arXiv ScienceSearch

arXiv subjects

Marija Ilic

Publications and source records attributed to Marija Ilic.

At least 19 recordsLinked to original sources

Distributed Droop-Free Control of Grid-Forming Inverters with Dynamic Line and Load Modeling

Droop-free distributed control has emerged as a promising alternative to conventional linear droop control for coordinating inverter-based resources in AC microgrids. However, existing droop-free methods typically rely on quasi-steady state network models that neglect fast electromagnetic transients and assume a decoupled dependence of active and reactive power on frequency and voltage, respectively. In this paper, we propose a transient-aware droop-free distributed primary/secondary control framework for grid-forming (GFM) inverters. The control objective is to achieve proportional active and reactive power sharing, frequency regulation, and voltage regulation within prescribed bounds. The proposed approach incorporates dynamic models of network lines and loads into the control design, enabling accurate representation of transient behavior while capturing the coupled nature of power flow interactions. A gradient-based distributed control update is derived from an optimization formulation, in which voltage magnitude constraints are enforced through a projection operator. To validate the performance of the proposed control, real-time simulation studies are conducted on a four-inverter microgrid under load changes in both strong and weak grid conditions. The results demonstrate that the proposed method improves transient performance, particularly in weak grids, while preserving the same steady state operating conditions as existing QSS-based droop-free approaches.

eess.SY

Dynamic Storage Operation Under Uncertainty and the Reliability Externality: Implications for Capacity Investments

Energy storage is increasingly relied upon to meet short-term demand uncertainties from renewable variability and electrification. Unlike conventional generators, storage's contribution to reliability is policy-dependent and balances near-term arbitrage against future scarcity risk. We study how demand uncertainty alters such dynamic storage operation and how these operating decisions propagate into long-run investment outcomes. We formulate storage operation as an average-cost Markov decision process and embed the resulting stationary policies into a stylized capacity expansion framework. Demand uncertainty induces a precautionary storage policy which hedges against stochastic scarcity, leading to materially different post-storage demand distributions relative to perfect-foresight benchmarks. We additionally demonstrate that the reliability externality characteristic of electricity markets interacts with uncertainty in a manner that uniquely distorts both storage operation and investment.

eess.SY

Composite Control of Grid-Following Inverters for Stabilizing AI-Induced Fast Power Disturbances

AI data center loads create query-driven power transients on millisecond timescales. Such loads can violate the timescale separation assumptions underlying internal inverter control of grid-following resources collocated with data centers as supplementary generation. This paper develops a singular perturbation-based modeling and control for stabilizing fast power imbalances. We show that physically-implementable droop control is derived and valid by requiring reduced-system stability rather than being imposed a priori, and that AI workloads satisfy a bounded-rate disturbance class due to physical filtering in power delivery hardware. The analysis yields explicit gain bounds linking inverter parameters to disturbance rejection performance, a modulation admissibility condition ensuring physical realizability of the feedback linearizing control, and a feasibility condition identifying the maximum tolerable load ramp rate. Numerical simulations validate the theoretical predictions under stochastic AI transients.

eess.SY

AC-OPF Feasibility Analysis and Sensitivity-Guided Capacitor Placement in a High-PV Islanded Microgrid

This paper presents a comparative AC Optimal Power Flow study on a real world city scale islanded microgrid with high solar PV penetration, implemented within a Digital Twin framework. Four objective function cases economic dispatch, voltage stress exposure via PV power factor variation, then optimal load delivery, and capacitor enhanced economic dispatch as recovery options are evaluated over a 47 hour time series horizon on the same network under a shared loading scenario. Optimization sensitivities OSQ and OSV extracted from all cases are combined into a composite placement score used to rank candidate buses for shunt capacitor upgrades. A post processing planning optimization balances capacitor upgrade cost against avoided value-of-lost-load, enabling direct economic comparison of infrastructure investment versus reliability penalties. Results demonstrate that sensitivity guided capacitor placement restores full load service across the horizon and provides targeted reactive support at a quantifiable cost trade off against corrective load shedding.

eess.SY

Peak-Load Pricing and Investment Cost Recovery with Duration-Limited Storage

Energy storage shifts energy from off-peak periods to on-peak periods. Unlike conventional generation, storage is duration-limited: the stored energy capacity constrains the duration over which it can supply power. To understand how these constraints affect optimal pricing and investment decisions, we extend the classic two-period peak-load pricing model to include duration-limited storage. By adopting assumptions typical of solar-dominated systems, we link on- and off-peak prices to storage investment costs, round-trip efficiency, and the duration of the peak period. The bulk of the scarcity premium from on-peak prices is associated with the fixed costs of storage as opposed to variable costs stemming from round-trip efficiency losses. Unlike conventional generators, the binding duration constraints lead storage to recover energy capacity costs on a per-peak-event basis instead of amortizing these costs over total peak hours. A numerical example illustrates the implications for equilibrium prices and capacity investment.

eess.SY

External Data-Enhanced Meta-Representation for Adaptive Probabilistic Load Forecasting

Accurate residential load forecasting is critical for power system reliability with rising renewable integration and demand-side flexibility. However, most statistical and machine learning models treat external factors, such as weather, calendar effects, and pricing, as extra input, ignoring their heterogeneity, and thus limiting the extraction of useful external information. We propose a paradigm shift: external data should serve as meta-knowledge to dynamically adapt the forecasting model itself. Based on this idea, we design a meta-representation framework using hypernetworks that modulate selected parameters of a base Deep Learning (DL) model in response to external conditions. This provides both expressivity and adaptability. We further integrate a Mixture-of-Experts (MoE) mechanism to enhance efficiency through selective expert activation, while improving robustness by filtering redundant external inputs. The resulting model, dubbed as a Meta Mixture of Experts for External data (M2oE2), achieves substantial improvements in accuracy and robustness with limited additional overhead, outperforming existing state-of-the-art methods in diverse load datasets. The dataset and source code are publicly available at https://github.com/haorandd/M2oE2\_load\_forecast.git.

cs.LG

Distributed component-level modeling and control of energy dynamics in electric power systems

The widespread deployment of power electronic technologies is transforming modern power systems into fast, nonlinear, and heterogeneous networks. Conventional modeling and control approaches, rooted in quasi-static analysis and centralized architectures, are inadequate for these converter-dominated systems operating on fast timescales with diverse and proprietary component models. This paper adopts and extends a previously introduced energy space modeling framework grounded in energy conservation principles to address these challenges. We generalize the notion of a port interaction variable, which encodes energy exchange between interconnected components in a unified manner. A multilayered distributed control architecture is proposed in which dynamics of each component are lifted to a linear energy space through well-defined mappings. Distributed control with provable convergence guarantees is derived in energy space using only local states and minimal neighbor information communicated through port interactions. The framework is validated using two examples: voltage regulation in an inverter-controlled RLC circuit and frequency regulation of a synchronous generator. The energy-based controllers show improved transient and steady-state performance with reduced control effort compared to conventional methods.

eess.SY

ExARNN: An Environment-Driven Adaptive RNN for Learning Non-Stationary Power Dynamics

Non-stationary power system dynamics, influenced by renewable energy variability, evolving demand patterns, and climate change, are becoming increasingly complex. Accurately capturing these dynamics requires a model capable of adapting to environmental factors. Traditional models, including Recurrent Neural Networks (RNNs), lack efficient mechanisms to encode external factors, such as time or environmental data, for dynamic adaptation. To address this, we propose the External Adaptive RNN (ExARNN), a novel framework that integrates external data (e.g., weather, time) to continuously adjust the parameters of a base RNN. ExARNN achieves this through a hierarchical hypernetwork design, using Neural Controlled Differential Equations (NCDE) to process external data and generate RNN parameters adaptively. This approach enables ExARNN to handle inconsistent timestamps between power and external measurements, ensuring continuous adaptation. Extensive forecasting tests demonstrate ExARNN's superiority over established baseline models.

cs.LG

Dissipativity Conditions for Maximum Dynamic Loadability

In this paper we consider a possibility of stabilizing very fast electromagnetic interactions between Inverter Based Resources (IBRs), known as the Control Induced System Stability problems. We propose that when these oscillatory interactions are controlled the ability of the grid to deliver power to loads at high rates will be greatly increased. We refer to this grid property as the dynamic grid loadability. The approach is to start by modeling the dynamical behavior of all components. Next, to avoid excessive complexity, interactions between components are captured in terms of unified technology-agnostic aggregate variables, instantaneous power and rate of change of instantaneous reactive power. Sufficient dissipativity conditions in terms of rate of change of energy conversion in components themselves and bounds on their rate of change of interactions are derived in support of achieving the maximum system loadability. These physically intuitive conditions are then used to derive methods to increase loadability using high switching frequency reactive power sources. Numerical simulations confirm the theoretical calculations, and shows dynamic load-side reactive power support increases stable dynamic loadability regions.

eess.SY

A Mixed Integer Quadratic Program for Valuing the Impact of Price and Forecast Uncertainty for Wind Generators

Owners of wind power plants are exposed to financial risk in wholesale electricity markets due to the uncertain nature of wind forecasts and price volatility. In the event of a wind shortfall, the plant may have to repurchase power at a higher price in the real-time market. However, reducing the power offered in the day-ahead market may also be interpreted by regulators as physical withholding. We formulate and solve a mixed-integer quadratic program (MIQP) that prices the uncertain portion of a wind generator's forecast to hedge against uncertainties and which addresses concerns around withholding. We exploit the structure of the MIQP inputs to introduce additional constraints to improve computation time. Additionally, we provide a qualitative approach for generators and regulators to interpret the results of the MIQP. Finally, we simulate a real-world application for a wind farm in New York using past wind forecasts and NYISO prices.

eess.SY

Uniform Pricing vs Pay as Bid in 100%-Renewables Electricity Markets: A Game-theoretical Analysis

This paper evaluates market equilibrium under different pricing mechanisms in a two-settlement 100%-renewables electricity market. Given general probability distributions of renewable energy, we establish game-theoretical models to analyze equilibrium bidding strategies, market prices, and profits under uniform pricing (UP) and pay-as-bid pricing (PAB). We prove that UP can incentivize suppliers to withhold bidding quantities and lead to price spikes. PAB can reduce the market price, but it may lead to a mixed-strategy price equilibrium. Then, we present a regulated uniform pricing scheme (RUP) based on suppliers' marginal costs that include penalty costs for real-time deviations. We show that RUP can achieve lower yet positive prices and profits compared with PAB in a duopoly market, which approximates the least-cost system outcome. Simulations with synthetic and real data find that under PAB and RUP, higher uncertainty of renewables and real-time shortage penalty prices can increase the market price by encouraging lower bidding quantities, thereby increasing suppliers' profits.

eess.SY

Dynamic Reserves for Managing Wind Power

In this paper, we suggest that reserves must be computed dynamically to account for wind power volatility. We formalize the notion of dynamic reserves in support of sequential Day-Ahead-Market (DAM) and Real-Time-Market (RTM) clearing and discuss several ways of accounting for wind power uncertainties. Three different wind integration protocols are considered and their impact on stakeholders, and, notably, on the need for dynamic reserves is analyzed. An optimal strategy is proposed which requires the least reserve as long as the bids are truthful and implementable with high confidence. This can be checked by the existing market monitor and/or by a separate risk buro. Concepts are illustrated in part using publicly available information about NYISO.

eess.SY

Data-based optimal estimation of frequency bias: The case of Southwest Power Pool

In this paper, we introduce a method to optimally estimate time-varying frequency bias. Current industry practice is to assume that frequency bias is changing only on annual basis. We suggest that this improved time-dependent bias estimate can be used to reduce the cost of frequency regulation needed to meet industry standards requested by the North American Electric Reliability Corporation (NERC). Optimization of time-varying frequency bias is posed as a parameter estimation (calibration) problem whose implementation utilizes online system measurements. It is further shown how this result can be used to estimate intra-dispatch load deviations. This knowledge is needed to estimate more accurately regulation reserve needed, and to therefore reduce overall regulation cost. Methods can be introduced to give incentives to demand response to participate in frequency regulation. Overall, we show the importance of incorporating knowledge of physics-based models for data-enabled parameter estimation of physical systems.

eess.SY

Interactive Planning and Operations using Peak Load Pricing in Distribution Systems

The emergence of Distributed Energy Resources (DERs) provides both challenges and opportunities for the planning and operations of distribution systems. These resources can be deployed in a manner that is either complementary to or in competition with traditional network operations and planning as the DERs can provide numerous important services to grid operators and utilities. This paper presents a novel method to estimate the trade off between DERs and traditional investments using a dynamic Peak-Load Pricing (PLP) methodology. PLP is a pricing strategy for a time-dependent quantity of a non-storable commodity and is based on the theory of long-run marginal costs. Importantly PLP deals with the trade-off between capacity utilization and consumer welfare. Therefore, the capacity price is set at a point where the cost of investment is exactly offset by the additional social welfare that the investment would bring. Importantly it allows for capital cost recovery with no uplift payments. This dynamic PLP methodology is an interactive planning and operations model based on the Dynamic Monitoring and Decision Systems (DyMonDS) Framework which helps to align physical, information, and economic incentives across many stakeholders within the electric energy system. The DyMonDS framework helps to solve the drawbacks of PLP, which are related to the computational complexity of the PLP models considering different technologies with different payback periods over a long investment horizon. Results show that the dynamic PLP model can accurately value the impact of different technologies in reducing congestion and increasing the number of customers served. This allows the network operator to easily identify which technologies should be chosen in each investment cycle.

eess.SY

Advisory Tool for Managing Failure Cascades in Systems with Wind Power

This paper concerns the resilience of systems with wind power upon wind reduction by evaluating the potential of corrective actions, such as generation and load dispatch, on minimizing the effects of transmission line failures. Three functions (grid, consumer-centric loss, and resilience impact) are used to statistically evaluate the criticality of initial contingent failures and wind reductions. Our model is learned with Monte Carlo, convex optimization, and adaptive selection, illustrated on the IEEE-30 and IEEE-300 bus systems with both AC and DC models. We highlight the impact of wind reductions and propose physically implementable solutions.

eess.SY

Towards Statistical Methods for Minimizing Effects of Failure Cascades

This paper concerns the potential of corrective actions, such as generation and load dispatch on minimizing the effects of transmission line failures in electric power systems. Three loss functions (grid-centric, consumer-centric, and influence localization) are used to statistically evaluate the criticality of initial contingent failures. A learning scheme for both AC and DC grid models combine a Monte Carlo approach with a convex dynamic programming formulation and introduces an adaptive selection process, illustrated on the IEEE-30 bus system.

eess.SY

Valuing Uncertainties in Wind Generation: An Agent-Based Optimization Approach

The increasing integration of variable renewable energy sources such as wind and solar will require new methods of managing generation uncertainty. Existing practices of uncertainty management for these resources largely focuses around modifying the energy offers of such resources in the quantity domain and from a centralized system operator consideration of these uncertainties. This paper proposes an approach to instead consider these uncertainties in the price domain, where more uncertain power is offered at a higher price instead of restricting the quantity offered. We demonstrate system-level impacts on a modified version of the RTS-GMLC system where wind generators create market offers valuing their uncertainties over scenario set of day-ahead production forecasts. The results are compared with a dispatch method in which wind energy is offered at zero marginal price and restricted based on the forecast percentile.

eess.SY

Cyber-resilient Automatic Generation Control for Systems of AC Microgrids

In this paper we propose a co-design of the secondary frequency regulation in systems of AC microgrids and its cyber securty solutions. We term the secondary frequency regulator a Micro-Automatic Generation Control (Micro-AGC) for highlighting its same functionality as the AGC in bulk power systems. We identify sensory challenges and cyber threats facing the Micro-AGC. To address the sensory challenges, we introduce a new microgrid model by exploiting the rank-one deficiency property of microgrid dynamics. This model is used to pose an optimal Micro-AGC control problem that is easily implemented, because it does not require fast frequency measurements. An end-to-end cyber security solution to the False Data Injection (FDI) attack detection and mitigation is developed for the proposed Micro-AGC. The front-end barrier of applying off-the-shelf algorithms for cyber attack detection is removed by introducing a data-driven modeling approach. Finally, we propose an observer-based corrective control for an islanded microgrid and a collaborative mitigation schemes in systems of AC microgrids. We demonstrate a collaborative role of systems of microgrids during cyber attacks. The performance of the proposed cyber-resilient Micro-AGC is tested in a system of two networked microgrids.

eess.SY