Power System Fault Diagnosis Based On Temporal Bayesian Knowledge-Base And Its Software Implementation

With the more comoplex structure of power grid and demand of higher level of power system supervisory control, when the faults occur, a large number of fault informations are sent to the dispatching center. This situation has brought operators difficulties in dealing with these informations efficiently in a short time. The power system fault diganosis method afterwards based on Temporal Bayesian Knowledge Bases is studied aiming at assisting operators in detecting the fault component and identifying the fault situation.Firstly, the theory of Temporal Bayesian Knowledge Bases is introduced. This mathematical method is applied successfully in other areas and has practical applicability and good theoretical basis. The TBKB has been introduced for power system fault diagnosis in this paper.The power system fault diganosis method based on TBKB is proposed. The fault diganosis models based on components are firstly bulit. The models can clearly express the operation logics between the protections and breakers. Temporal Casual Relationship is built to express the temporal constraint relationships among components and protection operations, protection operations and related breakers tripping. The automatic generation method of TBKB fault diagnosis models based on the current structure of power grid is discussed.When the action and temporal informations of protections and breakers are received, the examination method of TCR is used to identify the mal-function, abnormal informations and time scale error. If there is missing information, the state assumption method is adopted to create hypothetic state combinations. For these states, Bayesian backward reasoning is made to calculate the fault probability and detect the fault component, then the forward reasoning is applied to identify mal-function and rejection. Examples have illustrated the correctness and effectiveness of this fault diagnosis method.The development of the power system fault diganosis software based on TBKB is introduced. The software is capable of generating TBKB fault diganostic models based on the current structure of power grid automaticly, implementating TBKB power system fault diganosis method correctly. The graphical presentation based on5layed cause-effect graph is used to display diagnosis results on User Interface. The unusual operation state such as mal-function, rejection and time scale error can be identified through different background colours. The operation results of the software are correct. The developing processes of faults and diagnosis results are clearly displayed.