Search

This paper presents a deep learning-based approach for hourly power outage probability prediction within census tracts encompassing a utility company's service territory. Two distinct deep learning models, conditional Multi-Layer Perceptron (MLP) and unconditional MLP, were developed to forecast power outage probabilities, leveraging a rich array of input features gathered from publicly available sources including weather data, weather station locations, power infrastructure maps, socio-economic and demographic statistics, and power outage records. Given a one-hour-ahead weather forecast, the models predict the power outage probability for each census tract, taking into account both the weather prediction and the location's characteristics. The deep learning models employed different loss functions to optimize prediction performance. Our experimental results underscore the significance of socio-economic factors in enhancing the accuracy of power outage predictions at the census tract level., Comment: Accepted in 2024 IEEE PES General Meeting, Seattle, Washington (PES GM 2024)

Industrial sensor data provides significant insights into the failure risks of microgrid generation assets. In traditional applications, these sensor-driven risks are used to generate alerts that initiate maintenance actions without considering their impact on operational aspects. The focus of this paper is to propose a framework that i) builds a seamless integration between sensor data and operational & maintenance drivers, and ii) demonstrates the value of this integration for improving multiple aspects of microgrid operations. The proposed framework offers an integrated stochastic optimization model that jointly optimizes operations and maintenance in a multi-microgrid setting. Maintenance decisions identify optimal crew routing, opportunistic maintenance, and repair schedules as a function of dynamically evolving sensor-driven predictions on asset life. Operational decisions identify commitment and generation from a fleet of distributed energy resources, storage, load management, as well as power transactions with the main grid and neighboring microgrids. Operational uncertainty from renewable generation, demand, and market prices are explicitly modeled through scenarios in the optimization model. We use the structure of the model to develop a decomposition-based solution algorithm to ensure computational scalability. The proposed model provides significant improvements in reliability and enhances a range of operational outcomes, including costs, renewables, generation availability, and resilience., Comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

This paper proposes an incentive-based aggregator for electric vehicles (EVs) in workplace parking lots to participate in the energy and regulation markets. With the implementation of seasonal Autoregressive Integrated Moving Average (ARIMA) model to predict the market information, a Day-Ahead (DA) planning model coordinated with the EV's responses to the incentive is formulated to determine the days to activate the aggregation program and the optimal incentives that are broadcasted to the EV owners in the first stage. Given the determined incentives and EV's responses, an real-time (RT) optimal bidding algorithm complying with EVs' energy demand in the second stage is designed to maximize the aggregator's profits in the markets. The proposed models are tested using the data collected from PJM's energy and regulation markets. The results show the incentive-based aggregator can benefit both the EV owners and the aggregator from the credits obtained from the markets. Meanwhile, the results also indicate the proposed optimal bidding algorithm is capable of handling the uncertainty of regulation signals and following the signals with high precision., Comment: Submitted

The power system needs more transmission capacity to deal with the increasing integration of renewable energy resources. The battery energy storage systems (BESSs) are effective to enhance the grid capacity and relieve the transmission congestion. A comprehensive revenue analysis of BESSs is critical for market participants to install such systems in a market-based power system. Taking PJM as an example, this paper carries out a thorough revenue analysis for the entire system. Highly profitable nodes in the system are revealed and characterized for BESS participants. A comparison study of stationary and transportable BESSs shows the transportable energy storage can produce higher potential revenue in the energy and regulation markets. Based on the results of the revenue analysis and characterization of commercial pricing nodes, an optimal placement algorithm is proposed for finding the profitable sites for market participants to install BESSs in the system and the algorithm is validated with real PJM market data., Comment: Submitted

A competitive wholesale electricity market consists of thousands of interacting market participants. Driven by the variations of fuel costs, system loads and weathers, these market participants compete actively and behave variously in the power market. Although electricity markets tend to become more transparent, a large amount of market information is still not publicly available to market participants. Hence, data-driven analysis based on public data is crucial for market participants to better understand and model large-scale power markets, and ultimately to perform better in power trading. While most of the previous researches related to the large-scale power markets are based on the synthetic networks, a data-driven approach utilizing the real power market data is proposed in this paper. First, the power plants' monthly net generation and capacity data are obtained from U.S. Energy Information Administration (EIA) and aggregated to figure out the monthly regional capacity factors which are used to characterize the market's regional behaviors for market participants. Then, the regional capacity factors are analyzed against the metered system loads and natural gas prices to study the generation behaviors in the power market. The analysis reveals the impacts of regional natural gas prices on capacity factors and the responses of generating behaviors to the system loads. The analysis results present the solid evidence and rational references for market participants to model and validate the large-scale power market in the future., Comment: 51st North American Power Symposium, October 2019

Accurate identification of parameters of load models is essential in power system computations, including simulation, prediction, and stability and reliability analysis. Conventional point estimation based composite load modeling approaches suffer from disturbances and noises and provide limited information of the system dynamics. In this work, a statistic (Bayesian Estimation) based distribution estimation approach is proposed for both static (ZIP) and dynamic (Induction Motor) load modeling. When dealing with multiple parameters, Gibbs sampling method is employed. In each iteration, the proposal samples each parameter while keeps others fixed. The proposed method provides a distribution estimation of load models coefficients and is robust to measurement errors.

High penetration of photovoltaic (PV) generators can lead to voltage issues in distribution networks. Various approaches including the real power control through PV inverters have been proposed to address voltage issues. However, among different control strategies, communication delays are inevitably involved and they need to be carefully considered in the control loop. Those delays can significantly deteriorate the system performance with undesired voltage quality, and may also cause system instability. In this paper, according to the inverter based active power control strategy, a linearized state space model with communication delay is presented. A delay dependent stability criterion using linear matrix inequality (LMI) approach is used to rigorously obtain the delay margins based on different system parameters. The method can handle multiple PVs in the distribution network as well., Comment: 5 pages, 7 figures

Short-term electricity price forecasting has become important for demand side management and power generation scheduling. Especially as the electricity market becomes more competitive, a more accurate price prediction than the day-ahead locational marginal price (DALMP) published by the independent system operator (ISO) will benefit participants in the market by increasing profit or improving load demand scheduling. Hence, the main idea of this paper is to use autoregressive integrated moving average (ARIMA) models to obtain a better LMP prediction than the DALMP by utilizing the published DALMP, historical real-time LMP (RTLMP) and other useful information. First, a set of seasonal ARIMA (SARIMA) models utilizing the DALMP and historical RTLMP are developed and compared with autoregressive moving average (ARMA) models that use the differences between DALMP and RTLMP on their forecasting capability. A generalized autoregressive conditional heteroskedasticity (GARCH) model is implemented to further improve the forecasting by accounting for the price volatility. The models are trained and evaluated using real market data in the Midcontinent Independent System Operator (MISO) region. The evaluation results indicate that the ARMAX-GARCH model, where an exogenous time series indicates weekend days, improves the short-term electricity price prediction accuracy and outperforms the other proposed ARIMA models, Comment: IEEE PES 2017 General Meeting, Chicago, IL

Availability of a validated, realistic fuel cost model is a prerequisite to the development and validation of new optimization methods and control tools. This paper uses an autoregressive integrated moving average (ARIMA) model with historical fuel cost data in development of a three-step-ahead fuel cost distribution prediction. First, the data features of Form EIA-923 are explored and the natural gas fuel costs of Texas generating facilities are used to develop and validate the forecasting algorithm for the Texas example. Furthermore, the spot price associated with the natural gas hub in Texas is utilized to enhance the fuel cost prediction. The forecasted data is fit to a normal distribution and the Kullback-Leibler divergence is employed to evaluate the difference between the real fuel cost distributions and the estimated distributions. The comparative evaluation suggests the proposed forecasting algorithm is effective in general and is worth pursuing further., Comment: Accepted by IEEE PES 2018 General Meeting