11-06-2012, 05:28 PM
VOLTAGE STABILITY
VOLTAGE_STABILITY_USING_FACTS.doc (Size: 1.64 MB / Downloads: 4)
Abstract:
The impact that artificial neural network technique for the measurement of voltage stability using facts controllers is discussed here. For static voltage stability studies of a power system, the loading of the system is increased incrementally and slowly (in certain direction) to the point of voltage collapse. The MW-distance to this point is a good measure of system voltage stability limit. The voltage profile of the system is shown by the PV-curves which are plotted using continuation power flow programs as the loading varies from the base values to the point of collapse. Most reported studies consider constant PQ (or at best constant impedance, constant current, constant power, called ZIP) models for loads. Given, in average, 60% of the loads include induction motors which do not follow the ZIP-model, the results obtained from these PV curves may be erroneous. In this paper, voltage stability analysis is performed through continuation power flow program that accommodates the model for the induction motor components of the loads. The method is tested using the New England 39-bus Power System. Comparison of the presented method with the commonly used method shows that the voltage stability limits found commonly can be too optimistic. It was found that when induction motor components of the loads are considered, the obtained voltage stability limit is smaller and the overall voltage profile of the system is very much different than that found using only PQ-load models.
Analysis of stability:
The basic stability analysis model is referred in this paper. In addition, this model is extended to accept the following components. Standard types of turbines and governing systems, basic types of automatic voltage regulators (AVRs) including the power system stabilizer (PSS) and finally an aggregate model of an induction machine are also included. In this model the fast transient processes in all machines stator/network is ignored. The elimination of these transient leads to standard singular perturbation form of the description:
Application of the methodology:
The proposed methodology for stability analysis is based on ANNs with complete and reduced sets of input data. It can be tested on high voltage power system about 92 buses 12 number of generator buses and 174 lines. In doing so the following four cases may be considered for the complete input vector of system variables.
Indirect stability assessment (ISA):
The first approach is indirect and is performed so that for each outage the evaluation is first done for the vector of system variables by applying the fast-decoupled load flow calculation. Then on the basis of the vector of system variables and the specified identification code for the topology of the outage in question, the voltage stability margin of the observed outage is tested by applying one of the suggested ANNs.
Direct stability assessment (DSA):
The second approach is direct, and is performed so that the input of the ANN is the input vector y = (P, Q) consisting of actual injected active and reactive powers in all buses and the identification code of the outage.
