Distribution Network Reconfiguration in the Presence of Harmonic Loads: Optimization Techniques and Analysis

This paper deals with reconfiguration of distribution system to minimize power losses, deviation of nodes voltages and branches current constraint violations. This means alternation of feeders´ topological structure by changing the status of tie and sectionalizing switches. The work presented here is a step forward to define the reconfiguration problem closer to reality by considering the effect of harmonic loads. To solve this complicated combinatorial, non-differentiable constrained optimization problem, novel heuristic optimization techniques such as Shuffled Frog Leaping Algorithm (SFLA) and Imperialist Competitive Algorithm (ICA) are employed. The methods are compared with Discrete Particle Swarm Optimization (DPSO) algorithm to examine their performance including computational time, convergence rate, and accuracy of results. In this context, experiments are conducted on 69-bus and 84-bus distribution test systems for the short- and long-term reconfiguration scenarios. The obtained results establish that harmonic losses are a major deciding factor in reconfiguration and harmonic loads cannot be represented with only fundamental components.