In this paper, a two-stage cooperative-noncooperative game theory is used to study the problem of the optimizing power dispatch and negotiating tariff for the cooperative operation of the microgrid system which is composed of photovoltaic power generation and power consumers. Firstly, in the first stage, the overall revenue of the alliance is improved by optimizing the power dispatch, then a mathematical model is built for the operating characteristics of the microgrid, ADMM is adopted to avoid the situation that the electric quantities cancels each other due to the effect of superposition in the trading process due to the superposition effect, and the dispatch results are obtained by using cooperative game. Secondly, in the second stage, the impact of trading tariff on the trading model is analyzed, and a bargaining model is built between photovoltaic power generation and power consumers , and the BAS algorithm is improved, a bargaining model combined with adaptive moment estimation is adopted and to the optimal trading tariff is obtained by using non-cooperative game to allocate the revenue. Finally, the feasibility of the model is verified by arithmetic examples, and the results show that the level of PV consumption can be improved by using the built model through market guidance while improving the revenue of microgrid.