Abstract: The mismatch between source and load characteristics as well as the randomness and intermittency of distributed new energy block the on-site accommodation of distributed new energy in the distribution network. However, accommodation obstruction would cause line losses and a sharp increase in operation cost. To solve this issue, the on-site power balance and time-varying characteristics of line losses in distribution networks with high penetration distributed new energy is analyzed firstly. Next, a two-layer optimization model for on-site accommodation of distributed new energy with participation of energy storage in distribution networks is established with the maximum on-site accommodation of distributed new energy as well as the minimum line losses and operation cost as objectives. Finally, case studies are conducted on IEEE-39 system and IEEE-118 system using particle swarm optimization(PSO) algorithm and multi-verse algorithm(MVO). The results show that the two-layer optimization model considering line losses proposed in this paper has significant advantages compared to traditional two-layer optimization models and single-layer optimization models, i.e., cost, line losses and new energy curtailment are reduced by 6.08%, 8.02%, 13.12% and 24.07%, 18.13%, 4.72%, respectively. In addition, comparing with traditional PSO algorithm, cost and new energy curtailment are decreased by 15.83% and 49.10% using the MVO algorithm. Then the feasibility, effectiveness and superiority of the presented two-layer optimization method of the distribution network as well as the MVO algorithm are verified.