Power system security enhancement by NARX ANN model based static synchronous series compensator

Abstract

Ensuring the security of power systems has become an increasingly complex challenge due to the growing intricacies and dynamic nature of modern grids. This paper introduces a novel approach to strengthening power system security by integrating a Nonlinear Autoregressive with Exogenous Inputs (NARX) Artificial Neural Network (ANN) model with a Static Synchronous Series Compensator (SSSC). The NARX ANN model is applied to forecast key system parameters, facilitating real-time and adaptive control of the SSSC. A comparative performance between a conventional Proportional-Integral (PI) controller and the NARX-based SSSC shows that the NN controller outperforms the PI controller in respect of dynamic performance. The NARX controller is designed to quickly mitigate power oscillations and improve power flow control during disturbances, surpassing the conventional PI controller's capabilities. The proposed controller uses two input signals—reference voltage and measured voltage at the SSSC location. The training data is generated from the difference between these two signals. The model’s results were validated using MATLAB simulation.