Voltage Instability Detection Using Neural Networks

Voltage Instability Detection Using Neural Networks

Adnan Khashman (Near East University, Turkey), Kadri Buruncuk (Near East University, Turkey) and Samir Jabr (Near East University, Turkey)
Copyright: © 2009 |Pages: 7
DOI: 10.4018/978-1-59904-849-9.ch234
OnDemand PDF Download:
No Current Special Offers


The explosive growth in decision-support systems over the past 30 years has yielded numerous “intelligent” systems that have often produced less-than-stellar results (Michalewicz Z. et al., 2005). The increasing trend in developing intelligent systems based on neural networks is attributed to their capability of learning nonlinear problems offline with selective training, which can lead to sufficiently accurate online response. Artificial neural networks have been used to solve many problems obtaining outstanding results in various application areas such as power systems. Power systems applications can benefit from such intelligent systems; particularly for voltage stabilization, where voltage instability in power distribution systems could lead to voltage collapse and thus power blackouts. This article presents an intelligent system which detects voltage instability and classifies voltage output of an assumed power distribution system (PDS) as: stable, unstable or overload. The novelty of our work is the use of voltage output images as the input patterns to the neural network for training and generalizing purposes, thus providing a faster instability detection system that simulates a trained operator controlling and monitoring the 3-phase voltage output of the simulated PDS.
Chapter Preview


Artificial Neural Networks have been used to solve many problems obtaining outstanding results in various applications such as classification, clustering, pattern recognition and forecasting among many other applications corresponding to different areas.

Power system stability is the property of a power system which enables it to remain in a state of equilibrium under normal operating conditions and to regain an acceptable state of equilibrium after a disturbance. Beyond a certain level, the decrease of power system stability margins can lead to unacceptable operating conditions and/or to frequent power system collapses (Sjostrom M. et al., 1999) (Ernst D. et al., 2004). In 2003 and within less than two months, a number of blackouts happened around the world, affecting millions of people. These blackouts include (Novosel D. et al., 2004):

  • The 14th of August blackout in Northeast United States and Canada, which is considered one of the worst blackouts in the history of these countries, affecting approximately 50 million people.

  • The 28th of August blackout in London, which affected commuters during the rush hour.

  • The 23rd of September blackout in Sweden and Denmark, which affected approximately 5 million people.

  • The 28th of September blackout in Italy, which is considered the worst blackout in Europe ever, affecting approximately 57 million people.

In recent years voltage instability has been one of the major reasons for blackouts, and it is the root cause of the 14 August blackout. Voltage stability is threatened when a disturbance increases the reactive power demand beyond the sustainable capacity of the available reactive power resources. Although, progress in the areas of communication and digital technology has increased the amount of information available at the efficient supervisory control and data acquisition (SCADA) systems, however, during events that cause outages, an operator may be overwhelmed by the excessive number of simultaneously operating alarms, which increases the time required for identifying the main outage cause and then starting the restoration process (De Souza A.C.Z. et al., 1997) (Lukomski R. & Wilkosz K., 2003). Additionally, factors such as stress and human error can affect the operator’s performance; thus, the need for an additional tool to support the real-time decision-making process which currently exists. This tool can be in the form of an intelligent voltage instability detector.

Key Terms in this Chapter

SCADA: A system that performs Supervisory Control and Data Acquisition, independent of its size or geographical distribution.

Back propagation Algorithm: Learning algorithm of ANNs, based on minimizing the error obtained from the comparison between the outputs that the network gives after the application of a set of network inputs and the outputs it should give (the desired outputs).

Voltage Instability: Voltage instability analysis is concerned with the inability of assessing the power system to maintain acceptable voltages at all system buses under normal conditions and after being subjected to disturbances (Kundur P. et al., 2004). A major factor contributing to voltage instability is the voltage drop that occurs when active and reactive power flow through inductive reactance of the transmission network. Voltage instability can be caused when a disturbance increases the reactive power demand beyond the sustainable capacity of the available reactive power resources.

Artificial Neural Networks (ANN): A network of many simple processors (“units” or “neurons”) that imitates a biological neural network. The units are connected by unidirectional communication channels, which carry numeric data. Neural networks can be trained to find nonlinear relationships in data, and are used in various applications such as robotics, speech recognition, signal processing, medical diagnosis, or power systems.

Iterations: The number of epochs or repetitions of presenting a neural network with training input/output data.

Learning Coefficient: A numerical value that defines the learning capability of a neural network during training.

Power Distribution System (PDS): Systems that comprise those parts of an electric power system between the sub-transmission system and the consumers’ service switches. It includes distribution substations; primary distribution feeders; distribution transformers; secondary circuits, including the services to the consumer; and appropriate protective and control devices.

Momentum Rate: A numerical value that defines the learning speed of a neural network during training.

Blackout: A power outage (complete collapse), a large-scale disruption in electric power supply.

Pixel: A pixel (short for picture element, using the common abbreviation “pix” for “picture”) is a single point in a graphic image.

Complete Chapter List

Search this Book: