Electric Power Systems Research最新文献

This paper introduces a lightning current waveform measurement system developed at the University of Zagreb's high-voltage laboratory. It details the installation process and initial measurements of the prototype, which was deployed on a wind turbine in the southern part of Croatia, an area prone to winter lightning activity. The system employs two Rogowski coils - one high-frequency, high-amplitude (1 MHz, ±250 kA) and one low-frequency, low-amplitude (10 kHz, ±12.5 kA) - both fixed with magnets near the base of the wind turbine tower. The system efficiently captures both continuing-type currents, such as initial continuous currents, and pulse-type currents, such as return strokes and superimposed impulses. A detailed analysis of a typical upward lightning strike highlights the challenges in waveform measurement, including issues like 2-MHz oscillations in high-frequency sensor measurements and DC offsets in low-frequency sensor measurements. Validation against lightning location system data confirms the overall effectiveness of the current measurements, with the return stroke timestamps matching within the millisecond range.

The increasing integration of renewables, battery storage, and electric vehicles is leading communities to operate as microgrids within distribution networks. Managing multiple microgrids as a network of microgrids enhances benefits of individual microgrids. Although there are different approaches for energy management system of network of microgrids in the literature, this work presents a distributed communication approach for the energy management of network of microgrids by considering that each microgrid has a communication network with a limited number of participants, i.e., the communication network of each microgrid does not cover all other microgrids. The proposed method is compared to the centralized approach. In the test cases, the collective operational cost of microgrids are reduced from 1070$ to 54$, which corresponds to a cost reduction 93% of that achieved by the centralized approach. The results show that the proposed method can serve as an alternative to the centralized version considering investment costs and resilience. A simulation test case illustrates the indirect communication between two microgrids through a common neighbouring microgrid. Despite performing well, the proposed approach cannot surpass the centralized energy management system as expected. Furthermore, a scenario featuring an undesirable power transfer route is presented as a limitation case.

Real-time optimal control is crucial for the efficacy of Home Energy Management Systems (HEMS) in residential settings during actual operation. The time-varying and nonlinear nature of smart households — characterized by fluctuations in renewable energy generation, real-time electricity pricing, and load consumption — presents substantial challenges for both prediction and real-time control within HEMS. To tackle these issues, this paper introduces a real-time optimal control algorithm, augmented by predictive scheduling for HEMS. More specifically, the proposed real-time HEMS framework integrates an adaptive dynamic programming (ADP) algorithm, which is complemented by predictions of renewable energy generation and load consumption. Initially, data-driven methodologies generate accurate forecasts using available data collected and processed in real time. Gated Recurrent Unit (GRU) neural networks utilizing a range of data inputs such as electricity prices, battery charge/discharge rates, load consumption, and renewable energy generation, the system computes the optimal performance index function. Following this, we employ the ADP algorithm to reduce total electricity costs. This paper confirms the convergence properties of the value iteration ADP algorithm., demonstrating a monotonic approach of the iterative performance index function towards the optimal solution. The efficacy of the proposed algorithm is supported by numerical experiments, which verify its that solar energy efficiency has increased to 98% and electricity costs have been reduced by 64%.

This paper presents an experimental study on Power Quality (PQ) using indicators to assess the impact on loads supplied by a Low Voltage Open Structure Direct Current Nanogrid (LVOSDCN) utilizing off-the-shelf equipment and passively regulated in 24 V, implemented in real scale in the Amazon Region. Four tests were performed to evaluate the regular occurrence of PQ events with different irradiance profiles and commercial loads. The characteristics of the distribution grid, measurement instruments, and load groups are presented. As a result, variations in RMS voltage under the influence of the irradiance profile reach values above 1.1 PU with a duration of over 1 min, causing failures in some loads. Oscillatory transient events, resulting from the activation of DC-DC converters with time below a few milliseconds, without causing impacts. Finally, the evaluation of voltage ripple using the RMS ripple factor showed maximum values ​​close to 5 %, with different magnitudes over time and at different points of the distribution grid, also having an increase in circulating non-active power. These results are important because the operation of the real scale nanogrid and experimental setup show the magnitudes at which these events can cause failures or damage to the loads supplied by this system.

This paper presents a novel and non-iterative methodology for estimating the series capacitance ($C_s$) of power transformer windings based on terminal measurements through Frequency Response Analysis (FRA). Departing from conventional approaches that require detailed geometrical information, the proposed method utilizes practical terminal-based FRA, making $C_s$ estimation accessible to end-users. By categorizing the frequency response into low ($f_L$), mid ($f_M$), and high ($f_H$) frequency regions, targeted analysis is enabled. Extensive simulations and experimental studies on various windings, including single-layer, continuous disc, interleaved, and two-winding configurations, validate the method’s accuracy and versatility. The results show that estimated $C_s$ values closely match those from analytical calculations and Finite Element Method (FEM) simulations, with minimal errors. Key contributions of this work include a clear, step-by-step methodology for $C_s$ estimation through terminal-measured frequency response that eliminates the need for lookup tables, curve fitting, or model estimation, and demonstrates high accuracy under the impact of noise signals and transformer oil. The practical applicability of the proposed method is showcased through deformation diagnosis and voltage distribution analysis, highlighting its potential for widespread adoption in real-world scenarios.

This paper proposes a fuzzy credibility chance-constrained multi-objective optimization model to optimize market transactions in the electricity–gas–carbon sectors under uncertainty. The model aims to maximize the profit of an integrated energy service provider by incorporating a ladder-type carbon trading mechanism, which adjusts carbon prices based on emission levels, and detailed multi-energy flow constraints. To effectively manage uncertainties in electricity, gas, and carbon markets, we derive credibility distributions for uncertain variables and introduce fuzzy credibility chance constraints—tools that assess the likelihood of meeting multiple transactions under uncertainty. The proposed model hedges the risks associated with multiple uncertainties while maximizing the credibility of expected costs, effectively balancing risk and cost. Through simulation analysis on an IEEE 33-node power network and a 32-node heat network, the proposed model achieved a 9.8% reduction in total system cost and an 8.5% reduction in total carbon emissions. Additionally, the model effectively determined credibility levels under different risk preferences, demonstrating its robustness in enhancing electricity–gas–carbon trading and promoting a low-carbon economy. This research provides a novel planning method for formulating trading strategies in multi-energy markets, with significant real-world implications for energy management and environmental sustainability.

Heat pumps (HPs) are one of the most efficient heating technologies; their mass adoption will be required to decarbonize energy systems. However, to do so will require a better understanding of how they will impact electric grid load. Methods are needed to estimate not just their peak demand but also their impact on hourly load profiles. In this paper, we propose two methods, using easily accessible data, for estimating future hourly load profiles following the adoption of large populations of residential HPs. The first method uses feeder load data disaggregation while the second method uses annual space heating end-use energy consumption, both taking into account the temperature dependencies on coefficient of performance and output heat capacity. A case study based on data from Summerside, PE, Canada, is used to demonstrate and evaluate the two methods.

Accurate prediction of photovoltaic (PV) cluster power is crucial for the reliable and cost-effective operation of PV high penetration power systems. This paper introduces a method that utilizes time-frequency correlation. Firstly, the cluster power is decomposed using Complementary Ensemble Empirical Mode Decomposition with Adaptive Noise Algorithm (CEEMDAN) to extract more time-frequency information. Then, Kendall correlation coefficients are used to assess the consistency of time-frequency information across individual power plants and clusters within each frequency band. These coefficients are weighted according to the energy distribution in each frequency band to select the PV reference power station. Additionally, factors influencing PV power generation are taken into account to develop the PV impact factor. An Informer neural network is employed to predict the power output of the PV reference power plant. A trend inconsistency factor is introduced to adjust the PV cluster power variance. The final cluster prediction value is determined by correcting the linearly scaled variance using the adjusted variance. The method's feasibility and effectiveness are validated using real operational data from the PV cluster power plant in Alice Springs, Australia. This method offers a novel and highly accurate approach for forecasting future PV cluster power.