Search

We introduce a model-free scheme that is capable of identifying topology changes in distribution networks using data from distribution-level phasor measurement units (D-PMUs). To do so, we use algebraic tools of behavioral system theory to develop easy-to-implement algorithms. The intrinsic difficulties associated with D-PMU measurements and their proposed solution are discussed as an additional challenge for the implementation of the proposed algorithms. Experimental results using real D-PMUs and laboratory real-time hardware are provided as a proof of concept using a standard IEEE-test node feeder.

We present k-ShapeStream, a clustering method for streaming time-series data. In addition to the algorithmic novelty, the method represents a highly practical approach for electric grid data analytics, requiring no model assumptions or ground truth information, running sustainably on ever growing datasets, and providing intuitive and insightful results to grid operators. We demonstrate the effectiveness of k-ShapeStream using several months of real synchrophasor data from an operational distribution network in California. Through two case studies on (i) transformer tap changes; and (ii) voltage sags, we illustrate how k-ShapeStream assists in identifying and analyzing recurring grid events, a critical task for decision making in electric grids.

This chapter focuses on technological innovations in electric power infrastructure and their enabling potential for the integration of variable generation resources, electric vehicles (EVs), and microgrids. Many of the innovations described in this chapter were driven by evolving technical and economic constraints that long predated renewable grids: factors other than clean energy have dominated problem solving and investment in new electric power hardware, software, and design strategies over the past several decades. In the context of the renewable energy transition, however, some of these rather esoteric niche technologies take on a new relevance and interest for a broader community of researchers and practitioners.

The push to automate and digitize the electric grid has led to widespread installation of Phasor Measurement Units (PMUs) for improved real-time wide-area system monitoring and control. Nevertheless, transforming large volumes of high-resolution PMU measurements into actionable insights remains challenging. A central challenge is creating flexible and scalable online anomaly detection in PMU data streams. PMU data can hold multiple types of anomalies arising in the physical system or the cyber system (measurements and communication networks). Increasing the grid situational awareness for noisy measurement data and Bad Data (BD) anomalies has become more and more significant. Number of machine learning, data analytics and physics based algorithms have been developed for anomaly detection, but need to be validated with realistic synchophasor data. Access to field data is very challenging due to confidentiality and security reasons. This paper presents a method for generating realistic synchrophasor data for the given synthetic network as well as event and bad data detection and classification algorithms. The developed algorithms include Bayesian and change-point techniques to identify anomalies, a statistical approach for event localization and multi-step clustering approach for event classification. Developed algorithms have been validated with satisfactory results for multiple examples of power system events including faults and load/generator/capacitor variations/switching for an IEEE test system. Set of synchrophasor data will be available publicly for other researchers.

In this work, we present an ongoing, three year project funded by the Advanced Research Project Agency (ARPA-E) to develop an infrastructure to accelerate the development and deployment of artificial intelligence solutions for the electric grid. The project addresses the critical issues that we have identified as hampering the deployment of AI on electric grid measurements. These issues have stymied the potential for insight from high resolution grid measurements and generally hindered artificial intelligence innovation in the utility industry. The project consists of three main components. The first is the deployment of a diverse set of grid sensors to capture a variety of grid behaviour, both from the field and in simulation. The second is the deployment of a highly performant, scalable, cloud-based data management and AI platform designed for time series data to enable the easy storage, processing and analysis of grid sensor data. The third is the cultivation of an open research community of experts around the platform and data through useful educational material, code and data sharing, and data science competitions. Overall, the project will accelerate the development of analytics, machine learning, and AI by addressing existing gaps in data, tools, and people, with the aim of improving the electric grid.

This paper introduces two tools-the Effective Impedance State Estimator (EISE), and the Real Time Effective Impedance Estimator (RTEIE). The EISE is novel in that it is an entirely data-driven state estimator, and does not require a network model. Thus, the EISE is well suited for applications in which an accurate network model is not available. Instead of using a network impedance model, the EISE uses an effective impedance network model that is learned online by the RTEIE. The RTEIE recursively estimates the effective impedance network model using phasor measurement unit (PMU) measurements. The mathematical reasoning for using an effective impedance model, rather than a full network impedance model, is given. The recursive estimation framework allows the RTEIE to take the entire history of phasor measurements into account to produce a model estimate, while still adjusting to changes in the network impedance due to external factors such as topology changes and temperature. The merits of the RTEIE and EISE are demonstrated by comparing the accuracy of the EISE with a conventional state estimator on a toy network.

This publication presents an unsupervised network impedance estimation method for AC microgrid electrical networks using only voltage and current synchrophasor measurements. The proposed method combines the use of two techniques: the Kron reduction and a novel common mode elimination technique to improve estimation robustness to measurement noise. We motivate Kron reduction as an instrument for addressing the rank-deficiency of electrical network voltage measurement data. The Kron reduced model can be used in optimal power flow and other centralized and decentralized control algorithms. The proposed common mode elimination technique improves the robustness of the impedance estimate to sensor noise by improving the eigenstructure of the voltage measurement data matrix, without discarding any useful information. We present results demonstrating improved admittance matrix estimation accuracy on realistic data over a range of signal to noise (SNR) ratios. The proposed method is data-driven and will update the network impedance estimate based on sensor measurements alone, without any external input, making it is well suited for time-varying microgrid applications.