A radial basis function neural network based optimal placement of distributed generation sources in distribution networks

Abstract EN

Distributed Generation (DG) has been envisaged to play an escalating role in electrical power system in near future.

However, the benefits of DGs cannot be explored if they are not optimally placed.

The efficiency of the distribution system is increased with the employment ofDG units.

As the employment of the DG comes with the idea of loss minimization and voltage enhancement of the distribution feeders.

This paper presents a Radial Basis Function Neural Network (RBFNN) based strategy to penetrate the effective locations and sizing of the DG in a Radial Distribution Network.

Training and learning of the Neural Network (NN) has an ample importance while dealing with the real problems, however, the computation time associated with the process put a question mark on the effectiveness of the approach.

RBFNN with conventional learning algorithms does not achieve the desired speed and performance in the training process.

To overcome this difficulty a heuristic technique Particle Swarm Optimization (PSO) is employed for efficient learning of the RBFNN.

The convergence of RBFNN-PSO is compared with Conventional RBFNN.

Efficacy of the proposed approach is tested over the standard IEEE-33 bus system.