Optimal restoration for distribution system using PSO and ANN

Service restoration is an important aspect of power system design that caters the restoration of power to an un-faulted area under blackout after an emergency condition. The power system operators have the principal objective of minimizing the inconvenience occurred to consumers by isolating only the faulty area while providing the power to unaffected areas as much as possible. Their objective of providing service to customers is subjected to further constraints such as distribution system configurations, power available in the network, and the current carrying capacity of the distribution lines or feeders. Once a fault takes place, the number of customers in the blackout area mainly depends upon the effectiveness of the load restoration mechanism. Currently, the power system operators respond by implementing a pre-defined restoration schedule based on the previous human knowledge. While this may serve the purpose of power restoration to some extent. There is typically a large number of feeders with even larger number of switches in a distribution system and it is not humanly possible to restore the service to an out of service area solely based on past experiences. Many algorithms have been proposed to serve the purpose of restoration with each one having certain merits and demerits. This paper presents an effective and globally optimal restoration mechanism using Particle Swarm Optimization (PSO) and Artificial Neural Network (ANN)