Sustainable Energy Grids & Networks最新文献

Residential and commercial buildings, equipped with systems such as heat pumps (HPs), hot water tanks, or stationary energy storage, have a large potential to offer their consumption flexibility as grid services. In this work, we leverage this flexibility to react to consumption requests related to maximizing self-consumption and reducing peak loads. We employ a data-driven virtual storage modeling approach for flexibility prediction in the form of flexibility envelopes for individual buildings. The risk-awareness of this prediction is inherited by the proposed scheduling algorithm. A Mixed-integer Linear Program (MILP) is formulated to schedule the activation of a pool of buildings in order to best respond to an external aggregated consumption request. This aggregated request is then dispatched to the active individual buildings, based on the previously determined schedule. The effectiveness of the approach is demonstrated by coordinating up to 500 simulated buildings using the Energym Python library and observing about 1.5 times peak power reduction in comparison with a baseline approach while maintaining comfort more robustly. We demonstrate the scalability of the approach by solving problems with 2000 buildings in about 21 s, with solving times being approximately linear in the number of considered assets.

With the growth of Renewable Energy (RE) generation, the operation of power grids has become increasingly complex. One solution could be automated grid operation, where Deep Reinforcement Learning (DRL) has repeatedly shown significant potential in Learning to Run a Power Network (L2RPN) challenges. However, most existing DRL algorithms have only considered individual actions at the substation level. In contrast, we propose a more holistic approach by proposing specific Target Topologies (TTs) as actions. These topologies are selected based on their robustness. In this paper, we present a search algorithm to find the TTs and upgrade our previously developed DRL agent CurriculumAgent (CAgent) to a novel topology agent. We compare our upgrade with the CAgent and significantly increase its L2RPN score by 10%. Further, we achieve a 25% better median survival time with our TTs included. Later analysis shows that almost all TTs are close to the base topology, explaining their robustness.

In this paper we consider the problem of scheduling two types of home appliances: ones with fixed power and flexible operation time and others with fixed operation time and flexible power. We consider energy consumption bounds that depend on the time of the day, which modify considerably the computational complexity of the appliance scheduling problems. Additionally, we also take into account the possibility that appliances with flexible operation time could be interrupted. The goal of the decision-making process is to minimize the electricity bill paid by the users and the discomfort perceived by them, according to a desired-tradeoff, subject to the operation constraints involved.

Since these problems turn out to be NP-hard, we describe several heuristics we designed, explaining their foundations. Finally, we analyze the computational results obtained for instances we generated and present the best algorithms to employ according to different priorities the users might have.

The integration of solar and wind energy sources in microgrids has witnessed significant growth, giving rise to distinct challenges due to their intermittent nature when it comes to achieving efficient microgrid control. However, estimating the parameters of the dynamic microgrid components facilitates capturing the complex and time-varying characteristics of renewable energy generation. This requires an accurate estimation of the parameters from the dynamic differential equations for effective modeling and control. In this research paper, we presented a novel methodology based on the integration of Bayesian inference and Neural ODEs. The Bayesian inference quantifies the uncertainty, and the Neural ODEs model the dynamic systems. By combining the strengths of both methods, we aimed to achieve a precise and robust parameter estimation of the dynamic microgrid components. The methodology is validated on a simulated microgrid that consists of a diesel generator, Solar PV array, double-fed induction generator, and a battery energy storage system. The results showed promised inferences estimation obtained from the parameter posterior distribution even in the presence of uncertainty. This can enhance our understanding of the dynamics of renewable energy systems and can contribute to the advancement of decision-making microgrid control strategies.

The implementation of advanced digital technologies in the conventional electric grid has triggered a transformation towards an intelligent network, known as Smart Grid. The associated benefits are diverse, ranging from more efficient energy management and demand response to the distributed integration of renewable energy sources. Ultimately, this transition promotes a more reliable, sustainable, and cost-effective energy supply. In this context, there is increasing recognition of the advantages of employing intelligent at edge to provide redundancy, virtualize functions that were previously in different proprietary hardware in the same device, or introduce new functionalities into the electric grid. This study focuses on conducting a comprehensive analysis on the key aspects to consider when implementing virtualized solutions in substations. Strategies have been sought to ensure the optimal deployment of virtualized nodes within the electrical sector, taking into account factors such as functional requirements, facility types, virtualization methodologies, and node specifications, among others. Furthermore, throughout the study, several virtualization tools have been analysed to determine their feasibility and the advantages they offer when integrated into the Smart Grid.

Amidst escalating environmental concerns and energy scarcity, the integration of distributed generation (DG) within distribution networks (DN) has emerged as a pivotal developmental trend. The uncertainty inherent in renewable energy output often disrupts DG networks. Notably, the dynamic correlation between key renewable sources, such as wind and solar energy, significantly influences the reliability analysis of these networks.To comprehensively assess the impact of wind-solar power output uncertainty and its dynamic correlation on DN reliability, this study leverages copula theory to express the dynamic correlation coefficient between wind and solar power. This coefficient is formulated as the dynamic correlation of wind-solar power through copula dynamic correlation coefficient. Employing an auto-regressive moving average (ARMA) model with constraints solved using maximum likelihood kernel (MLK), we construct the wind-solar joint output (WSJO) model. Subsequently, utilizing sequential Monte Carlo simulation (MCS) with the WSJO model, we analyze DN reliability. In case of DN failure, the WSJO model generates random samples of the wind-solar joint output sequence. Subsequent power restoration to governed islands enables the calculation of DN reliability indices. The WSJO model constructed in this study accounts for wind resource output uncertainty and dynamic correlation, aligning more closely with actual distributed generation output and enhancing the accuracy of reliability assessment. Finally, we simulate the improved IEEE-RBTS-BUS6-F4 system to underscore the crucial role of considering wind-solar energy's dynamic correlation in DN reliability assessment.

This study introduces a significant advancement in peer-to-peer (P2P) energy trading systems within smart grids, addressing a crucial gap in existing research by incorporating optimal energy storage capacities to accommodate varying energy demands resulting from lifestyle changes. Through a two-level optimization approach, aimed at maximizing self consumption and optimizing energy flow within the grid, we propose a novel energy management strategy. Our contribution lies in the introduction of a new layer of deep learning and rules control, forming a self-energy sharing system for each prosumer. This architecture, termed the smart node, integrates deep learning techniques, to predict and customize energy services through dynamic adjustment of lower and upper bounds of battery capacities. Additionally, we leverage cellular automaton (CA) approaches to establish sustainable consensus among P2P network users, enhancing the adaptability and efficiency of the energy management system. The results show that the proposed algorithm could reduce the energy consumed by the P2P community from the utility by around 20% and maximize the collective self-consumption by around 8% compared to conventional energy trading in microgrids.

In this study, we introduce an innovative deep learning framework designed to achieve precise detection, localization, and rate estimation of gas distribution pipeline system leakages. Our method surpasses conventional statistical approaches, particularly those based on Bayesian inference, by accommodating the system’s intricate behaviors, including variable usage and production from both sources and sinks. Notably, our approach demonstrates remarkable accuracy in localizing leakages even amidst multiple occurrences within the system. Specifically, achieving over 98% accuracy in single-leakage scenarios underscores its effectiveness. Furthermore, through data augmentation involving the introduction of noise into the training dataset, we significantly enhance the model’s performance, particularly when tested against real-world-like noisy data. This study not only showcases the efficacy of our proposed deep learning framework but also underscores its adaptability and robustness in addressing complex challenges in gas pipeline systems.

This paper presents a week-long scheduling approach to address the issues associated with uncertain stochastic generation. Specifically, the method is designed for active distribution networks (ADNs) hosting hybrid energy storages, composed by a hydrogen energy storage system (HESS) and a battery energy storage system (BESS). The inclusion of a pressurized HESS allows to balance energy over longer time periods, as opposed to methods considering only BESSs. To this end, this paper combines linearized models for the electricity grid with linearized models of the HESS to solve a tractable scheduling problem. The proposed optimal schedule consists of an active power trajectory at the grid connection point (GCP), called the dispatch plan, and the unit commitment schedule of a PEM fuel cell and electrolyzer system interfacing the electricity network with the HESS. Additionally, a bilevel model predictive control strategy is proposed, where the upper layer MPC computes a storage target accounting for the full horizon, while the lower layer computes the controllable resource setpoints to minimize the dispatch tracking error in each period. A numerical experiment shows the effectiveness of the proposed scheduling and control to accurately compute and track a dispatch plan over a full week. The results clearly show the benefits of combining a HESS with a BESS especially in periods where the prosumption is highly uncertain. Finally, we discuss the computational challenges associated with the weekly horizon and the use of a HESS that exhibits different dynamics than a BESS and propose an approach to mitigate the computational cost.