Search

The Energy Communities (EC) framework facilitates the active engagement of energy entities. It brings about a fundamental shift in the energy sector by effectively managing Distributed Energy Resources (DERs) and advancing a decentralized energy system. The implementation of this technology allows the electrification of rural or mountainous regions by addressing the obstacles associated with power grid maintenance through substantial restructuring of the underlying energy distribution framework. The present review aimed to investigate and examine the significance of EC structures and to start an inclusive foundation for the broader implementation of EC for energy decentralization to figure out the research gaps and ensure that Machine Learning (ML) based solutions are essential tools to study and further discussed the several ML-based algorithms based on their objectives. Moreover, a comprehensive literature review is conducted to compare the possible strategies and tools that could be implemented in EC. Furthermore, different solution tools are organized based on their advantages, such as Demand Response (DR), forecasting, and load management goals. Hence, it can be inferred that Reinforcement Learning (RL) methodologies exhibit considerable potential in the control field. In contrast, supervised and unsupervised approaches are essential in predicting tasks. Based on the existing knowledge, the present study can conclude that ML-based solution methods are of significant importance for developing an effective energy decentralization platform.

To redefine the mobility and energy usage patterns in urban cities, the exponential rise in fast-charging loads of electric vehicles (EVs) necessitates a new operating paradigm. In this context, a competitive market emerges where multiple oligarchic Charging-as-a-Service (CaaS) providers vie for profit. However, analyzing such systems, especially in a long-term rolling horizon, poses significant challenges due to complex strategic interactions among multiple entities. This article proposes an agent-based model to analyze the evolution of spatial CaaS markets. Firstly, we propose a strategic pricing game among CaaS firms based on multinomial logit user choice model, and formally prove the existence of Nash equilibria. For equilibrium finding, an iterative algorithm method leveraging the Nikaido–Isoda function and Barzilai-Borwein step size rule is presented. Furthermore, an agent-based market simulation algorithm is developed to model investment decisions of CaaS firm agents based on equilibrium prices. The proposed model is validated using real world data from Shenzhen, and results indicate the efficacy of the equilibrium solving algorithm in reducing computational burden. This study reveals that deregulated competition among CaaS providers, while promoting social welfare, poses risks to long-term sustainability due to regional inequality and market saturation. Timely policy intervention is recommended to guide the sustainable evolution of the CaaS economy.

As the number of electric vehicles (EVs) increases, EV charging demand is also growing rapidly. In the smart grid environment, there is an urgent need for green charging stations (GCS) to effectively manage the internal photovoltaic (PV), energy storage system (ESS), charging behaviors of EVs and energy transactions with entities. In this paper, a novel EV classification approach was proposed for GCS, of which the objective is to minimize the total cost of energy trading between charging station and entities. Based on EV charging requirements and risk preference for participating in demand response (DR), EVs are classified into three categories: (1) regular; (2) conservative; and (3) vehicle to grid(V2G). For different types of EVs, charging price schemes are proposed for GCS to stimulate more EVs to interact with the grid and duly dis-charging their power in the capacity of V2G runners. In dealing with the uncertainty issues caused by photovoltaic generation, chance-constrained and probabilistic sequence methods are employed to solve the proposed framework with satisfactory computational efficiency. Case studies based on genuine PV generation, load and pricing data are carried out to verify the proposed framework. It is demonstrated V2G EVs can dramatically reduce the total cost to the GCS. Besides, V2G EVs also can greatly reduce the charging cost compared to regular and conservative EVs.

Since the number of electric vehicles (EVs) grows rapidly, the EVs charging schedule problem has attracted more attention than before. On account of continuous and random power requirement from EVs, the burden of grid will increase and thus jeopardize the stability of grid. In this paper, an accessible close-loop vehicle-to-vehicle (V2V) charging mechanism with interaction of energy and information under charging station is proposed to cope with EVs charging schedule problem. And the non-cooperative game model is adopted to analyze the V2V charging mechanism. Furthermore, the optimal strategy algorithm is introduced to get the Nash Equilibrium (NE) of the game. Simulation with 10 EVs illustrates that V2V charging with non-cooperative game can decrease the load variance (0.0701) compared to V2V charging without the game (1.7248). And the convergence of the optimal strategy algorithm can reach in less than 20 iterations.

Electric vehicles (EVs) combined with low-carbon generator sets can significantly reduce CO2 emissions in the transportation sector. EVs can provide flexible auxiliary services for the grid through proper power dispatching, which benefits both EVs and the grid. This paper discusses the optimal charging and discharging plan for EVs for the purpose of a distribution system operator (DSO) while obtaining the optimal day-ahead bidding strategy for DSO in the spot market. Considering the different charging demands of EV users, two contract models of vehicle-to-grid (V2G) and smart charging services are introduced. Then, K-means clustering is adopted to manage the spatiotemporal uncertainties of EVs, in order to maximize the expected marginal revenue of DSO. Then the optimization model is proposed by considering the real conditions campus grid in Shenzhen and the electricity market in Guangdong province. The result shows that the DSO could obtain the marginal revenue not only by the optimal day-ahead bidding strategy but also by scheduling EVs’ charging and discharging. In addition, EV users could reduce the charging cost by charging their EVs during low electricity price periods and get extra revenue by exporting power to the utility grid during high price periods.

The number of electric vehicles (EVs) is growing rapidly due to environmental concerns and political supports. To better coordinate the charging behaviors of large numbers of EVs, the EV aggregator is an important intermediate between the electricity market and the EVs. This work proposes a two-stage optimal operational regime for an EV aggregator who participates in multiple markets under multiple uncertainties. In the day-ahead stage, a joint offering model is proposed to help the aggregator concurrently participate in both the energy and reserve markets under market price uncertainties. The concept of reserve capacity ratio is introduced in the joint offering model to allow different decisions for aggregators with various risk preferences. In the real-time stage, a rolling horizon control-based multi-resolution scheduling model is proposed to minimize the energy deviation cost under reserve deployment uncertainty. The uncertainties are modeled using representative scenarios and the stochastic optimization approach is applied to acquire the optimal solutions. The numerical results suggest that the proposed operational regime can significantly improve the profitability of the studied EV aggregator. The impact of reserve capacity ratios is also analyzed.

Water is a precious resource for agriculture and most of the land is irrigated by tube wells. Diesel engines and electricity-operated pumps are widely used to fulfill irrigation water requirements; such conventional systems are inefficient and costly. With rising concerns about global warming, it is important to choose renewable energy source. In this study, SPVWPS has been optimally designed considering the water requirement, solar resources, tilt angle and orientation, losses in both systems and performance ratio. A PVSyst and SoSiT simulation tools were used to perform simulation analysis of the designed solar photovoltaic WPS. After designing and performance analysis, farmers were interviewed during fieldwork to assess socioeconomic impacts. In the result section, performance of PV system is analyzed at various tilt angles and it is established that system installed at a 15° tilt angle is more efficient. The annual PV array virtual energy at MPP of designed photovoltaic system is 33342 kWh and the annual energy available to operate the WPS is 23502 kWh. Module array mismatch and ohmic wiring losses are 374.16 kWh and 298.83 kWh, respectively. The total annual water demand of the selected site is 80769 m³ and designed SPWPS pumped 75054 m³ of water, supplying 92.93% of the irrigation demand. The normalized values of the effective energy, system losses, collection losses and unused energy in the SPVWP system are 2.6 kW/kWp/day, 0.69 kW/kWp/day, 0.72 kW/kWp/day and 0.48 kW/kWp/day, respectively. The annual average performance ratio of the proposed system is 74.62%. The results of the interviews showed that 70% of farmers are extremely satisfied with the performance of SPVWPS and 84% of farmers indicated that they did not incur any operating costs. The unit cost of the SPWPS is 0.17 €/kWh, which is 56.41% and 19.04% less expensive than the cost of diesel and grid electricity.

The implementation of machine learning methods in home energy management have been shown to be a feasible alternative in the minimization of electricity cost. These methods regulate the home electric appliance systems, which contribute to the most critical loads in a household, thus enabling consumers to save electricity while still enhancing their comfort. Furthermore, renewable energy supplies are continuously integrating with other electricity resources in number of homes that is an important component to optimize energy consumption which result in the reduction of peak load and can bring economic benefits. In this paper, a reinforcement learning algorithm is explored for monitoring household electric appliances with the intention of lowering energy consumption through properly optimizing and addressing the best use renewable energy resources. The proposed method does not necessitate any previous information or knowledge of the uncertain dynamics and parameters of different household electric appliances. Simulation-based findings using real-time data validate the efficiency and reliability of the proposed method.

To increase the energy utilization efficiency, it becomes fairly promising to convert the surplus electricity from renewable generation to other forms of energy for multi-dimensional consumption. In this paper, we propose a hybrid energy scheduling model for a multi-energy microgrid with the integration of the hydrogen energy storage system (HESS) and the heat storage system (HSS). In our study, the operational uncertainties induced by renewables and loads (including electrical, hydrogen, and heat demand) are comprehensively considered. We investigate such an operating regime that HESS stores the surplus electricity in case of abundant renewable generation and generates electricity through hydrogen fuel cells otherwise. Further, heat units including HESS, combined heat and power (CHP), and external heat suppliers are modeled in this paper. We split the decision-makings of energy scheduling for both the day-ahead stage and real-time stage to tackle the power balancing issues. To effectively solve the aforementioned optimization model, a flexible weighted Model Predictive Control (weighted-MPC) strategy is proposed, in which the receding horizon can be suitably adjusted according to the forecasting accuracy of system uncertainties. The effectiveness of the proposed hybrid model for microgrid energy scheduling is comprehensively verified through extensive case studies.

Machine learning approaches have diverse applications in forecasting electrical energy consumption using smart meter data. Various classification techniques and clustering methods analyze smart meter data for accurately forecasting the electrical appliance consumption and peak demand. Electrical appliance forecasting and peak demand forecasting play a vital and key role in planning, maintenance and automation development for electrical power system. However, there is always a variation between electrical appliance consumption and appliance energy demand due to certain parameters including losses in lines and appliance and mismanagement of appliance energy demand. Detail scrutiny of smart meter data is required to identify the decisive attributes and major cause of variation between electrical appliance consumption and customers’ peak demand. This paper proposed a hybrid method based on Machine learning for forecasting appliance consumption and peak demand. We have deployed faster k-medoids clustering, support vector machine and artificial neural network for forecasting appliance consumption and customers’ peak demand. The proposed algorithm achieves 99.2% accuracy in forecasting electrical appliance consumption which is much better compared to state-of-the-art in same field. Experimental results validate the effectiveness of the proposed method in forecasting the electrical appliance consumption using smart meter data.

In recent years, there has been notable progress in the development of perovskite solar cells (PSCs), marked by significant advancements in efficiency, stability, and scalability. Among different device architectures and technical routes, mesoporous perovskite solar cells (MPSCs) based on TiO2/ZrO2/carbon scaffold and screen-printing fabrication process have shown unique advantages for mass production and commercialization due to the low material cost and scalable fabrication process. Through efforts on material optimization, interface engineering, and device design, the power conversion efficiency of MPSCs has steadily increased from the initial 6.64 ​% in 2013 to current 22.22 ​%. Attributing to the screen-printing-based fabrication process, printable mesoscopic PSCs have enabled large-area devices, from mini-modules to modules, and solar panels. In this review, we discuss the development and latest advances of MPSCs, and provide outlooks for further improving the performance of this promising photovoltaic technology.