Fault Modeling and Traceability of Power Grid Dispatching Control System based on Data-driven method

The power grid dispatching control system is an important part of the power system operation, including a large number of components. Once a system component fails, it will cause a series of component alarms, which will endanger the security of the power system. Aiming at this problem, a fault tracing method based on the combination of system fault time series modeling and knowledge graph is proposed. Firstly, the hardware components of the power grid dispatching control system are modeled with prior knowledge, and then the fault time series of the modeled system components are simulated and transformed into the fault network represented by the knowledge graph using the Monte Carlo-based Choesky factor linear transformation. A network fitting similarity algorithm calculates the similarity of the fitted network, and finally uses the real network and the fault network fitting to trace the source of the fault. Numerical experiments were carried out on simulated data. The test results demonstrate the effectiveness and correctness of the method.