Research of immune algorithms for reconfiguration of distribution network with distributed generations

The structure of traditional distribution network is changed from a single supply radial network to a more power ring network when grid is connected with distributed generations (DG), and island phenomenon occurs higher in the process of reconstruction, which becomes uncontrolled and high risk operation and increased the difficulty of distribution network reconstruction. The improved immune genetic algorithm introduced the principle of biological immune mechanism into standard genetic algorithm is put forward, the problem of genetic algorithm premature convergence in distribution network calculation is overcome, constraint conditions of preventing islanding phenomenon in the reconstructed model is added, encoding method based on the loop is used. The calculation results show that, in the case of a distributed power supply, the loss of network can be decreased and voltage quality is improved by using this method, what is the most that island phenomenon is avoided. IEEE33 node is an example to prove the feasibility of the method.