Capsule Network with Long Short-Term Memory for Fault Detection and Diagnosis in Electrical Power Grids

The identification of fault types and their locations is crucial for power system when a fault occurs in the lines. Subsequently various type of faults and producing maximum capacity of electrical power grids is processing in the static mode. However, the electrical power transmission is access in static mode, dynamic power generated is failure to analysis the traditional mechanism. This paper proposed Capsule Network with Long Short-Term Memory (LSTM) for detection various type of faults as parameters of the model to expand the scope of the fault identification. Data collected from the simulated three phase of the grid power that change in the voltages of current, which is enhance capsule network combined the LSTM in electrical power transmission system. The combination of capsule network and LSTM technique enhance the detection of different types of faults. The capsule network captures the spatial hierarchies and analysis relationship between various feature, while LSTM handle the temporal dependencies in grid power. The static and dynamic behaviours handle the capsule network and LSTM dynamically generation pattern of power grid data. The performance of proposed method is evaluated with various metrics such as accuracy, precision, recall and f1-score metrics on grids power.