美国几大牛完成的最新项目报告

wxli · 发表于 2009-1-26 06:05:25

马上加入，结交更多好友，共享更多资料，让你轻松玩转电力研学社区！

您需要登录才可以下载或查看，没有账号？立即加入

×

venkatasubramanian_pserc_s-29_project_report_2008.rar (4.34 MB, 下载次数: 81)

liuhd405 · 发表于 2009-1-26 22:30:34

怎么没有介绍一下具体的内容啊

p0o9i8u7 · 发表于 2009-1-27 04:48:57

请问楼主是否可以详细的介绍一下报告的主要内容，谢谢

郑昌伟 · 发表于 2009-1-27 20:28:53

应该不错吧,下来看看,比较前沿了

cosmic_tao · 发表于 2009-1-27 21:25:07

请楼主补充一下内容。：）

rocketsun · 发表于 2009-1-28 11:04:14

Executive Summary
The risk of cascading outages in power systems manifests itself in a number of ways.
With the advent of structured competitive power markets, and with the lack of needed
investment in the transmission grid, electric power systems are increasingly being
operated close to their limits. Potential terrorist threats raise concerns about power system
being placed in unforeseen operating conditions. Recent blackouts in North America and
in Europe show that the risk of cascading events is real with the cost of the 2003 North
American blackout estimated to be around $10 Billion. As the power system becomes
more stressed there are a host of reasons (such as weak connections, unexpected events,
hidden failures in protection system, and human errors) for the loss of stability eventually
leading to catastrophic failures.
When a power system is subjected to large disturbances control actions need to be taken
to steer the system away from severe consequences and to limit the extent of the
disturbance. This is particular true if system is in an operating condition that makes it
unusually vulnerable to catastrophic failure. In a previous PSERC project (S-19 that
ended in 2005), we developed novel algorithms for each of the following steps:
1. Detection of major disturbances and protective relay operations leading to
cascading events. The detection algorithms improved capabilities of real-time
fault detection and analysis to classify the impact of a fault towards initiating
cascading outages.
2. Wide-area measurement based detection and remedial control actions. The
wide-area mitigation algorithms include methods for reliably extracting modal
information on critical wide-area modes through real-time wide-area Phasor
Measurement Unit (PMU) measurements. They also provide specific control
actions to damp out the oscillations when problems are detected.
3. Adaptive islanding with selective under-frequency load shedding. The
adaptive islanding algorithms suggest methods for controlled islanding of the
system should mitigation strategies fail.
The algorithms were shown to be effective using realistic computer models of test power
systems. In this project, we focused on prototype implementations of those algorithms at
collaborating PSERC member utilities.
Volume I: Detection of major disturbances and protective relay operations leading
to cascading events (Lead: Mladen Kezunovic, Texas A&M University).
This research volume proposes a system-wide monitoring and control solution intended
for use at control centers. The solution includes steady state and dynamic analysis tools.
The tools may be used to check system stability, find the vulnerable elements, and send
commands to the local tools for initiating detailed monitoring.
The local monitoring and control tools are intended for installation at substations. They
have the capability for advanced fault analysis and relay monitoring using neural network
ii
for fault detection and classification, synchronized sampling for fault location, and event
tree analysis for relay operation verification. The local monitoring and control tools can
characterize a disturbance and make a correction if there is an unintended relay operation.
This information can be sent to a system monitoring and control tool for better security
control.
Implementation of the proposed algorithms requires assessment of data handling
requirements. Data handling includes obtaining, converting and storing data.
Consideration must also be given to integrating data from various sources: Supervisory
Control and Data Acquisition (SCADA) systems, Intelligent Electronic Devices, and
other add-on high speed data acquisition systems. Data from system simulation packages
is also needed to perform steady state analysis.
Steady state and dynamic analysis cases were studied using a model of the entire Entergy
system. Vulnerability Index and Margin Index for each bus, generator, and transmission
line were calculated to identify the vulnerability and security margin information for each
individual power system element. A 500kV Extra High Voltage transmission line was
modeled for evaluating local monitoring and analysis tools. The results indicate that
when using adequate training sets and time synchronized data, the new algorithms are
quite accurate and provide assessment of system conditions during cascades that are not
feasible with any of the existing techniques.
Volume II: Wide-area measurement based detection and remedial control actions
(Lead: Mani V. Venkatasubramanian, Washington State University)
If persistent over an extended period of time (e.g., 30 minutes), poorly damped
oscillations in a power system can lead to permanent damage of expensive power system
equipment and pose power quality issues. Negatively-damped oscillations can be even
more problematic by resulting in sudden tripping of generators and/or widespread system
blackout such as occurred in the August 10, 1996 western power system blackout.
In this research project, we designed, developed and implemented an Oscillation
Monitoring System (OMS) that uses wide-area PMU measurements for automatically
monitoring for poorly damped and/or negatively-damped oscillatory modes. OMS
includes two complementary engines that provide real-time modal analysis: 1) an
automatic Prony-type analysis of power system responses following routine events such
as line tripping and generator outages; and 2) an engine for continuous estimation of
poorly damped mode frequencies and their damping ratios from routine ambient noise
PMU measurements. The algorithms were structured as a rule-based expert system for
simplifying the Prony analysis of nonlinear PMU responses in real-time while avoiding
false alarms and incorrect estimates. The OMS can issue operator alerts as well as initiate
triggers to start appropriate control actions to improve damping of problematic oscillatory
modes. A wide-area damping control strategy that uses Static VAR Compensators was
developed in the previous PSERC project. Another damping control strategy was
developed in this project. It uses HVDC modulation to improve the damping of inter-area
oscillatory modes.
iii
A prototype version of OMS has been integrated into real-time monitoring capability of
the Phasor Data Concentrator at TVA. At present, it monitors PMU data from within
TVA for estimating the frequency, damping, and mode shape of oscillatory modes. We
are also collaborating with the Electric Power Group and Bonneville Power
Administration (BPA) to implement a version of the Prony-type first engine of OMS at
two PSERC companies: BPA and the California Independent System Operator. An offline
version of three Prony type algorithms from this project has been integrated into an
event analysis software package being developed by Electric Power Group Inc., and the
package aimed at off-line modal analysis of PMU measurements is available to PSERC
members.
Volume III: Adaptive islanding with selective under-frequency load shedding (Lead:
Vijay Vittal, Arizona State University)
The main objective of this portion of the research project was to develop a fast and
accurate assessment tool to determine the timing of conducting controlled islanding
scheme in real time for preventing cascading events that could lead to a large scale
blackout. This work demonstrated an event-initiated controlled islanding scheme using
phasor measurement units and decision trees to stabilize power systems following severe
contingencies. The demonstration was performed on a test system provided by Entergy.
A control scheme for must be designed very carefully since the costs can be quite high
for unneeded controlled separation operation and for failure to operate when needed.
Normally, there are two major issues associated with the design of a controlled separation
scheme:
i. Where to island? This issue mainly focuses on searching for the optimal cut set
in the system to satisfy certain constraints, such as (a) coherent generators should
stay in one island and (b) the load/generation imbalance of each island should be
minimized. Much of the research effort of a previous PSERC project (S-19)
focused on addressing this issue. Due to the large computational burden
associated with this aspect, the controlled separation cut set is usually obtained
offline.
ii. When to island? The second issue is to accurately determine the timing of
controlled separation. This is the main focus of this research. The objective was to
develop an online transient stability prediction scheme to determine whether
certain contingencies can initiate severely unstable swings and cause cascading
events. If so, controlled separation should be initiated to prevent a large scale
blackout.
Taking advantage of synchronized phasor measurements, we developed a decision tree
based transient stability assessment scheme for online application. The synchronized
signals and high sampling frequency give PMUs the capability of observing different
states across the whole system in a common time frame with great accuracy. System
transient behaviors can also be accurately captured compared to the traditional SCADA
system. The decision tree technique is an effective data mining tool to solve the
classification problems in high data dimensions. It can be used to uncover critical system
attributes that contribute to an objective such as transient stability or voltage security. The
splitting rules and the corresponding thresholds in the decision trees help to build a
nomogram in terms of these critical system parameters and guide system operators
effectively.
The project results indicate that out of step swings can be accurately predicted only a few
cycles after the initial disturbance using properly trained decision trees. This is much
faster than the traditional analysis method. The decision trees also identify transient
stability indicators that are good candidates for new PMU locations. With an effective
controlled islanding strategy, the system can be stabilized faster with less amount of load
to be shed than the uncontrolled separation case.
We developed a software platform that implements the entire approach based on data
provides the DSATools suite of software from Powertech Labs. The approach also uses
commercially-available software called CART developed by Salford Systems for
decision tree training and testing.
Next Steps to Advancing this Research
At Texas A&M University, future steps to implementing algorithms for the detecting,
classifying, and mitigating cascading events require selection of a utility test set-up.
Currently discussions with several utilities under auspices of DOE-CERTS, EPRI, and
ERCOT are under way to determine whether such arrangements for the final testing work
can be made.
The Washington State University team will continue to work with PSERC utilities and
software vendors for field implementations as well as testing, tuning and enhancement of
OMS capabilities. Future research needs to focus on determining correct operator actions
as well as automatic control actions to improve the damping of problematic oscillatory
modes when such problems are detected by OMS.
At Arizona State University, future steps to implementing the controlled islanding work
would require testing and implementation at an electric utility. Some of this work is being
done in a CERTS project and discussions are under way to consider implementation at a
specific company.

ghomis · 发表于 2009-1-28 14:14:02

具体是哪几个大牛出的报告啊？

wxli · 发表于 2009-1-29 12:05:11

Mani V. Venkatasubramanian, Project Leader,
Washington State University
Mladen Kezunovic, Texas A&M University
Vijay Vittal, Arizona State University

loadflow · 发表于 2009-1-29 17:24:57

Vijay Vittal还写了HVDC的书。

newseer · 发表于 2009-1-29 21:38:33

提示: 作者被禁止或删除内容自动屏蔽

账号		自动登录	找回密码
密码			立即加入

美国几大牛完成的最新项目报告

马上加入，结交更多好友，共享更多资料，让你轻松玩转电力研学社区！

评分

相关帖子

详细一些