A Study Of Reconfiguration In Distribution Network Based On Improved Adaptive Genetic Algorithms

Network reconfiguration is an important means of the distribution system operation and control, is an important component part of the distribution management system. In the current circumstances of economic has been developing rapidly and electricity tense to increasingly, through distribution network reconfiguration, giving full play the potential of the existing distribution network and enhancing the security and the economy of the system has great economic and social benefits.This paper simplifies the distribution network model, seen the switching equipment in feeder line as a node, seen the node power as the load flowed this node. The distribution feeders and distribution transformers between two adjacent nodes are seen as a comprehensive slip road; Anew approach to modeling distribution system is presented. Based on special hiberarchy characteristic, this paper proved the back/forward sweep distribution load flow algorithm. It can compute parallely the loss of voltage and power of branches in the same layer, which can save computation time.Elaborates the genetic algorithm basic principle, the characteristic, mathematics foundation and the application, in view of basic genetic algorithm limitation, selection, selection question of crossover rate and mutation rate, improves the sequence of genetic operation, uses choice method which the proportional fitness assignment method and the best individual preservation law unified; In the process of optimizing, the use of improved adaptive adjustments in the cross-rate and the rate of mutation, reduces the possibility that the search to a local optimum, improves the efficiency.Combining the characteristics of distribution network with the adaptive genetic algorithms, the paper discusses specifically how to achieve reconfiguration in distribution network based on improved adaptive genetic algorithms, proposes the code method that taking the ring circuit switch number as the gene, the system ring circuit leg number as the chromosome length, reduced the chromosome length greatly. Takes the effective means to eliminate the isolated island and the loop in the genetic manipulation, thus greatly reducing the production of the feasible solutions in genetic manipulation, improve the algorithm search ability and speed. Unify the adaptive control and the genetic algorithm, carries on the adaptive adjustment to crossover rate and the mutation rate, sped up the algorithm search speed, avoid a premature convergence effectively.The paper uses MATLAB as a simulation tool to carry on the example simulation for optimized reconfiguration in distribution network, analysis and results of examples show that the method is effective.