Ensemble Forecasting Applied to Power Systems

Signal-devices management and data-driven evidential constraints based robust dispatch strategy of virtual power plant

Nanogrids are becoming an essential part of modern home power systems, offering sustainable solutions for residential areas. These medium-to-low voltage, small-scale grids, operating at medium-to-low voltage, enable the integration of distributed energy resources such as wind turbines, solar photovoltaics, and battery energy storage systems. However, ensuring power quality, stability, and effective energy management remains a challenge due to the variability of renewable energy sources and evolving customer demands, including the increasing charging load of electric vehicles. This paper reviews the current research on nanogrid architecture, functionality in low-voltage distribution systems, energy management, and control systems. It also explores power-sharing strategies among nanogrids within a microgrid framework, focusing on their potential for supplying off-grid areas. Additionally, the application of blockchain technology in providing secure and decentralized energy trading transactions is explored. Potential challenges in future developments of nanogrids are also discussed.

Electricity load forecasting is of high importance for electricity management. Modern power systems are complex and diverse, resulting in increased randomness and nonlinear factors of electricity load data, which greatly increases the difficulty of forecasting. This paper proposes a hybrid-deep-learning-based load forecasting method, named DCFformer (DFT-CNN-FEDformer), for short-term load forecasting (STLF) tasks. The method first employs the discrete Fourier transform (DFT) to denoise time-sequence data on electricity load, so that fluctuations caused by incidents can be reduced. Secondly, it utilizes a convolutional neural network (CNN) that produces sequences of local features extracted from the denoised time sequences. Thirdly, a FEDformer network is applied to perform load forecasting by using extracted feature sequences. In the experiments, we utilize datasets from three regional power systems or apparatuses to compare the proposed DCFformer with other approaches, and the results show that, under the same conditions, DCFformer outperforms the competitors in forecasting precision, which proves the significance of its performance and practicality.

Guest Editorial: Situational awareness of integrated energy systems

Cybersecurity is crucial for modern power systems due to their high digitization. The open information and communication technology, that is (ICT) being used for the operation of such systems, is highly vulnerable to cyber threats. The adopted smart grid concept around the globe enables the utilization of demand-side for providing ancillary services based on well-known demand response (DR) programs. These programs aggregate smart appliances in homes and electric vehicles (EVs) for providing vital services such as frequency regulation and voltage support. Since the aggregation is based on cyber layer, any cyber threat could affect the ancillary services that are being delivered from the aggregators, which might lead to stability and security issues resulting in brownout or massive blackouts. This chapter discusses the cybersecurity in DR program and shows its importance for modern and future smart power systems due to their stability and security margins. Furthermore, the cyberattack case study is implemented in a power system with demand side program responsible for providing primary frequency support ancillary service, where the results confirm the high vulnerability of modern power systems to cyber threats on DR-active power reserve providers. Moreover, technical suggestions are provided for enhancing the cybersecurity in DR programs in power systems with high power share from renewable energy sources.

Unit scheduling in hybrid microgrids (PV/wind–generator–battery) is nonlinear, multi-constraint, and affected by uncertainties in load and renewable energy forecasts. Conventional rule-based or deterministic optimization methods often require accurate models and are less robust to forecast errors, while large-dimensional exact solutions are not always feasible for real-time operations. This study proposes a Deep Reinforcement Learning (DRL)-based generator scheduling optimization framework that formulates the problem as a Markov Decision Process. The state vector includes multi-horizontal load/renewable energy forecasts, battery state of charge, fuel price, and unit operating limits; actions are the genset power setpoint and battery charge/access rate. A reward function internalizes fuel costs, battery degradation, emissions, curtailment, and unsupplied energy penalties, while also encouraging reserve provision. To ensure operational safety, we add a safety layer that projects policy actions onto the feasible set (SOC limits, ramp rate, minimum on/off, and converter capacity). Training is performed offline with domain randomization over weather and load profiles, and then evaluated in a rolling horizon scheme with minute resolution. Simulation results demonstrate operating cost savings and curtailment reduction compared to the MILP/MPC baseline, with high constraint compliance and sub-second inference times, making it suitable for implementation in edge controllers. This approach demonstrates scalability across a wide range of microgrid configurations and remains robust to uncertainties, offering a practical path to low-cost and low-emission operation.

The electric power system is under rapid development under the context of smart grid. As electric demand evolves, the power system must evolve too. New loads in the system are changing the ways system operators ensure system security. One such load recently attracting attention are plug-in hybrid electric vehicles (PHEVs). Seen as the stepping stone to an all electric transportation sector, PHEVs are expected to significantly penetrate the distribution system in the foreseeable future. PHEVs offer unique challenges to system operators. Not only are they additional loads on the system, but PHEVs with vehicle to grid functionality can also behave as distributed generators. PHEVs can potentially offer ancillary services such as contingency reserves for the system operator. This paper will investigate the potential of PHEVs in providing contingency reserves. Analysis will be completed using mixed integer programming solved via the commercially available software AIMMS.

Flexible AC Transmission System (FACTS) controllers are advanced power electronic devices employed in electrical power systems to enhance system stability, reliability, and power transfer capabilities. These controllers enable dynamic and flexible control of various transmission parameters such as voltage, current, impedance, and phase angle, thereby improving the overall power system performance. FACTS devices are broadly classified into series, shunt, series- series, and series-shunt configurations, each designed to manage reactive power and regulate voltage either by injecting voltage in series or current in shunt with the transmission line. Well-known FACTS devices include Static Var Compensators (SVC), Static Synchronous Compensators (STATCOM), Thyristor Controlled Series Capacitors (TCSC), Static Synchronous Series Compensators (SSSC), and Unified Power Flow Controllers (UPFC), each serving specific roles in power flow control, voltage regulation, and transient stability enhancement. The integration of these controllers provides significant benefits such as increased transmission line capacity, improved voltage stability, minimized transmission losses, and damping of power system oscillations. They are capable of rapid response during transient disturbances, effectively mitigating voltage collapses and enhancing system resilience against faults and contingencies. FACTS controllers not only improve steady-state operational parameters but also address dynamic stability and transient conditions, enabling power systems to operate closer to their stability limits while maintaining security. Their deployment helps avoid costly infrastructural expansions by optimizing existing transmission assets, which is critical given environmental and economic constraints. Recent advances in semiconductor technologies and control algorithms have made FACTS devices more reliable, efficient, and adaptable to modern grids, including renewable integration and deregulated power markets. In summary, FACTS controllers play a vital role in modern power systems by providing flexible, efficient, and reliable solutions for power flow management, voltage control, and system stability enhancement, thereby supporting the evolving demands of today's complex electrical networks.

Nowadays, the increment of energy demand in the world as well as the development of smart grids and the combination of different types of energy systems have led to the complexity of power systems. On the other hand, ever-expanding energy consumption, development of industry and technology systems, and high penetration of renewable energies have made electricity networks operating in more complex and uncertain conditions. Also, consumers and especially users of sensitive load tend to have access to a reliable and sustainable power supply. Therefore, power producers need a variety of long- and short-term planning methodologies for attaining to sustainable investment, production, and operation. Analysis of traditional power and energy systems requires physical modeling and extensive numerical computation. To analyze the behavior of these systems, advanced metering, and condition monitoring devices and systems are utilized, which generate a huge amount of data. Evaluation of these data is approximately impossible by conventional or numerical methods, and it requires powerful data mining procedures. Regression, classification, and clustering applications of machine learning and deep learning methods are powerful tools to use for dealing with such issues. These procedures can be utilized for load/demand forecasting, renewable energy generation forecasting, demand response evaluation, and power system analysis. Understanding the problem and functioning of each learning methods is, therefore, one of the most important issues in the application of such approaches to solve power system problems. Accordingly, in this chapter, the authors will introduce and discuss selected applications of machine learning and deep learning based on their learning, structure, mode of operation, and application in the load forecasting of power systems. Literature review on machine learning and deep learning applications in load forecasting will be presented in this chapter.

A review on artificial intelligence based load demand forecasting techniques for smart grid and buildings

Load forecasting plays a crucial role in fields such as energy management and power system planning. Accurate forecasting can effectively reduce energy costs and enhance the stability and reliability of power systems. Traditional forecasting methods struggle to balance forecasting accuracy and efficiency when dealing with complex and variable load data. Although the Long Short-Term Memory network (LSTM) can address the long-term dependency problem in time series and shows certain advantages in load forecasting, the setting of its initial parameters has a significant impact on the forecasting results, and it is prone to getting trapped in local optima. The Particle Swarm Optimization algorithm (PSO), based on swarm intelligence, has powerful global search capabilities and fast convergence. This paper proposes a load forecasting method based on PSO-LSTM, which uses PSO to optimize the parameters of LSTM, thereby enhancing the generalization ability and forecasting accuracy of the model. Through case-based analysis and comparison, it can be seen that compared with traditional forecasting methods, this method effectively improves the accuracy of load forecasting and provides strong support for the efficient and stable operation of power systems.

With the deployment of advanced metering infrastructure (AMI), an avalanche of new energy-use information became available. Better understanding of the actual power consumption patterns of customers is critical for improving load forecasting and efficient deployment of smart grid technologies to enhance operation, energy management, and planning of electric power systems. Unlike traditional aggregated system-level load forecasting, the AMI data introduces a fresh perspective to the way load forecasting is performed, ranging from very short-term load forecasting to long-term load forecasting at the system level, regional level, feeder level, or even down to the consumer level. This paper addresses the efforts involved in improving the system level intraday load forecasting by applying clustering to identify groups of customers with similar load consumption patterns from smart meters prior to performing load forecasting.

Categorization of residential electricity consumption as a basis for the assessment of the impacts of demand response actions

Accurate load forecasting is essential for reliable and efficient operation of power systems. Traditional forecasting methods often struggle with capturing complex nonlinear patterns in load data. Artificial neural networks (ANNs) have emerged as a promising alternative due to their ability to learn complex relationships from historical data (Syed et al. in IEEEA 9:54992–55008, 2021. https://doi.org/10.1109/ACCESS.2021.3071654). This study investigates the potential of ANNs for short-term peak load forecasting in a 150 kV power system in Semarang, Indonesia. The study examines the impact of different input variables, including historical peak load, minimum load, population, and energy production, on forecasting accuracy. Several ANN architectures are trained and evaluated using mean absolute percentage error (MAPE) and mean squared error (MSE) metrics as reported by Demuth and De Jesús (neural network design). The results indicate that ANNs can achieve high accuracy in predicting peak load, with MAPE values below 10%. The study also demonstrates the importance of carefully selecting input variables and training parameters for optimal model performance. The findings highlight the potential of ANNs for improving load forecasting accuracy in power systems, contributing to enhanced grid reliability and operational efficiency. The findings of this study contribute to a deeper understanding of the application of ANNs in power system load forecasting. They demonstrate the potential of ANNs to achieve high accuracy and provide valuable insights into the factors influencing model performance. The findings are relevant for power system operators, researchers, and policymakers working to improve grid reliability and efficiency as reported by Prabha Kundur and Malik (Power System Stability and Control, McGraw-Hill Education, New York, 2022. https://www.accessengineeringlibrary.com/content/book/9781260473544).

The power system that we have today has gone through several transformations over the years. These changes are due to either advancement in science and technology or changes in customer demands. Today we have the vision for a future electrical power system known as the ‘Smart Grid’. This power system has several functional and operational capabilities which are envisioned to solve the reliability, generation deficit and cost issues in the power system. Majority of the power systems in the world are still being operated as the traditional power system with one-way communication and one-way power flow. Transitioning to a Smart Grid (SG) influences the protection schemes and principles of the power system as the smart grid is to incorporate Distributed Generation (DG) units and other Distributed Energy Resources (DERs) hence there is going to be a bi-directional flow of power and information. Therefore, there is also a need for development and advancement in the protection schemes. Over the years there have been a lot of transformational stages in the protection schemes for power system operation. Understanding the protection needs is important for power system operators to know the reason why certain protection schemes cannot work on new and modern power systems as they seek to transform their utilities to the future electrical grid known as the smart grid. This paper presents the transitional stages that power system protection has gone through and the requirements needed for smart grid protection.