Wenyuan Tang

The issue of transparency has been a persistent concern from the stakeholders in electricity markets. While the markets throughout the U.S. have been redesigned after the California electricity crisis, there remain a lot of ambiguities arising from the sophisticated market rules, resulting in numerous long-lasting litigations. Moreover, emerging market mechanisms including demand response may bring in undesirable gaming opportunities, which can be further exacerbated by the lack of transparency. It is envisioned that improving market transparency is the key to limiting the room for manipulation. This five-year project aims to build analytical frameworks to understand, evaluate, and adapt the evolving market mechanisms, toward the long-term goal of creating market mechanisms and providing policy recommendations that enhance the transparency, reduce the manipulation, and ultimately improve the efficiency of electricity markets. In light of the fact that the underlying engineering details are typically highly abstracted in policymaking, this project will recognize and incorporate the intricate operational constraints in power systems into the economic modeling, elevating the role of engineers in market design. This project will investigate virtual transactions and demand response, as integral parts in today's market design, and accomplish the following objectives: (1) laying the analytical foundation for wholesale electricity market design, through the formulation of a two-settlement, networked market model which characterizes the spatiotemporal interdependency of market operations; (2) scrutinizing the design of virtual transactions, and evaluating their impact on both price convergence and economic efficiency; (3) quantifying the manipulation of baseline methods in demand response programs, and creating novel mechanisms that balance the design trade-offs; and (4) integrating research with education to promote energy literacy, through innovative outreach activities which foster inclusive opportunities in energy education.

Microgrids (MG) deliver highly resilient power supply to local loads in the event of a power outage, while improving distribution system reliability by reducing the load on the system under stress conditions. Networked microgrids that coordinate with each other and the distribution system operator increase reliability and resilience further by improving the diversity of the generation assets and loads. Microgrids are also widely considered as an essential technology to address the challenges of climate change.

The EVOPT Planner and Controller both aim to provide users with an accurate understanding of the power availability for charging throughout a given day, whether from the grid or from on-site generation sources. The Controller utilizes a number of signals to determine where and when fleet energy is going to be provided from the electricity grid or on-site generation such as solar and storage. For the Controller to minimize operations cost, it is critically important to have an accurate energy availability forecast for the upcoming day. As such, an algorithmic approach using available data sources and sans any hardware is needed to forecast the availability of solar energy at a given site, and at multiple points in which a fleet may install solar capacity for charging vehicles.

In the current state-of-the-art machine learning based real-time control and decision-making in large-scale complex networks such as electric power systems is largely bottlenecked by the curse of dimensionality. Even the simplest linear quadratic regulator design demands cubic numerical complexity. The problem becomes even more complex when the network model is unknown, due to which an additional learning time needs to be accommodated. In this 3-year NSF CPS proposal, we take a new stance for solving this problem, and propose a hierarchical or nested machine learning-based scheme for real-time control of extreme-dimensional networks. Our approach will be to design appropriate projection matrices by which a network can be divided into disparate sets of non-overlapping groups depending on the low-rank properties of their controllability grammian, and multiple sets of composite controllers can be learned independently for each group using model-free reinforcement learning. Accordingly, the control goals of the network will also be decomposed into local (microscopic) and global (macroscopic) reward functions. Local controllers will be designed via privacy preserving group learning, and the global controllers via model reduction and averaging. Sparsity-promoting structures will be imposed on top of the local controllers to reduce their communication complexity. Deep learning algorithms based on historical events will be used to train recurrent neural networks so that they can rapidly predicting these sparse projections after any disturbance event in the network. Throughout this entire exercise, wide-area control of power systems using streaming Synchrophasor data from Phasor Measurements Units (PMUs) will be treated as a driving example. Results will be validated using standard IEEE models, a simplified model of the Japanese power grid with high-scale solar penetration, and an Opal-RT model of the Duke Energy power grid integrated with the ExoGENI cloud computing network at the FREEDM Systems Center.

Through multidisciplinary doctoral education in Cybersecurity for Electric Power Systems (CEPSE), North Carolina State University (NCSU) will increase its commitment to graduate training in two areas designated by the GAANN Program as critical to national need: Cybersecurity and Electrical Engineering. The goal of is to enlarge the pool of U.S. citizens and permanent residents who will pursue teaching and research careers in cybersecurity for electric power systems, thereby promoting workforce development and technological innovation impacting, national security, energy security, and environmental sustainability.

With the increasing penetration of behind-the-meter solar and energy storage, it is favored to leverage recent advances in artificial intelligence to enhance the accuracy of net-load forecasting, the observability of net-load variability, and the understanding of the coupling between net-load and demand response potentials. The proposed project will develop two models to address the hybrid probabilistic forecasting when small and large data sets are available. The first model will incorporate a new gradient boosting machine, in which a projection of the distribution into a Riemannian space is considered, whose corresponding natural gradient is expected to give better updates at each iteration than the state of the art. Meanwhile, a data-driven type-2 fuzzy system which generates monotone if-then rules will be developed to preprocess inputs. The second model consists of graph attention networks, transformers, and variational autoencoders. The graph attention networks overcome the theoretical issues with spectral based methods. The transformers ensure each time step to attend over all the time steps in the input sequence, compared with recurrent neural networks. The combination can give better spatiotemporal information. Moreover, those two models will be extended to forecast net-load with the consideration of demand response potentials, as a multi-target forecasting task.

In this project, we will develop a Photovoltaic Analysis and Response Support (PARS) platform for improving solar situation awareness and providing resiliency services. The team will focus on developing new operation modes for solar energy systems and a PV+DER situation awareness tool to enable accurate estimation and predication of PV and DER operation conditions in both normal operation conditions and in emergency operation when there is a wide spread outage caused by natural disasters or coordinated cyber attacks. Real-time dynamic studies will be conducted to compute system operation conditions for different operation options. This tool will be run on real-time simulation platform so that optimal restoration plans can be developed in real-time using operation modes enabled by Tasks 1 and parameters derived in Task 2. The team will model transmission, distribution, and all the way down to each DER and inverter units at utility scale PV farms on a multi-core OPAL-RT real-time simulation platform.

This project will focus on investigating the impact of distributed generation (DG) on a utility distribution system from the cost-of-service perspective and developing a methodology to quantify those costs. The NCSU team will closely collaborate with the related groups in Duke Energy to determine which impacts can be reasonably quantified, present the developed methodology to the Public Staff and other stakeholders, and potentially file expert testimony in a general rate case. This project is primarily an engineering study, providing inputs to the rates department in Duke Energy, which will handle the rate design (i.e., how those costs should be allocated to DG).