Improved improved harmony search algorithm (IBEGA) is used to solve distribution network reconfiguration with distributed generation problem. The algorithm introduces the basic harmony search algorithm in a random location update operation, reverse learning technology, small probability of mutation mechanism to fine-tune pitch correction probability strategies to enhance the algorithm"s mining and exploration capacity and improve the ability of global optimization, avoiding algorithm premature convergence problem. The distributed power as a PQ node processing, computing network trends before use backward substitution method. using IHS algorithm IEEE33 node system for network reconfiguration, and comparing with harmony search algorithms,the results show that the proposed IHS algorithm has better performance, can effectively solve distribution network reconfiguration with distributed generation problem.