Artificial Intelligent Method and Singular Value Decomposition based System Identification for Digital Twin of Inverter based Power Grid

This paper proposes two methods to construct the digital twins (DT), based on the artificial neural network and singular value decomposition. DTs are designed to increase observability for state estimation of the inverter-dominant distribution grid. With the increasing complexity of the power grid, demand to use state estimation increases to maintain the grid stability of operation. Based on the metered data from the phasor measurement unit, an online model, the so-called DT, can be created to monitor the current grid status. It can be used for state prediction and event forecasting also. In this paper, the state estimation of the inverter dominated grid formulated by the DT model. To build the model, the artificial neural network and singular value decomposition-based system identification are utilized to emulate the dynamic feature of the inverter. The training data set of the neural network in the DT model is obtained from the reference model, which is created based on traditional dynamic state equations of inverter and power grid. To validate the accuracy of the DT model, the simulation results of the reference model are first compared with the results of the DT model after the training process. Afterward, two additional scenarios are simulated using DT based on the Cigré benchmark grid to validate the accuracy of state estimation in the power grid.