01-12-2012, 02:35 PM
Voltage Stability Assessment Using Neural Networks in The Deregulated Market Environment
1Voltage Stability Assessment.pdf (Size: 344.06 KB / Downloads: 22)
Abstract
The project was done in Nanyang technological
university, Singapore at Electric power research Lab during the
period from Sep 2006 to Nov 2007. The liberalization of electric
market has generated the unbundling of large utilities into
separated generation, Transmission and Distribution
companies with subsequent changes in the operating condition
at electric power systems. Most of the breakdowns in the
electric power systems are caused by unfavorable dynamic
response of the networks following system disturbances. In
addition, environmental and economic consideratioNS are
forcing power systems to be operated closer to their limits of
stability .Therefore dynamic security assessment of power
systems is becoming increasingly important. Voltage collapse is
one of the instability which occurs when the system is heavily
loaded. In addition, the mechanism of voltage collapse is due to
the load reacting to the voltage changed which leads to further
voltage changes. In this project, indicators utilizing the
generator-load mismatch and active/reactive power margins for
on-line assessment of proximity of voltage instability of power
system is to be developed.
INTRODUCTION
The primary purpose of the system generation control is to
balance the total system generation against load demands and
losses, so that the desired frequency and power interchange
with neighboring systems (tie flows) is maintained. In this
structure, in addition to the controllers operating directly on
individual system elements such as the generators, there is
usually some form of overall plant controller that coordinates
the controls of closely linked elements. The plant controllers
are in turn supervised by system controllers at the operating
centers.
METHODOLOGY:
A P-V curve method which is a traditional static voltage
stability analysis method is a kind of computation analysis
method based on the physical conception. Setting the result
of the basic system power flow, increasing the system load
gradually, and calculating the system voltage corresponding
to each operating point, a series of (P,V) points reflecting the
relationship between the load power actually absorbed and
the node voltage can be obtained and a P-V curve can be
formed . The inflection point of P-V is considered as the
voltage stability curve, the area above the inflection point is
the voltage stability region and the area below is considered
as the voltage instability region. The distance from the
current system operating point to the inflection point(Y.
Kataoka,1994) is known as the system voltage stability
region. For a definite five bus power system we can use the
Newton Raphson method in multi bus system.
CONCLUSION:
Most of the ANN applications to SA and SE, reported in
the literature, are of explanatory nature. They attempt to
determine the feasibility of using different ANN
architectures and determine which inputs and types of
training produces the most accurate results. This study can be
considered to be similar to the selection of the good selection
of security indices or approximate system performance (ASP)
models for SA. That is finding what ASP is fast and accurate.
Although the computational speed is a non issue when using
ANNs training time is a crucial issue in power system
security assessment. The experience acquired from the
development of the ANN based model strongly indicates that
the ANN technology has matured enough to be applied
successfully in many power system problems.