Machine Learning Based Transient Stability Emulation and Dynamic System Equivalencing of Large-Scale AC-DC Grids for Faster-Than-Real-Time Digital Twin

Abstract

Modern power systems have been expanding significantly including the integration of high voltage direct current (HVDC) systems, bringing a tremendous computational challenge to transient stability simulation for dynamic security assessment (DSA). In this work, a practical method for energy control center with the machine learning (ML) based synchronous generator model (SGM) and dynamic equivalent model (DEM) is proposed to reduce the computational burden of the traditional transient stability (TS) simulation. The proposed ML-based models are deployed on the field programmable gate arrays (FPGAs) for faster-thanreal-time (FTRT) digital twin hardware emulation of the real power system. The Gated Recurrent Unit (GRU) algorithm is adopted to train the SGM and DEM, where the training and testing datasets are obtained from the off-line simulation tool DSATools TM /TSAT R . A test system containing 15 ACTIVSg 500-bus systems interconnected by a 15-terminal DC grid is established for validating the accuracy of the proposed FTRT digital twin emulation platform. Due to the complexity of emulating large-scale AC-DC grid, multiple FPGA boards are applied, and a proper interface strategy is also proposed for data synchronization. As a result, the efficacy of the hardware emulation is demonstrated by two case studies, where an FTRT ratio of more than 684 is achieved by applying the GRU-SGM, while it reaches over 208 times for hybrid computational-ML based digital twin of AC-DC grid.INDEX TERMS AC-DC grid, digital twin, dynamic equivalents, faster-than-real-time, field programmable gate arrays, gated recurrent unit, machine learning, parallel processing, power system stability, real-time systems, recurrent neural networks, synchronous generator.

Review of Data-Driven Techniques for On-Line Static and Dynamic Security Assessment of Modern Power Systems

“…Although a recent review of HVDC modeling for power system stability assessment [144] made no mention of ML model applications, one example has recently been presented in [145]. Furthermore, the use of data-driven modeling techniques and machine learning methods for the development of dynamic models of microgrids is advocated in [146].…”

CNN-Based Digital Twin Model for Ultra-High Voltage Direct Current System Loss Measurement

“…Machine learning is commonly used for fault warning and diagnosis in HVDC systems [13] [14], which can automatically identify early onset faults based on measured data. Shiqi Cao et al established a dynamic equivalent model of an AC-DC power grid and constructed a digital twin based on machine learning hybrid computing [15]. However, when machine learning is used for calculating the energy consumption of individual components, it is necessary to calculate the losses of different components separately and then add them up, making it difficult to consider the cross coupling relationship of multidimensional heterogeneous data, which affects computational efficiency and accuracy.…”

An Intelligent Location Method for Power System Oscillation Sources Based on a Digital Twin

“…Digital twin technology has been used in the simulation, analysis and control of power systems. Reference [8] develops an approach that treats the OPF problem as a functional mapping between the system operating status and OPF solutions. Reference [9] proposes a confidence-oriented model updating strategy, which only requires small sample data to update the model.…”