Abstract: According to the relationship between faults and topology of the distribution network as well as the coupling between repair and recovery, a multi-fault repair and optimization model of distribution network based on fault adjacency state is established in this paper. In the stage of emergency repair, the electrified state matrix of load nodes is established based on the power supply type. The electrified state matrix of the fault adjacent load is proposed by extracting the electrified state of the fault adjacent load. The fault adjacency state is obtained by extending the electrified state matrix of the fault adjacent load, then the faults are classified. The optimal repair task at each stage is determined by the alternate update of repair and fault adjacent state. During the reconstruction calculation period, the node voltage based adaptive ordered ring matrix for the ring network is set up as the solution space of the algorithm. Then, the Levy coefficient quantum particle swarm optimization was applied using the decreasing cosine function and Levy flight to improve the quantum particle swarm optimization algorithm. The practicability and the effectiveness of the proposed methods are verified by simulation on PG&E69 bus system.