Abstract: With the access of large-scale distributed energy resources in active distribution network，the

traditional centralized control architecture system will not effectively deal with a large number of distributed unit，so it is necessary to study the optimal scheduling of distributed resource cluster in active distribution network. Taking the active distribution network based on multi-agent architecture as the research object，a double-layer distributed optimal scheduling strategy based on the power deviation of "source-load" cluster was proposed according to the hierarchical regulation and control idea of "intra group autonomy and coordination among groups". In the upper global optimization stage，an improved consistency algorithm based on power deviation of "source-load" cluster was proposed to optimize the distributed coordination among clusters. In the lower level local optimization stage，the distributed units were optimized in coordination based on distributed consistency algorithm，and the "source-load"demand were obtained under the optimal benefit. At the same time，the convergence of the proposed strategy was proved. The IEEE 33 system simulation results show the effectiveness of the proposed strategy.

      Key words: active distribution network（ADN）；distributed resource cluster ；consistency algorithm；cluster

power match ；distributed dispatch