Searcharxiv⌕ Search

arXiv subjects

Geert Deconinck

Publications and source records attributed to Geert Deconinck.

At least 19 recordsLinked to original sources
arXiv2025

Uncertainty quantification in load profiles with rising EV and PV adoption: the case of residential, industrial, and office buildings

The integration of photovoltaic (PV) generation and electric vehicle (EV) charging introduces significant uncertainty in electricity consumption patterns, particularly at the distribution level. This paper presents a comparative study for selecting metrics for uncertainty quantification (UQ) for net load profiles of residential, industrial, and office buildings under increased DER penetration. A variety of statistical metrics is evaluated for their usefulness in quantifying uncertainty, including, but not limited to, standard deviation, entropy, ramps, and distance metrics. The proposed metrics are classified into baseline-free, with baseline and error-based. These UQ metrics are evaluated for increased penetration of EV and PV. The results highlight suitable metrics to quantify uncertainty per consumer type and demonstrate how net load uncertainty is affected by EV and PV adoption. Additionally, it is observed that joint consideration of EV and PV can reduce overall uncertainty due to compensatory effects of EV charging and PV generation due to temporal alignment during the day. Uncertainty reduction is observed across all datasets and is most pronounced for the office building dataset.

eess.SY↗
arXiv2025

Grid-Aware Flexibility Operation of Behind-the-Meter Assets: A review of Objectives and Constraints

The high penetration of distributed energy resources (DERs) in low-voltage distribution networks (LVDNs) often leads to network instability and congestion. Discovering the flexibility potential of behind- the-meter (BTM) assets offers a promising solution to these challenges, providing benefits for both prosumers and grid operators. This review focuses on the objectives and constraints associated with the operation of BTM flexibility resources in LVDNs. We propose a new classification framework for network-aware flexibility modelling that incorporates prosumer objectives, flexibility sources, and both local and grid-level constraints. This review identifies research gaps in prosumer-centric grid considerations, control strategies, flexibility preferences, and scenarios in the use of BTM resources.

eess.SY↗
arXiv2024

A concise introduction to Koopman operator theory and the Extended Dynamic Mode Decomposition

The framework of Koopman operator theory is discussed along with its connections to Dynamic Mode Decomposition (DMD) and (Kernel) Extended Dynamic Mode Decomposition (EDMD). This paper provides a succinct overview with consistent notation. The authors hope to provide an exposition that more naturally emphasizes the connections between theory and algorithms which may result in a sense of clarity on the subject.

math.NA↗
arXiv2024

Improved Physics-Informed Neural Network based AC Power Flow for Distribution Networks

Power flow analysis plays a critical role in the control and operation of power systems. The high computational burden of traditional solution methods led to a shift towards data-driven approaches, exploiting the availability of digital metering data. However, data-driven approaches, such as deep learning, have not yet won the trust of operators as they are agnostic to the underlying physical model and have poor performances in regimes with limited observability. To address these challenges, this paper proposes a new, physics-informed model. More specifically, a novel physics-informed loss function is developed that can be used to train (deep) neural networks aimed at power flow simulation. The loss function is not only based on the theoretical AC power flow equations that govern the problem but also incorporates real physical line losses, resulting in higher loss accuracy and increased learning potential. The proposed model is used to train a Graph Neural Network (GNN) and is evaluated on a small 3-bus test case both against another physics-informed GNN that does not incorporate physical losses and against a model-free technique. The validation results show that the proposed model outperforms the conventional physics-informed network on all used performance metrics. Even more interesting is that the model shows strong prediction capabilities when tested on scenarios outside the training sample set, something that is a substantial deficiency of model-free techniques.

eess.SY↗
arXiv2024

Impact of Dynamic Operating Envelopes on Distribution Network Hosting Capacity for Electric Vehicles

The examination of the maximum number of electric vehicles (EVs) that can be integrated into the distribution network (DN) without causing any operational incidents has become increasingly crucial as EV penetration rises. This issue can be addressed by utilizing dynamic operating envelopes (DOEs), which are generated based on the grid status. While DOEs improve the hosting capacity of the DN for EVs (EV-HC) by restricting the operational parameters of the network, they also alter the amount of energy needed for charging each EV, resulting in a decrease in the quality of service (QoS). This study proposes a network-aware hosting capacity framework for EVs (EV-NAHC) that i) aims to assess the effects of DOEs on active distribution networks, ii) introduces a novel definition for HC and calculates the EV-NAHC based on the aggregated QoS of all customers. A small-scale Belgian feeder is utilized to examine the proposed framework. The results show a substantial increase in the EV-NAHC with low, medium, and high-daily charging energy scenarios.

eess.SY↗
arXiv2024

Privacy-Preserving Distributed Learning for Residential Short-Term Load Forecasting

In the realm of power systems, the increasing involvement of residential users in load forecasting applications has heightened concerns about data privacy. Specifically, the load data can inadvertently reveal the daily routines of residential users, thereby posing a risk to their property security. While federated learning (FL) has been employed to safeguard user privacy by enabling model training without the exchange of raw data, these FL models have shown vulnerabilities to emerging attack techniques, such as Deep Leakage from Gradients and poisoning attacks. To counteract these, we initially employ a Secure-Aggregation (SecAgg) algorithm that leverages multiparty computation cryptographic techniques to mitigate the risk of gradient leakage. However, the introduction of SecAgg necessitates the deployment of additional sub-center servers for executing the multiparty computation protocol, thereby escalating computational complexity and reducing system robustness, especially in scenarios where one or more sub-centers are unavailable. To address these challenges, we introduce a Markovian Switching-based distributed training framework, the convergence of which is substantiated through rigorous theoretical analysis. The Distributed Markovian Switching (DMS) topology shows strong robustness towards the poisoning attacks as well. Case studies employing real-world power system load data validate the efficacy of our proposed algorithm. It not only significantly minimizes communication complexity but also maintains accuracy levels comparable to traditional FL methods, thereby enhancing the scalability of our load forecasting algorithm.

cs.LG↗
arXiv2022

Combined Peak Reduction and Self-Consumption Using Proximal Policy Optimization

Residential demand response programs aim to activate demand flexibility at the household level. In recent years, reinforcement learning (RL) has gained significant attention for these type of applications. A major challenge of RL algorithms is data efficiency. New RL algorithms, such as proximal policy optimisation (PPO), have tried to increase data efficiency. Additionally, combining RL with transfer learning has been proposed in an effort to mitigate this challenge. In this work, we further improve upon state-of-the-art transfer learning performance by incorporating demand response domain knowledge into the learning pipeline. We evaluate our approach on a demand response use case where peak shaving and self-consumption is incentivised by means of a capacity tariff. We show our adapted version of PPO, combined with transfer learning, reduces cost by 14.51% compared to a regular hysteresis controller and by 6.68% compared to traditional PPO.

eess.SY↗
arXiv2022

Anomaly Detection in Automatic Generation Control Systems Based on Traffic Pattern Analysis and Deep Transfer Learning

In modern highly interconnected power grids, automatic generation control (AGC) is crucial in maintaining the stability of the power grid. The dependence of the AGC system on the information and communications technology (ICT) system makes it vulnerable to various types of cyber-attacks. Thus, information flow (IF) analysis and anomaly detection became paramount for preventing cyber attackers from driving the cyber-physical power system (CPPS) to instability. In this paper, the ICT network traffic rules in CPPSs are explored and the frequency domain features of the ICT network traffic are extracted, basically for developing a robust learning algorithm that can learn the normal traffic pattern based on the ResNeSt convolutional neural network (CNN). Furthermore, to overcome the problem of insufficient abnormal traffic labeled samples, transfer learning approach is used. In the proposed data-driven-based method the deep learning model is trained by traffic frequency features, which makes our model robust against AGC's parameters uncertainties and modeling nonlinearities.

cs.LG↗
arXiv2022

Optimal Power Flow in Four-Wire Distribution Networks: Formulation and Benchmarking

In recent years, several applications have been proposed in the context of distribution networks. Many of these can be formulated as an optimal power flow problem, a mathematical optimization program which includes a model of the steady-state physics of the electricity network. If the network loading is balanced and the lines are transposed, the network model can be simplified to a single-phase equivalent model. However, these assumptions do not apply to low-voltage distribution networks, so the network model should model the effects of phase unbalance correctly. In many parts of the world, the low-voltage distribution network has four conductors, i.e. three phases and a neutral. This paper develops OPF formulations for such networks, including transformers, shunts and voltage-dependent loads, in two variable spaces, i.e. current-voltage and power-voltage, and compares them for robustness and scalability. A case study across 128 low-voltage networks also quantifies the modelling error introduced by Kron reductions and its impact on the solve time. This work highlights the advantages of formulations in current-voltage variables over power-voltage, for four-wire networks.

math.OC↗
arXiv2021

Model-predictive control and reinforcement learning in multi-energy system case studies

Model-predictive-control (MPC) offers an optimal control technique to establish and ensure that the total operation cost of multi-energy systems remains at a minimum while fulfilling all system constraints. However, this method presumes an adequate model of the underlying system dynamics, which is prone to modelling errors and is not necessarily adaptive. This has an associated initial and ongoing project-specific engineering cost. In this paper, we present an on- and off-policy multi-objective reinforcement learning (RL) approach, that does not assume a model a priori, benchmarking this against a linear MPC (LMPC - to reflect current practice, though non-linear MPC performs better) - both derived from the general optimal control problem, highlighting their differences and similarities. In a simple multi-energy system (MES) configuration case study, we show that a twin delayed deep deterministic policy gradient (TD3) RL agent offers potential to match and outperform the perfect foresight LMPC benchmark (101.5%). This while the realistic LMPC, i.e. imperfect predictions, only achieves 98%. While in a more complex MES system configuration, the RL agent's performance is generally lower (94.6%), yet still better than the realistic LMPC (88.9%). In both case studies, the RL agents outperformed the realistic LMPC after a training period of 2 years using quarterly interactions with the environment. We conclude that reinforcement learning is a viable optimal control technique for multi-energy systems given adequate constraint handling and pre-training, to avoid unsafe interactions and long training periods, as is proposed in fundamental future work.

eess.SY↗
arXiv2020

A Comprehensive Multi-Period Optimal Power Flow Framework for Smart LV Networks

This paper presents an extensive multi-period optimal power flow framework, with new modelling elements, for smart LV distribution systems that rely on residential flexibility for combating operational issues. A detailed performance assessment of different setups is performed, including: ZIP flexible loads (FLs), varying degrees of controllability of conventional residential devices, such as electric vehicles (EVs) or photovoltaics (PVs), by the distribution system operator (DSO) (adhering to customer-dependent restrictions) and full exploitation of the capabilities offered by state-of-the-art inverter technologies. A comprehensive model-dependent impact assessment is performed, including phase imbalances, neutral and ground wires and load dependencies. The de-congestion potential of common residential devices is highlighted, analyzing capabilities such as active power redistribution, reactive power support and phase balancing. Said potential is explored on setups where the DSO can make only partial adjustments on customer profiles, rather than (as is common) deciding on the full profiles. The extensive analysis can be used by DSOs and researchers alike to make informed decisions on the required levels of modelling detail, the connected devices and the degrees of controlability. The formulation is computationally efficient, scaling well to medium-size systems, and can serve as an excellent basis for building more tractable or more targeted approaches.

eess.SY↗
arXiv2020

Dual-decomposition-based peer-to-peer voltage control for distribution networks

The increasing penetration of distributed energy resources (DERs) creates new voltage issues in distribution networks. This study proposes an algorithm that mitigates these issues, by actively managing the active and reactive power of those DERs. The control problem is formulated as an optimisation problem. The study proposes solving the problem in a fully distributed manner and presents a methodology to convert a centralised constraint optimisation problem into a fully distributed constraint optimisation problem based on dual decomposition, linearised model of the distribution network and peer-to-peer communication protocol. A real low voltage residential semi-urban feeder from the region of Flanders, Belgium has been used as a case study. The simulation results show the ability of the proposed peer-to-peer control algorithm to control the voltage effectively within limits.

math.OC↗
arXiv2020

Optimal Combination of Frequency Control and Peak Shaving with Battery Storage Systems

Combining revenue streams by providing multiple services with battery storage systems increases profitability and enhances the investment case. In this work, we present a novel optimisation and control framework that enables a storage system to optimally combine the provision of primary frequency control services with peak shaving of a consumption profile. We adopt a dynamic programming framework to connect the daily bidding in frequency control markets with the longer term peak shaving objective: reducing the maximum consumption peak over an entire billing period. The framework also allows to aggregate frequency control capacity of multiple batteries installed at different sites, creating synergies when the consumption profile peaks occur on different times. Using a case study of two batteries at two industrial sites, we show that the presented approach increases net profit of the batteries significantly compared to using the batteries for only peak shaving or frequency control.

math.OC↗
arXiv2019

Techno-Economic Analysis and Optimal Control of Battery Storage for Frequency Control Services, Applied to the German Market

Optimal investment in battery energy storage systems, taking into account degradation, sizing and control, is crucial for the deployment of battery storage, of which providing frequency control is one of the major applications. In this paper, we present a holistic, data-driven framework to determine the optimal investment, size and controller of a battery storage system providing frequency control. We optimised the controller towards minimum degradation and electricity costs over its lifetime, while ensuring the delivery of frequency control services compliant with regulatory requirements. We adopted a detailed battery model, considering the dynamics and degradation when exposed to actual frequency data. Further, we used a stochastic optimisation objective while constraining the probability on unavailability to deliver the frequency control service. Through a thorough analysis, we were able to decrease the amount of data needed and thereby decrease the execution time while keeping the approximation error within limits. Using the proposed framework, we performed a techno-economic analysis of a battery providing 1 MW capacity in the German primary frequency control market. Results showed that a battery rated at 1.6 MW, 1.6 MWh has the highest net present value, yet this configuration is only profitable if costs are low enough or in case future frequency control prices do not decline too much. It transpires that calendar ageing drives battery degradation, whereas cycle ageing has less impact.

eess.SY↗
arXiv2019

Grid-Constrained Distributed Optimization for Frequency Control with Low-Voltage Flexibility

Providing frequency control services with flexible assets connected to the low-voltage distribution grid, amongst which residential battery storage or electrical hot water boilers, can lead to congestion problems and voltage issues in the distribution grid. In order to mitigate these problems, a new regulation has been put in place in Belgium, imposing a specific constraint: in any circle with a radius of 100m, there can be at maximum 10 connection points providing frequency control at any time. This paper presents an impact analysis and a coordination strategy of a Flexibility Service Provider (FSP) that operates a pool of assets and is exposed to this new regulatory constraint. Results show that at 5% participation, only 90% of total control capacity can be used, with a large difference between neighbourhoods with different population densities. A distributed optimization framework to coordinate the assets arises naturally, in which the assets are able to keep their local cost functions private and only have to communicate with neighbouring assets that are geographically close, and with the FSP. Analysis of the proposed distributed optimization algorithm shows a clear trade-off between optimality gap, owing to the mixed-integer nature of the problem, and iterations to convergence.

eess.SY↗
arXiv2017

Peer-to-Peer Control of Microgrids

In this paper, the motivation to develop microgrids as an effective solution for the control of distribution networks with high level penetration of Distributed Energy Resources(DERs) is discussed. As many different control methods for microgrids can be found in literature, this paper proposes a classification from highly centralized to distributed peer-to-peer control architectures. A peer-to-peer control paradigm is proposed as a way to control the distribution network with a high penetration of distributed energy resources. Different control algorithms suited for the proposed peer-to-peer control strategy are discussed.

math.OC↗
arXiv2017

Combined Stochastic Optimization of Frequency Control and Self-Consumption with a Battery

Optimally combining frequency control with self-consumption can increase revenues from battery storage systems installed behind-the-meter. This work presents an optimized control strategy that allows a battery to be used simultaneously for self-consumption and primary frequency control. Therein, it addresses two stochastic problems: the delivery of primary frequency control with a battery and the use of the battery for self-consumption. We propose a linear recharging policy to regulate the state of charge of the battery while providing primary frequency control. Formulating this as a chance-constrained problem, we can ensure that the risk of battery constraint violations stays below a predefined probability. We use robust optimization as a safe approximation to the chance-constraints, which allows to make the risk of constraint violation arbitrarily low, while keeping the problem tractable and offering maximum reserve capacity. Simulations with real frequency measurements prove the effectiveness of the designed recharging strategy. We adopt a rule-based policy for self-consumption, which is optimized using stochastic programming. This policy allows to reserve more energy and power of the battery on moments when expected consumption or production is higher, while using other moments for recharging from primary frequency control. We show that optimally combining the two services increases value from batteries significantly.

math.OC↗