New algorithm based on CLPSO for controlled islanding of distribution systems

Controlled islanding operation of distribution systems having significant penetration of distributed generation (DG) is becoming an important option for economical and technical reasons. Implementation of intentional islanding of DG improves the continuity of supply and reliability of power system. In this paper, comprehensive learning particle swarm optimization (CLPSO) is used to optimally partition the distribution system in case of main upstream loss. The objective is to find the optimal islanding scheme of distribution system to achieve minimum active power generation cost, minimum reactive generation cost and minimum cost of the un-served power while satisfying system operational constraints. The solution proceeds by splitting the system into islands after the loss of main upstream feeder. In each island, the power balanced is achieved through load shedding. An optimal dispatch of the generating units of each island is then carried out to achieve minimum active power generation cost. A power flow calculations are carried out to calculate the reactive power generated by each unit and to check the operational constraints of the system. Finally, the effects of controlled islanding with and without utility owned DGs on the system reliability indices is studied. The proposed algorithm is applied to two radial and meshed test systems. (c) 2012 Elsevier Ltd. All rights reserved.