Study Of Self-healing Techniques For Smart Distribution Networks With DGs

Distribution network is directly connected with the users in the whole system, whose failures will affect people’s daily lives and economic activities, so fault locating, isolating and self-healing reconfiguration are important tasks in power system. As more and more distributed generations connected to system, they will have some impacts on the distribution network because of their own characteristics. So attentions should be taken to the self-healing process of distribution networks with DGs when faults occur. This paper has done some work to achieve self-healing algorithm and program.First of all, this paper analyzes some related technical problems and the option of access solutions when distributed generations are connected to system. The capacities of DGs are determined according to the voltage level. The distribution network will become instability because of the particularities of distributed generations. The proposed concept of micro-grid is a good solution to this problem. Micro-grids are independent operation units which can generate, transmit and distribute power in small grid networks, and keep their reliability. They can connect and disconnect to external power grid with no blackout.Based on the basic concept, achievement conditions and basic requirements of self-healing reconfiguration in the distribution network, a common used framework "2-3-6" is analyzed, which divides the self-healing control process into the follow parts: two rings control logic, three control structures, and six control links. Three most popular methods of self-healing are compared, and the application of Agent technology is emphasized in this paper.According to the above analysis, simulated annealing algorithm is proposed to in this paper realize self-healing reconfiguration of the distribution network after fault occurs. The effectiveness of this algorithm is verified by using two examples: simple distribution network and complex distribution network with multiple power sources. In addition, the effect is analyzed when DGs are connected to the system.The best network structure can be got through simulated annealing algorithm, but there are still some defects, so two improvement of the algorithm are proposed in this paper. First, combining heuristic algorithm with intelligent algorithm, in which way, these arithmetic can learn from each other’s strong points to offset respective weaknesses. Second, adding the concept of particle swarm algorithm and introducing the region segmentation optimization ideas. The self-healing process is well improved by these two methods.