Wide Area Control and Improvement of Dynamical State of Power System

Abstract

Some approaches to achieving full observability and controllability in power systems are explored in this thesis. The core structure of this work consists of observability and controllability, focusing on two key phases of optimization and control. In the first phase, system observability is addressed through optimal placement of Phasor Measurement Units (PMUs). The Connectivity Matrix Algorithm (CMA) is used to determine the optimal PMU locations, initially without consideration of Zero Injection Buses (ZIBs), and then considering them to ensure full system observability. The transmission system simulation results for the IEEE 9, 14, 24, 30, 57, and 118 bus systems has been shown. MATLAB R2024a is utilized to validate the methodology, providing a robust framework for achieving optimal PMU configuration across different system scales and configurations. In the second phase, firstly the IEEE-9 bus test system has been used without a controller, testing three different scenarios (tie-line, branch data, and solar plant integration) to assess system stability. Thereafter, a Deep Deterministic Policy Gradient (DDPG) algorithm-based controller is introduced, which is grounded in Reinforcement Learning (RL). In Wide-Area Measurement Systems (WAMS), the controller uses frequency information to receive its global input signal from PMU devices. This method is used to improve controllability and damp out low frequency oscillations in power systems after PMU integration. The development of a Wide Area Control (WAC) strategy for stable power system operations is a key focus of this phase. Through these objectives, this thesis aims to contribute to the field of power system stability and control by optimizing system observability and implementing effective control strategies to improve dynamic performance. The proposed approaches provide a foundation for future research and practical applications in wide area control and power system dynamics.