The invention relates to a method and system for reconfiguring a distributed generation-containing power distribution network. The method comprises the following steps of obtaining system parameters and a network loss objective function of a to-be-reconfigured power distribution network and generating an initial frog population; and updating frogs by using a cloud generator combination method during local depth searching. The distributed generation-containing power distribution network is reconfigured by using an improved leap-frog algorithm, a cloud model theory with stable tendency and randomness is introduced into the local depth searching process of the leap-frog algorithm, that is to say, a traditional leap-frog algorithm is replaced with a cloud generator for updating the worst frog-related information in various sub-populations by using random numbers, so that the convergence rate of the leap-frog algorithm is greatly improved, the global optimal solution can be quickly obtained, and the local optimum condition is avoided.