| The distribution network reconfiguration is an important means for optimization operation and loss reduction. The distribution network reconfiguration is a complicated problem of multi- objective and nonlinear combinational optimization in theory. Then the intelligence technology developed rapidly. Meanwhile, it is impossible to use intelligence algorithm to achieve the distribution network reconfi- guration for loss minimization. The Particle Swarm Optimization(PSO)is an evolution calculation method based on swarm intelligence. PSO is an efficient parallel optimizat- ional method and it can be used to solve large number of nonlinear, non-differentiable and multi-peak value complicated optimization. PSO was success- fully used in system discriminate and neural network training, and this prove that it would be successful in distribution network reconfiguration also. If PSO could be improved according to the characteristic of the distribution network, it would be applied more comprehensively.A method based on PSO with mutation operator is proposed for distribution network reconfiguration in this paper. Firstly, we analysis the characteristic of the distribution network and propose binary PSO which is more suit the operation of the distribution network. Then the mathematical model, in which the objective is loss minimization, is set up. The convergence property of PSO is seriously influenced by the neighborhood and accelerative coefficient. So the adaptive theory is used to adjust the parameter and enhance efficiency of optimization. The connatural defect of PSO is the problem of local optimization. So the theory of mutation in Genetic Arithmetic (GA) is added to help get rid of the bondage of local optimization.The result of two testing systems of distribution network justify that the method proposed in this paper can solve the distribution network reconfiguration efficiently and get a correct result. It is of good convergence property.
|
PDF Full Text Request