In the power distributed economic dispatch based on multi-agent consensus algorithm, the information interaction between generation units relies on the communication network, so the topological structure of the communication network significantly impacts the performance of the dispatch system. Aiming at the dispatch system under the event-triggered consensus algorithm, a communication network topology optimization method based on greedy algorithm is proposed to further improve the convergence rate of the algorithm while reducing the increased communication burden of the system. The method introduced the characteristic ratio as the performance evaluation index to take into account the convergence rate of the consensus algorithm and the communication frequency of the system. By successively adding edges to the communication network topology to ensure that the characteristic ratio increment is maximized at each step, so as to realize the optimization of the communication topology and improve the overall performance of the dispatch system. The simulation results show that the proposed method improves the convergence rate of the consensus algorithm by 19.1% compared with the topology optimization method that only considers system communication frequency. Compared with the topology optimization method that only considers convergence rate of the consensus algorithm, the system communication frequency is saved by 5.6%. This indicates that the proposed method is more balanced in taking into account the convergence rate of consensus algorithm and system communication frequency. Furthermore, the simulation experiments also verify the dispatch system exhibits strong robustness after topology optimization.