Power system static and dynamic state estimation methods using heterogeneous measurements

Abstract

Power system state estimation (SE) constitutes an essential function of energy management systems, enabling operators to maintain a comprehensive awareness of system operating conditions through available field measurements. This dissertation introduces several contributions to the research domain of hybrid state estimation (HSE), aimed at optimally integrating heterogeneous supervisory control and data acquisition (SCADA) and phasor measurement unit (PMU) data. Initially, fundamental concepts of static and dynamic SE are elaborated from both mathematical and practical implementation perspectives, followed by an introduction to the principles of HSE. Subsequently, key challenges associated with HSE implementations are identified, accompanied by a com-prehensive literature review focusing on novel static and dynamic HSE methods designed to overcome these challenges. Furthermore, a classification of existing methods is proposed based on their scope and underlying mathematical formulations. The contributions of this thesis first focus on static HSE methods. A weighted least squares (WLS)-based static HSE formulation is developed, separately handling SCADA and WAMS measurements. The principal advantages of the proposed method include its modular design and practical applicability, making it particularly suitable for PMU integration into existing SE software through minimal modifications. Moreover, considering the widespread adoption of high-voltage direct current (HVDC) transmission technology, a model suitable for current source converter (CSC)-HVDC links in static HSE implementations is proposed and validated via numerical simulations involving both SCADA and PMU measurements on the AC side, along with diverse combinations of DC-side measurements. Additionally, the thesis investigates the inclusion of current injection phasors from PMUs in static HSE algorithms, examining how various current measurement configurations – whether flows or injections – influence HSE performance, a topic inadequately addressed in prior literature. Recognizing the increasing complexity and stochastic behavior of contemporary power systems, transitioning toward advanced SE algorithms capable of providing enhanced system visibility and situational awareness becomes imperative. In response, this thesis proposes a hybrid forecasting-aided state estimation (FASE) approach leveraging an extended Kalman filter (EKF) framework. The method supplements existing static state estimators by incorporating additional information derived from the temporal evolution of system states through multi-sensor data fusion, employing a transition model that combines dense, real-time PMU measurements with forecasted state estimates. To address synchronization discrepancies between SCADA and PMU data, a post-processing correction step based on the modified Bryson-Frazier fixed-interval smoothing algorithm is implemented. In the final two chapters, algorithms dedicated to detecting and suppressing bad data within the context of the proposed HSE approaches are formulated. Additionally, practical aspects of the research are demonstrated using a laboratory-scale experimental setup that integrates both commercial and low-cost PMUs with a digital real-time power system simulator, thereby enabling comprehensive testing and validation of synchrophasor-based monitoring and control algorithms.