Abstract: When multi-agent virtual power plant takes part in the electric energy market and peak shaving market, the virtual power plant operator presents the characteristics of bilateral interaction with the external market and the internal resources. The operator's bidding decision needs to consider the coordination between the external market and the internal members. This paper proposes an internal and external coordinated bidding strategy for the virtual power plant operator to participate in the electric energy market and the peak shaving market and coordinate with each member internally. The coordination and management of the operator and the internal members are realized by compensating for the operation of energy storage, fully consuming wind and photovoltaic power, and conducting dynamic games between price and electricity or between the operator and the controllable distribution generation and flexible load through the Stackelberg game. A multi-agent double-layer bidding model of the operator and the internal members is established and solved by the combination of the particle swarm optimization algorithm and the interior point method. Finally, an example is given to prove the rationality and effectiveness of the proposed strategy.