Aiming at the infeasibility of energy trading due to network constraints in virtual power plants (VPP), there are many ways to optimize the internal power balance. However, the fact that VPPs can break geographical restrictions and achieve large-scale energy interconnection and sharing allows them to better integrate geographically dispersed distributed resources, but it also makes traditional methods unsuitable for the optimization of energy within VPPs. This paper proposes a peer-to-peer (P2P) transaction optimization method based on the alternating direction multiplier method. The overall goal of this method is to minimize the total cost of transactions within the VPP. First, the network constraints of the line are represented by the load vector matrix in the built-in form, and the network constraints are written into the objective function through the augmented Lagrange multiplier method. Secondly, the alternating direction multiplier method is used to split the original optimization problem into several easy-to-solve sub-problems for iterative solution, which effectively improves the accuracy of the data and the execution efficiency of the algorithm. Finally, the feasibility of the algorithm is verified through case study, and the error value converged to less than 0.1%.