Distribution Network Reconfiguration Based On Immune Discrete Particle Swarm Optimization

Distribution system is an important part in power system which connects power production to consumer. Distribution network reconfiguration is an important aspect of distribution system automation and optimization. Along with the continual development of the national economy and the unceasing enhancement of the people's lives, the people set the more and more request to the reliability, security and efficiency of the distribution system. Distribution network reconfiguration is an optimal method taking the improvement of the distribution network movement, which has been the focus of the study, because it doesn't need to increase the extra investment to improve the distribution network movement quality.Distribution network reconfiguration is a complicated combinatorial optimization problem, nonlinear and multi-objective. The problem is one of the N on-Polynomial (NP) hard problems in power system. The destination of the distribution network reconfiguration is to reduce the system power loss, to relieve the overloads in the network and to improve the system reliability by changing the state of some sectionalizing switches.Firstly, it analyses the significance, some common objective functions and the research actuality of the distribution network reconfiguration in this paper.Some kinds of methods on power flow calculation in distribution networks was discussed in this dissertation, especially highlighting the technique of backward/ forward iterative algorithm. Then it uses depth first search (DFS) to analysis the network structure and the calculating sequence of the nodes. The results indicate that the algorithm has a good astringency.The paper analyses the particle swarm optimal optimization (PSO), the discrete particle swarm optimization (DPSO) and the immune algorithm (IA), proposes the immune discrete particle swarm optimization algorithm (IDPSO). Then according to the structural feature of distribution network with close-loop designing and open-loop operating mode, the discrete particle swarm optimization is improved to ensure the radial structure of distribution network and evidently reduce the number of infeasible solutions, connecting with immune algorithm to avoid the premature convergence problem. The algorithm enhances the computation speed and the ability searching in the entire solution space.In this paper, reconfiguration to reduce power loss of distribution network is studied. The program of distribution network reconfiguration based on immune discrete particle swarm optimization is developed and the result of the three IEEE examples is presented, which shows validity and correctness.