Abstract: In the background of the new power system, a large number of small-scale microgrids will inevitably be integrated into the distribution network at multiple points, and the cooperative optimization operation of the distribution network and multiple microgrids faces difficulties such as high dimensionality of decision variables, complex optimization objectives of multiple subjects,and strong randomness of single small subject access interaction. This paper proposes a self-organized cooperative regulation strategy for the microgrid-distribution network（MG-DN）based on stochastic evolutionary dynamics. It aims to improve the spontaneous evolutionary optimization ability of the multiple benefit subjects and the anti-interference ability of single small entities with random access. First, a multi-dimensional resource regulation model based on the public goods price response for the multiMG-DN system is established to reduce the dimensionality of variables for multi-subject optimization. Then, a self-organized evolution strategy based on the stochastic evolutionary stability criterion is proposed. The multi-optimization objective is transformed into a two-dimensional objective of “cooperation and non-cooperation” through an evolutionary utility function under the binomial distribution of multiple benefit subjects. And the spontaneous coordinated operation capability is obtained by the evolutionary dynamics process taking into account the stochastic fluctuation forces to effectively cope with the infulence of stochastic characteristics of the single small microgrid. Finally, the modified IEEE 37-bus system is simulated. The effectiveness of the proposed strategy is verified through multi-algorithm comparison experiments in multiple typical scenarios.