IET Energy Systems Integration最新文献

Accurate frequency modelling of inverter-based resource (IBR)-dominated power systems is crucial for ensuring stable, reliable and resilient operations, particularly given their inherent low-inertia characteristics and fast dynamics that traditional swing equation-based models inadequately capture. This paper explores neural ordinary differential equations (Neural ODEs) as a computationally efficient, data-driven framework for modelling power system frequency dynamics, specifically within microgrids integrating high penetrations of distributed energy resources (DERs). The developed neural ODEs framework incorporates a neural network architecture designed to capture input dynamics. By actively perturbing the system with a known signal, the Python-based neural ODEs framework was trained using measured system states and inputs, without the need for detailed system information. The framework, tested on a model of the Cordova, AK, microgrid, achieved a goodness of fit ranging from 60% to 99% across different state variables and maintained a mean square error in the $��1�0^{�-�6�}�$ p.u. range under square and step excitation signals. The proposed approach demonstrated robustness to measurement noise and initial condition variations while maintaining low computational complexity suitable for real-time power system control applications. Furthermore, transfer learning enabled the neural ODEs model to adapt to the following changes in system topology or generator dispatch, highlighting its effectiveness for dynamic microgrids with frequently evolving configurations and diverse DERs.

The global power grid is undergoing a significant transition with the increasing integration of renewable energy sources (RESs). This shift, characterised by a decline in conventional synchronous generators and a rise in inverter-based RESs, has resulted in reduced grid inertia, posing substantial challenges to frequency stability. To address these challenges, this paper proposes a reliability-aware optimisation framework for virtual inertia planning in grids with high RES penetration. The framework accounts for stochastic variations in load demand and renewable generation and optimises the required virtual inertia to ensure frequency stability during a contingency involving the largest generating unit. A reliability-driven methodology is integrated to determine the necessary virtual inertia constants and energy reserves in battery energy storage systems (BESSs) on an hourly basis, ensuring that the frequency nadir and RoCoF constraints are met under varying system conditions. The case study illustrated how the proposed approach determines BESS virtual inertia support requirements on an hourly basis, adapting to fluctuations in available system inertia and RES output. For instance, at 12:00 corresponding to peak solar generation, the required BESS inertia constant and energy reserve were determined to be 1.65 s and 2.56 MWh for 50% reliability increasing to 3.57 s and 5.27 MWh for 90% reliability. The total BESS capacity required was 76.3 MWh for 90% reliability compared to 46.25 MWh and 57.78 MWh for 50% and 70% reliability, respectively. These results underscore the significance of reliability aware inertia planning as higher reliability levels necessitate greater inertial support to mitigate the impact of contingencies involving the largest generating unit. Notably, during peak solar hours (e.g., 11:00, 13:00 and 15:00), the percentage of frequency constraint violations is reduced from nearly 100% to 4.38%, 9.86% and 7.12%, respectively, when virtual inertia support from BESS is applied at 90% reliability. An economic evaluation was carried out to assess BESS performance under different reliability targets. The results show that although annual benefits remain positive, the benefit–cost ratio declines from 1.12 at 50% to 1.06 at 90% and the NPV decreases from 2.06 M AUD to 1.50 M AUD. This indicates that higher reliability improves frequency stability but requires greater BESS investment, leading to diminishing economic returns. In this context, the proposed framework offers grid operators a decision-making tool to set reliability targets consistent with grid codes, contingency standards and investment capacity.

Addressing climate change, resource depletion and environmental degradation driven by fossil fuel dependence requires urgent development of sustainable energy systems. Virtual power plant (VPP) enhance economic performance and grid flexibility by aggregating distributed resources and flexible loads. Building on this concept, this study proposes a hydrogen park-level integrated energy system (H-PIES) under the VPP framework, incorporating user-side demand response (DR) and a tiered carbon trading mechanism for low-carbon, efficient energy management. First, a hybrid strategy combining alkaline electrolyzers (AEL) and proton exchange membrane electrolyzers (PEMEL) is proposed for power-to-gas (P2G) applications, considering hot/cold starts and load range limitations to maximise renewable energy utilisation. Second, a multi-dimensional flexible load response model is also introduced to enhance operational flexibility. Last, the tiered carbon trading mechanism is embedded into the objective function to incentivise low-carbon transitions. Simulation results indicate that the refined hydrogen utilisation strategy and hybrid electrolyser operation mode effectively accommodate renewable energy output, reducing total costs by 0.28% and 0.65%, respectively, compared to using only AEL or PEMEL. The tiered carbon trading mechanism effectively curbs emissions, whereas DR reduces system costs and emissions by 4.15% and 9.50%, respectively, and smooths load profiles, promoting overall system sustainability.

Achieving net-zero emissions requires a comprehensive transformation of energy infrastructure. The expansion of intermittent renewable generation together with increased demand, for example, from electric vehicles and heat pumps, exerts significant stress on grid stability and reliability, making power system flexibility essential for maintaining power and energy balance and enhancing system resilience. Urban buildings, with integrated heating, cooling, on-site generation, storage and flexible demand, constitute a substantial, cost-effective flexibility resource. In this paper, building flexibility is first defined and its physical drivers are identified. Quantification methods at the individual building scale, including physics-based simulations, machine-learning models and hybrid approaches supported by recent experiments, are then surveyed. Aggregation frameworks that include Virtual Power Plants and Virtual Energy Storage Systems are examined, comparing centralised, decentralised and hierarchical control strategies, ICT requirements and market-integration pathways. The participation of aggregated building flexibility in wholesale markets and its provision of distribution-level services are analysed. Enabling measures such as dynamic tariffs, demand-response programmes, smart-readiness indicators and relevant standards are critically evaluated. Finally, future research needs in integration, standardisation and policy development are outlined. That leveraging urban building flexibility may be crucial for developing a reliable, cost-effective, low-carbon power system is suggested by our analysis.

This paper deals with an updated frequency response model (FRM) for modern power grids incorporating a grid-forming (GFM) converter. This model aims to accurately capture the dynamic frequency behaviour of the GFM converters integrated into large-scale power systems. To achieve this, this paper estimates unknown parameters of the updated FRM to ensure a frequency response that closely matches the behaviour of the actual power grid. The updated FRM comprises two interconnected areas: the aggregated synchronous generator (ASG) area and the aggregated grid-forming (AGFM) area, linked by a virtual power synchronising interface. The parameter estimation is conducted using a curve-fitting-based methodology applied to frequency deviation data. First, the moment of inertia of each generation unit is determined based on observed frequency and active power deviations. Subsequently, the total load disturbance is estimated using the computed inertia values and the centre of inertia frequency deviation. Finally, the AGFM parameters including virtual damping, droop coefficient and virtual interconnection are estimated based on the ratio of the AGFM to the ASG speed deviations. The accuracy and effectiveness of the parameter estimation approach for the updated FRM are validated through simulations on a modified IEEE 39-bus test system using MATLAB/Simulink environment.

This article proposes a novel frequency and voltage control scheme for low-inertia electrical systems with high penetration of renewable energy sources (RES). A multi-timescale coordinated control scheme was proposed to optimally control inverter-based resources in different timescales. Accordingly, a two-stage stochastic optimisation framework has been developed for optimal operation of battery energy storage system (BESS) and voltage source converters (VSC) in hour-ahead and intra-hourly timescales, to counteract the effects of uncertainties in solar photovoltaic (PV) and load. Additionally, a novel real-time coordination framework was developed for fast frequency control, triggered by appliance switching/scheduling information through energy internet. Thus, real-time control is implemented as a pre-disturbance preventive action, appropriately acting with the load switching event. Furthermore, the proposed real-time frequency control is developed as a coordination strategy for primary regulation by adaptive VSC control and recovery control by the grid. Extensive simulations were performed to verify suitability of the proposed optimisation and control strategy in mitigating the effects of unforeseen uncertainties and scheduled events on system stability. Effectiveness of the proposed control is further verified by experimental validation on laboratory-scale hardware test setup.

As global energy demand continues to rise alongside the push for green technologies, artificial intelligence (AI) based power management systems play a pivotal role in achieving energy efficiency, grid stability and carbon footprint reduction, making them a vital component of future-ready building infrastructures. In this paper, an integrated AI based method for power management of building electrical systems is proposed. The main goal is to develop an accurate model to estimate the indoor temperature of building thermal zones, which is a critical aspect of energy management and occupant comfort. To achieve this, advanced modelling techniques are applied, specifically system identification and artificial neural networks (ANNs). Moreover, a sophisticated approach to real-time building energy management through accurate estimation of internal thermal zone gains is suggested, by applying heuristic parameter estimation techniques. This problem involves using the proposed ANN building model in the process of internal thermal gains estimation for each building thermal zone. By developing and validating these models, the aim is the efficiency of building electrical systems to be enhanced, the energy consumption be reduced, and the thermal comfort within buildings be improved, contributing to more sustainable and cost-effective building power management methods.

ON/OFF flexible loads (FLs), which account for a large proportion of demand side resources, have been proved to be effective in the frequency control of the power systems. The recovery process, which is essential for ON/OFF FLs, requires FLs to recover to their original ON/OFF status after activated. Other than thermostatically controlled loads (TCLs), non-TCLs cannot realise recovery through natural switching. The recovery control of non-TCLs may limit the capacity of bi-directional frequency control. Focusing on this problem, this paper proposes a bi-directional frequency control method which fully utilises the ON/OFF FLs in the recovery-process. By defining a combined fast-and-slow recovery-process (CFSRP), the activated ON/OFF FLs can still provide active power regulation service during their fast recovery-process (FRP). By designing a hierarchical strategy that considers bi-directional frequency control, the available regulation capacity of FLs and the frequency control capacity are increased. The case study results show that, compared to traditional methods that do not consider the CFSRP strategy, the proposed method reduces the maximum frequency deviation by over 4%.

In this work, the battery thermal management system (BTMS) using heat pipe and forced air cooling for NMC lithium-ion batteries was designed. The effect of air velocity on cooling performance was studied and compared between experimental and simulation results. All studies were conducted on lithium nickel manganese cobalt oxide (NMC) pouch cells with a 20 Ah capacity in seven series connections, under air velocities of 6.3, 9.5, and 12.7 m/s, with 4C discharge rates, at room temperature 22°C. The cooling performance was considered from two variables: the maximum temperature of the battery in the pack (T_max) and the maximum temperature difference of the battery in the pack (▵T_max). Both the experimental and simulation results indicated that increasing the air velocity has the effect of decreasing the T_max, while the ▵T_max did not differ significantly. The appropriated air velocity was 9.5 m/s. The behaviour from the simulation method was consistent with the experimental method, but the magnitude of the temperature fluctuations was still very large. At an air velocity of 9.5 m/s, T_max from simulation was only 33.1°C, while the experimental result was 44.3°C. The ▵T_max from the simulation was only 1.9°C, while the experimental result was 11.1°C. The main reason for the large difference was the properties of the materials used in the experiments, including batteries and heat pipes.

Global concern about climate change has accelerated the integration of renewable energy. To accommodate the high penetration of renewables at the distribution level and maintain system flexibility under a fully distributed architecture, this paper develops a voltage control strategy based on federated learning coordinated by a virtual power plant. A dynamic network partitioning method is introduced using a comprehensive performance index, along with an adaptive genetic algorithm featuring elite retention. An enhanced alternating direction method of multipliers with adaptive penalty modulation is employed to improve the convergence efficiency. Additionally, a two-stage encryption mechanism is applied to protect user privacy and ensure cybersecurity during distributed coordination. The effectiveness and feasibility of the proposed method are validated on a modified IEEE 33-bus system.