Multi-agent Approach For Service Restoration Of Distribution System Containing Distributed Generations

With the development of the social economy and the improvement of the living level, people are more dependent on the electric power. And the requirements of the service stability and energy quality are advancing. The distribution system connects to the user directly. How to diagnose, isolate and restore the faults becomes an important and emergent task of the power industry. The main role of the distribution automation is to diagnose the area of faults exactly and timely, then isolate it quickly, at last restore the service of the field without faults. It is just called service restoration, which has multi-goals, multi-periods and multi-restrictions. The result is a series of combinations of the switches. And we'll get the optimization service restoration approach, which satisfies the restrictions of radiant structure, the limitation of the feeder and achieves the objects including restoring the most load with the least number of switches.The distributed generation will be more rivalrous and important in the power system, owing to the extending of system scope and request to the power, the diversity of needs. This document sets up a layered control frame composing of distribution system control center, microgrid and distributed generation. This kind of layered syetem is fit for implementing the multi-agent controlling, which is independent, alternating and corresponding.The main works of this document: 1) Build models of the aim functions and restrictions of the distribution system service restoration considering the distributed generations; 2) The multi-agent system is used to solve the distribution service restoration with microgrid; 3) On the base of the multi-agent restoration system, analyse different situations of the distribution restoration in an example. Through comparing the numbers of operating switches, restored loads and the overplus capability feeders with the results of without distributed generation. We can conclude that the connecting of the distributed genenration improves the system stability and decreases the amount of the de-energized loads. The microgrid can be isolating or connectable with the network, which strengthens the controlling of distributed generations. It improves the flexibility and stability of the system greatly.