 40 kHz, 20ºC to 260ºC
Pulse waveform
27
Accelerated Electrical Degradation of Machine Winding Insulation
Partial discharge inception voltage (PDIV) as a function of time
Aging conditions: a) 1300 V square pulse at 40 kHz
MW 35-C at 180 ºC, and MW 80-C at 140 ºC
b) 1350 V square pulse at 20 kHz
MW 16-C at 216 ºC, and MW 73-C at 200 ºC
Breakdown Voltage as a function of time
Aging conditions: a) 1300 V square pulse at 40 kHz
MW 35-C at 180 ºC, and MW 80-C at 140 ºC
b) 1350 V square pulse at 20 kHz
MW 16-C at 216 ºC, and MW 73-C at 200 ºC
28
Measurement of Partial Discharge in Machine Winding Insulation
During 50 ns Rise Time Pulse Voltage
1300V Square Pulse
<40 ns Rise Time
Measurement System
Vm
Z
Sample 2
Z
Sample 3
Pulse Generator
Z
Sample 1
Samples are
Inserted in the
Cylindrical
Shielding
Sample Shielding
Diff. Probe Ch.1 1 GHz
Diff. Probe Ch.2 Oscilloscope
Outer Shield
Ch.3
Heat
Injection
1500
1000
500
0
-500 0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
-400 0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
600
400
200
0
-200
-400 0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
400
Chan. 2
Pulse
Source
System
Conn.
Chan. 1
Generator and Probes
Wire Connectors:
Ground
Pulse Voltage
200
0
-200
time (ms)
Heat Path
Measurement of Partial Discharge in Machine Winding Insulation
During 50 ns Rise Time Pulse Voltage
Partial Discharge Patterns
1300 V Square Pulse, <40 ns Rise Time
Partial Discharge at 20C
Partial Discharge at 150C
30
Electrical Degradation of 15 kV XLPE and EPR
Cables Energized by Switching Impulses
250/2500 μs of 100 kV
Cross-section of XLPE and EPR cables
Electrical Degradation of 15 kV XLPE and EPR Cables
Energized for 100, 500, 1000, 5000 and 10000 switching impulses
(a)
(b)
Inception voltage of partial discharge: (a) EPR cable samples
(b) XLPE and EPR cable samples
Posters and Tours on Wednesday and Thursday
32
MSU ESRDC Activities
Power
Electronics
Human
Systems
Controls
Power
Systems
High
Voltage and
Materials
Computational
Tools and
Visualization
33
Power System Research Activities Related to
SPS Including MVDC System

Objective: To increase the survivability of SPS with increased
security and reliability
 Key Issues:
– Analysis of MVDC system
– Reconfiguration

- Stability
- Protection
Approach at MSU:
– MVDC (Architecture, small signal & transient stability, optimal voltage and
power control)
– Stability (Tool development for index based and continuation power flow)
– Reconfiguration (Survivability index, cognitive analysis and uncertainty)
– Protection (Protection scheme, relay model, hardware in the loop with
multiple platform)
Stability of Shipboard Power System

Index based approach
 Continuation power flow based approach
• Motivation

Conventional tool used for stability analysis of utility can not be extended for
SPS due to pulse load, EMALS, AC/DC, 3 phase unbalanced network
 Need tools for electric ship voltage stability assessment to investigate and
study stability margin and comparing different SPS architecture
• Objective
 Index based: To develop fast and robust algorithm to find out static and
dynamic voltage stability index for all-electric ship to calculate approximate
stability margin
 Continuation power flow: To develop a tool to accurately calculate stability
margin for setting up benchmarks for index based approach of voltage
stability
Index based approach for Stability of SPS
Load
factor
Minimum Eigenvalue
1.0
5.709206867
Eigenvalue based
1.1
5.627212602
sensitive analysis
1.2
5.529382811
1.3
5.413750157
Second order
performance index based
on hessian matrix by
calculating derivative of
Jacobean matrix
1.4
5.277671071
1.5
5.117537598
1.6
4.928309384
1.7
4.702347895
1.8
4.428685008
1.9
4.08863468
Dynamic stability index is
being developed
2.0
3.648121503
2.1
3.031325541
2.2
1.97980143
2.25
0.639127222
2.3
4.328855049
• Approach



IEEE 9-Bus
Continuation Power Flow for Stability of SPS

Approach:
– Predictor-Corrector based
– Tangent vector based
– Pseudo arc length and local
parameterization
– Adaptive step length
Healy System
G
G
1
G
2
G
4
3
6
5
7
10
8
9
11
1
2
13
16
14
17
15
18
Stability Studies of Shipboard Power
System for MVDC architecture.
Motivation:
•
The adoption of a new architecture
(MVDC) for a shipboard power system,
requires investigating the stability of the
system under various system faults or
disturbance conditions.
Results:
•
•
Work carried out:
•
•
•
Detailed analysis of small signal stability
of the system has been carried out.
Participation analysis used to identify the
controls required to improve the stability.
Currently studying the transient stability.
•
•
Small signal stability based on
Eigenvalue analysis has been carried out
on MVAC as well as MVDC
architectures of the zonal shipboard
power system.
Participation analysis reveals that the
excitation/load bus voltage states are
contributing maximum to the critical
modes.
Lower order exciter control (type-0
exciter) exhibited poor damping with
MVAC and MVDC architectures.
Use of Higher order (IEEE type-1)
exciter or shunt compensating device
like SVC improved the damping of the
system.
Summary of Small Signal Stability Results
Table: Few Eigenvalues close to imaginary axis
39
Protection Research Activities for Shipboard
Power System




Development of protection scheme
Relay modeling using NI, dSpace, RTDS
Hardware in the loop with multiple platform
Multiple relay operation and coordination
• Work Completed

Developed multiple hardware in the loop platform using dSpace, National
Instruments and RTDS

Relay modeling and prototyping using Simulink & LabView for overcurrent
and differential relay

Multiple relay coordination for shipboard power system protection
Protection Research Activities
Shipboard Power
System (SPS)
RTDS
VTB-RT
Relay
Prototype
Relay model
PXI
dSPACE
Shipboard Power
System (SPS)
RTDS
VTB-RT
Relay
Hardware
Schweitzer Relays
(SEL- 421, SEL- 351)
Lab tours available Wednesday
and Thursday!
Planned Tasks for 2009-2010

–
–
–
–
–
–
–
–
–
Major planned tasks:
Development of dynamic voltage stability index
Development of continuation power flow for
distribution system based on current injection
Reconfiguration using bacteria foraging optimization
User interface for developed reconfiguration
algorithm for SPS
Multiple relay operation using RTDS and SEL and
GE relay
Integrating reconfiguration technique with controller
in the loop to perform real time reconfiguration
Detailed Converter dynamics to be considered, while
performing small signal stability analysis.
Transient stability analysis considering faults at AC
and DC bus.
Design of supplementary controllers to Voltage
Source Converters(VSCs), if required to improve the
system stability.
Intelligent reconfiguration
and stability index algorithm
Restored load with
optimized survivability
Transition Plan for MSU ESRDC Team
• Herb Ginn will transition to the MSU Representative on the
ESRDC Board of Directors
• Noel Schulz will continue as a researcher on the MSU team for
one year from Kansas State.
• MSU has hired a new faculty member in the power systems
area, Yong Fu from IIT-Chicago
• An additional hire is expected in 2009-2010
43