| Large-scale blackout is the most serious disaster to power system operation. With a lot of examples and lessons, Large-scale blackout has been under domestic and international attention to the power system. Therefore, to the large area power outages, it is necessary for effective prevention and restoration.For fault diagnosis, this paper analyzes and compares the power system fault diagnosis of the various methods. According to the different characteristics of the fault diagnosis in the power system, we divide the fault diagnosis into the systemic-level fault diagnosis and device-level fault diagnosis. Systemic-level fault diagnosis is as an online support tools for dispatchers handling the blackout. Systemic fault diagnosis's objective is to determine fault in nature, provide online help for the dispatcher. In the following paper, we shall judge the power outage's nature by a multi-factor fuzzy qualitative reasoning fault discriminant method. This judgment is not deliberately seeking complete and detailed description to the power outages. To the judgment of the fault device, we shall divide the process into two steps. In the first step, the district of fault is decided. The second step is deciding the fault device and the switch of refusal action by the forward and backward reasoning. By processing step, the diagnosis rate is greatly accelerated.The paper study and compare the various methods of restoration control in the power system. we put forward a path search technology based on multi-agent in this thesis. The restoration process is divided into two steps. The first step is serial electrical delivery of power. The second step is simultaneous delivery of power. The main objective of the first step is to establish the main power supply network. The second step is to quicken the electricity supply. For every stage of the power restoration, there are different optimal goals. The appropriate algorithm is implemented in each stage.At last, we introduce the design and the frame of the software. |