Research On Dynamic Continuous Diagnosis Mechanism Of Power System Faults

With the continuous development of power grids, interconnection between different regional power systems are more and more close, which makes the system fault more influential to the power systems. Once complex fault occurs, lots of alarm and fault information floods into the dispatch centre in a short time, which is beyond the dispatchers’ handling ability and usually leads to incorrect diagnosis. So it is necessary to rely on a fault diagnosis system for fast and correct fault identification. It can help dispatchers make control decisions to ensure the safety of the power system.However, the traditional fault diagnosis methods have to wait for completion of gathering all the fault alarm information to carry out the diagnosis process. As a result,the dynamic process and the time continuity are ignored, which is actually a static and isolated thinking way of fault diagnosis. In this thesis, a new dynamic continuous method of fault diagnosis for power systems is proposed, in which a “Self-Adaptive Dynamic Event Window” is used to directly monitor the dynamic process of fault occurrence and development as well as capture the monitoring information for detecting the fault in time.The proposed method combines the forward reasoning with the backward reasoning,tracks the fault developing process in real-time till the fault ends, and takes the newly increased alarm information into consideration at the same time, assuring fast and accurate diagnosis. Firstly, according to different actions of breakers and the scope of faults, the line faults are divided into three patterns, that is, Self-sufficient Mode, Breaker Failure Mode and Remote Mode. A Self-Adaptive Fault Mode is built based on the three patterns and WAMS measurement information. Then the Self-Adaptive Dynamic Event Window is used to monitor the newly increased alarm information and the power network real-time information, and track fault developing process. Finally, the mechanism of dynamic continuous fault diagnosis is established, based on the Self-Adaptive Fault Mode and Self-Adaptive Dynamic Event Window. The feasibility and effectiveness of the proposed method are verified by the diagnosis results of the fault examples in the Yunnan power grid.