This paper introduces a data-driven framework that harnesses residential flexibility to support distribution network operation and enhance end-user comfort. The proposed framework consists of three modules: (i) a forecasting module that predicts day-ahead flexible and non-flexible power profiles; (ii) a grid-aware coordination module that utilizes these forecasts to calculate residential operating envelopes, ensuring network security and preventing voltage violations; and (iii) a home energy management system that schedules the operation of flexible assets to optimize comfort, reduce energy costs and minimize CO2 emissions while respecting the operating envelopes. Numerical simulations conducted on the IEEE European low-voltage test feeder using real smart meter data demonstrate the effectiveness of the proposed framework. Additionally, its fully data-driven design offers a significant advantage in cases where the network configuration is unknown.

Despite the rapid expansion of smart grids and large volumes of data at the individual consumer level, there are still various cases where adequate data collection to train accurate load forecasting models is challenging or even impossible. This paper proposes adapting an established Model-Agnostic Meta-Learning algorithm for short-term load forecasting in the context of few-shot learning. Specifically, the proposed method can rapidly adapt and generalize within any unknown load time series of arbitrary length using only minimal training samples. In this context, the meta-learning model learns an optimal set of initial parameters for a base-level learner recurrent neural network. The proposed model is evaluated using a dataset of historical load consumption data from real-world consumers. Despite the examined load series’ short length, it produces accurate forecasts outperforming transfer learning and task-specific machine learning methods by 12.5%. To enhance robustness and fairness during model evaluation, a novel metric, mean average log percentage error, is proposed that alleviates the bias introduced by the commonly used MAPE metric. Finally, a series of studies to evaluate the model’s robustness under different hyperparameters and time series lengths is also conducted, demonstrating that the proposed approach consistently outperforms all other models.

Recent advancements in home energy management systems (HEMS) have focused on integrating end-user habits into day-ahead appliance scheduling. In this paper, a multi-objective HEMS assisted by a non-intrusive load monitoring (NILM) module is examined. The multi-objective HEMS aims to optimize total energy cost, reduce environmental footprint, and alleviate grid stress while ensuring end-user comfort and respecting individual habits. The scope of the paper is to evaluate the robustness and generalization capabilities of the NILM-assisted HEMS across varying conditions. The impact of each objective on the overall system performance is investigated with a particular focus on the role of personalized disaggregation in empowering homeowners to participate in energy-saving initiatives actively.

The proliferation of residential distributed energy resources, such as photovoltaic (PV) systems, into low-voltage distribution networks has introduced new challenges in phase grouping for distribution companies, which is critical for grid monitoring, balancing, and planning. This paper benchmarks seven algorithms for solving the phase grouping problem, taking into account the increasing penetration of PV systems and analyzing the impact of different levels of PV integration on phase connectivity. Additionally, the analysis is conducted on a monthly basis to monitor temporal variations and assess the impact of seasonal fluctuations on algorithm performance. Among the examined methods, the algorithm based on voltage correlation consistently outperforms clustering approaches, achieving 100% accuracy across all examined scenarios.

This paper introduces a holistic, multi-objective home energy management system (HEMS) designed to optimize residential electrical and thermal demands. By incorporating dynamic electricity tariffs and integrating solar production with storage units, the system aims to minimize energy costs while ensuring thermal comfort through energy-efficient solutions. Environmental sustainability and the impact on the distribution grid are also considered within the formulated optimization problem. A non-intrusive load monitoring tool is employed to schedule the operation of flexible appliances tailored to user habits, achieving a mean absolute error below 14W and an $\rm {F_{1}}$ score exceeding 0.8 for the most energy-intensive appliances across three public datasets. Analysis of five operating scenarios highlights the impact of each objective on system performance. The proposed HEMS outperforms baseline solutions, substantially reducing operational and environmental costs up to 42% and 27%, respectively. Moreover, it diminishes user discomfort by up to 86% and alleviates stress on the grid by 33%.

The increasing penetration of distributed renewable energy sources and the adoption of new power-intensive appliances, such as electric vehicles and heat pumps, poses unprecedented technical challenges to the power grid, especially on the distribution level. Furthermore, with the widespread roll-out of advanced metering infrastructure (AMI), new data-driven services can be leveraged to improve distribution networks’ performance, robustness, and flexibility. Accurate peak demand forecasting is a good example of a service that can play a vital role in smart grid operations. It can unlock demand response potential and allow more cost-efficient asset management and better planning for various stakeholders, i.e., market participants or generation units. This work presents a comparative analysis of 11 state-of-the-art machine learning (ML) approaches regarding day-ahead peak demand forecasting, along with the data analysis and feature engineering processes.

Given the additional awareness of the increasing energy demand and gas emissions’ effects, the decarbonization of the transportation sector is of great significance. In particular, the adoption of electric vehicles (EVs) seems a promising option, under the condition that public charging infrastructure is available. However, devising a pricing and scheduling strategy for public EV charging stations is a non-trivial albeit important task. The reason is that a sub-optimal decision could lead to high waiting times or extreme changes to the power load profile. In addition, in the context of the problem of optimal pricing and scheduling for EV charging stations, the interests of different stakeholders ought to be taken into account (such as those of the station owner and the EV owners). This work proposes a deep reinforcement learning-based (DRL) agent that can optimize pricing and charging control in a public EV charging station under a real-time varying electricity price. The primary goal is to maximize the station’s profits while simultaneously ensuring that the customers’ charging demands are also satisfied. Moreover, the DRL approach is data-driven; it can operate under uncertainties without requiring explicit models of the environment. Variants of scheduling and DRL training algorithms from the literature are also proposed to ensure that both the conflicting objectives are achieved. Experimental results validate the effectiveness of the proposed approach.

Non-intrusive load monitoring (NILM) is a topic that lately attracts both the academic and the industrial interest. NILM is used to reveal useful information regarding the consumption breakdown on appliance or activity level, thus can be a key solution to unlock various smart-home services and opportunities. To that end, deep learning has arisen as a prominent solution. Although most of the known solutions so far focus on a predefined number of home appliances, this paper proposes a multi-class NILM system which can detect in real-time any number of appliances and can be efficiently embedded into simple microprocessors. The key feature for the identification of the appliances is the processing of measured turn-on active power transient responses sampled at 100 Hz. The NILM system includes three stages; adaptive thresholding event detection method, convolutional neural network and k-nearest neighbors classifier. For future extensions, it is capable to automatically identify new appliances; thus, no retraining and additional modeling is required. (c) 2021 Elsevier B.V. All rights reserved.

Non-intrusive load monitoring (NILM) is a topic that lately attracts both the academic and the industrial interest. NILM is used to reveal useful information regarding the consumption breakdown on appliance or activity level, thus can be a key solution to unlock various smart-home services and opportunities. To that end, deep learning has arisen as a prominent solution. Although most of the known NILM techniques so far focus on a predefined number of home appliances, this paper proposes a system which can detect in real-time any number of appliances. In the proposed solution switched-on appliances are identified by processing the measured active power transient response sampled at 100 Hz. The NILM system includes three stages; statistics-based event detection, convolutional neural network and k-nearest neighbors classifier. For future extensions, it is capable to automatically identify new appliances; thus, no retraining and additional modeling is required.

Smart-meter technology advancements have resulted in the generation of massive volumes of information introducing new opportunities for energy services and data-driven business models. One such service is non-intrusive load monitoring (NILM). NILM is a process to break down the electricity consumption on an appliance level by analyzing the total aggregated data measurements monitored from a single point. Most prominent existing solutions use deep learning techniques resulting in models with millions of parameters and a high computational burden. Some of these solutions use the turn-on transient response of the target appliance to calculate its energy consumption, while others require the total operation cycle. In the latter case, disaggregation is performed either with delay (in the order of minutes) or only for past events. In this paper, a real-time NILM system is proposed. The scope of the proposed NILM algorithm is to detect the turning-on of a target appliance by processing the measured active power transient response and estimate its consumption in real-time. The proposed system consists of three main blocks, i.e., an event detection algorithm, a convolutional neural network classifier and a power estimation algorithm. Experimental results reveal that the proposed system can achieve promising results in real-time, presenting high computational and memory efficiency.

In this paper, a bottom-up modelling approach of households equipped with appliances, photovoltaic and battery energy storage system (BESS) is proposed. In particular, to simulate the daily household power profiles, a high-resolution dataset of appliances is created, including ground truth data as well as laboratory measurements. The household model applies both to single- and three-phase installations and incorporates a technique to limit the irrational BESS operation under frequent power fluctuations. Indicative simulation examples demonstrate the performance of the developed model in terms of power consumption and evaluates the effectiveness of some of the model parameters.

In this paper, a novel version of the Optimal Distributed Generation Placement (ODGP) problem regarding the siting and sizing of Renewable Energy Sources (RESs) units is presented, called Optimal RES placement (ORESP). Power losses constitute the objective function to be minimized, subject to operational constraints. The simultaneous installation of a mix of RESs is considered and the Capacity Factor (CF) ratio is used as an aid for taking into account: (a) the geographical characteristics of the area, in which the examined Distribution Network (DN) is placed, (b) the different weather conditions, and (c) the availability of RESs, all of that at the same time, while keeping the problem complexity at minimum. The contribution of this work is that the proposed methodology bypasses the weather uncertainties and, thus, the RESs' power generation stochasticity and provides an adequate solution with minimum computational burden and time, since the proposed CF use allows solving the problem under a straightforward way. Unified Particle Swarm Optimization (uPSO) is used for solving ODGP and ORESP. Moreover, a sensitivity analysis regarding the CFs variations is performed and finally a comparison of the proposed method with a more realistic one is performed, to consolidate further the claims of this paper. The proposed method is evaluated on RES-region-modified 33- and 118 bus systems.

This paper provides an extended review of projects and implementations for high-temperature superconducting (HTS) transmission systems worldwide. Projects are categorized based on a multi-criteria analysis that considers transmission technology, i.e., ac or dc, laboratory, or commercial activity, and classified according to operating characteristics, e.g., length, voltage/current level, power capacity, and cooling means. A survey about applications that can be supported by HTS links is also presented, whereas technical challenges that impede the more massive grid penetration of HTS systems are highlighted. Finally, the current technology readiness for both ac and dc implementations is assessed and a discussion about future HTS-links adoption is provided.

Modeling of battery energy storage systems used for applications, such as electric vehicles and smart grids, emerged as a necessity over the last decade and depends heavily on the accurate estimation of battery states and parameters. Depending on the battery-cell type and operation, a combination of algorithms is used to identify battery parameters and define battery states. This paper deals with robust Li-ion batteries modeling with a specific focus on a hybrid approach for a more accurate state-of-charge (SOC) estimation. The analysis presents a detailed description of the state-of-the-art stand-alone SOC estimation methods and focuses on a hybrid SOC estimation technique to improve accuracy under varying conditions. Emphasis is given on performance improvements of the proposed hybrid approach compared to the conventional methods, whereas a thorough experimental validation is presented to evaluate the accuracy of the proposed method.

Information is the pathway to smart electricity distribution grids. Current trends point at phase measurement units (PMUs) becoming the gateway to getting distribution grid information to increase its observability, controllability and monitoring and deliver more advanced grid services and other functionalities. Grid-connected power electronics converters (GCPECs) are ubiquitous in distribution grids as the interface between the grid and distributed generation sources and modern electronic loads. The paper proposes the integration of PMU capability into GCPECs. The cyber-physical systems framework of smart grids is used to illustrate the concepts and benefits of PMU enabled GCPECs. The coordinated initiatives of solar and energy storage GCPECs becoming more grid supportive, providing more information to the grid operators and allowing their direct control by the utilities are summarized as the basis for further development of GCPECs. Finally, the need for communication infrastructure and the potential of using optical sensors for the PMU enabled GCPECs are discussed.

This paper deals with the appropriate modeling and control strategy of a multiterminal dc transmission (MTDC) system that incorporates high-temperature superconducting (HTS) dc cables. The system is based on voltage source converters for the interconnection of stiff ac grids. An overview of the high voltage direct current power transfer technology along with the possibilities it enables, as well as an introduction to the concept of multiterminal topology are presented. The operation principles of the superconducting technology are described and the control strategy of HTS-based MTDC networks is presented. To validate the performance and the dynamic response of the system under steady state as well as under fault and load change conditions, a typical four-terminal MTDC network in ring topology is developed in MATLAB/Simulink.

In this paper, two discretization techniques are combined to present a coupled electro-thermal model for the transient analysis of gas-insulated transmission lines (GILs). The analysis of the thermal problem is based on the volume-element-method, representing heat transfer equations on a 2-D axisymmetric GIL model and determining temperature distribution over space and time. For the electrical formulation, the finite-difference time-domain method is adopted considering constant parameters calculated at the frequency of interest. The coupled electro-thermal problem in the case of transient conditions is solved bymeans of a bidirectional procedure. A set of test cases is imposed on the electric part, and transient analysis is conducted while considering the line coupled electro-thermal performance.

•Overview of BESS integration within an onboard DC system.•Battery model parameter identification using RLS-based method with variable forgetting factors.•Kalman filter and TLS-based techniques for the SOC and capacity estimation.•The combined solution indicates promising results, suitable for demanding ship operation.•Experimental validation is provided to further prove the authors claim.

Reliability of battery energy storage systems (BESS) used for online applications, such as electric vehicles and smart grid, depends heavily on the accuracy and rapidness of the state of charge (SOC) estimation. Moreover, to achieve a robust SOC estimation, the battery model parameter identification process is of significant importance. This paper examines a combination of the adaptive unscented Kalman filter (AUKF) and the fast upper diagonal recursive least square (FUDRLS) for the parameter identification and SOC estimation processes, respectively. The analysis focuses on on-line applications and the results are compared with previous work. Experimental validation based on various setups and load conditions is conducted, whereas the advantages of the proposed combination are highlighted.