Abstract: Service restoration strategy of the distribution network containing distributed generation is very important for the security of distribution networks. A graph theory-based service restoration algorithm is introduced in this paper for the distribution networks containing distributed generation and tie switches. Firstly, a service restoration model for distribution network containing distributed generation as well as an objective functions under service restoration are developed after consideration of the total quantities of restored load, the switching times of circuit breakers and the load losses. Secondly, a revision and adjustment mechanism of infeasible solutions is designed based on graph theory, which includes four steps, i.e., network state genera-tion, network region division, network structure revision and load checking. Then, the Quantum Discrete Particle Swarm Optimization(QPSO) algorithm is applied to achieve the optimization of the objective function. Finally, a case study has proved the superiority of the graph theory-based algorithm in revision of the infeasible solution and the effectiveness of the proposed graph theory-based algorithm in service restoration of distribution networks.

Key words: power grid, graph theory, service restoration, QPSO, power supply reliability