Extreme events, such as earthquakes and hurricanes, can simultaneously damage power systems and communication networks, making it difficult to restore either system independently due to their increasing cyber-physical interdependence. This article investigates the use of drones to establish an emergency communication network that supports postcontingency distribution network operation. An end-to-end learning-based, cyber-physical stability-constrained framework is proposed for emergency cyber network formation. In this framework, drones are strategically deployed across the distribution network to enable postcontingency communication and coordinate distributed energy resources. The framework incorporates multidrone connected communication optimization, wireless resource allocation, cyber-physical network modeling, and stability criteria extraction. These processes introduce significant computational complexity, including NP-hard challenges and nonlinearities, which are difficult to solve efficiently using traditional optimization methods. To address these limitations, this article develops an end-to-end deep reinforcement learning algorithm that determines optimal multidrone trajectories from the depot, providing a feasible multidrone deployment solution that ensures a stable emergency energy supply.

Successful machine learning involves a complete pipeline of data, model, and downstream applications. Instead of treating them separately, there has been a prominent increase of attention within the constrained optimization (CO) and machine learning (ML) communities towards combining prediction and optimization models. The so-called end-to-end (E2E) learning captures the task-based objective for which they will be used for decision making. Although a large variety of E2E algorithms have been presented, it has not been fully investigated how to systematically address uncertainties involved in such models. Most of the existing work considers the uncertainties of ML in the input space and improves robustness through adversarial training. We extend this idea to E2E learning and prove that there is a robustness certification procedure by solving augmented integer programming. Furthermore, we show that neglecting the uncertainty of COs during training causes a new trigger for generalization errors. To include all these components, we propose a unified framework that covers the uncertainties emerging in both the input feature space of the ML models and the COs. The framework is described as a robust optimization problem and is practically solved via end-to-end adversarial training (E2E-AT). Finally, the performance of E2E-AT is evaluated by a real-world end-to-end power system operation problem, including load forecasting and sequential scheduling tasks.

We investigate how environmental data collected from a low-cost, distributed Internet-of-Things platform can be used to identify factors affecting air quality in highly dense urban environments. In this work, we focus on the the case of Bangkok. We use data collected from the SEA-HAZEMON platform to analyze the incidence of several parameters on PMx level and assess its potential to forecast CO2 emission.

Second-order optimization has been shown to accelerate the training of deep neural networks in many applications, often yielding faster progress per iteration on the training loss compared to first-order optimizers. However, the generalization properties of second-order methods are still being debated. Theoretical investigations have proved difficult to carry out outside the tractable settings of heavily simplified model classes - thus, the relevance of existing theories to practical deep learning applications remains unclear. Similarly, empirical studies in large-scale models and real datasets are significantly confounded by the necessity to approximate second-order updates in practice. It is often unclear whether the observed generalization behaviour arises specifically from the second-order nature of the parameter updates, or instead reflects the specific structured (e.g. Kronecker) approximations used or any damping-based interpolation towards first-order updates. Here, we show for the first time that exact Gauss-Newton (GN) updates take on a tractable form in a class of deep reversible architectures that are sufficiently expressive to be meaningfully applied to common benchmark datasets. We exploit this novel setting to study the training and generalization properties of the GN optimizer. We find that exact GN generalizes poorly. In the mini-batch training setting, this manifests as rapidly saturating progress even on the training loss, with parameter updates found to overfit each mini-batch without producing the features that would support generalization to other mini-batches. In contrast to previous work, we show that our experiments run in the feature learning regime, in which the neural tangent kernel (NTK) changes during the course of training. However, changes in the NTK are not associated with any significant change in neural representations, explaining the lack of generalization.

Digitalization is one of the key drivers for energy system transformation. The advances in communication technologies and measurement devices render available a large amount of operational data and enable the centralization of such data storage and processing. The greater access to data opens up new opportunities for a more efficient and decentralized management of the energy system. At the distribution level of the energy system, local electricity markets (LEMs) provide new degrees of flexibility by trading and balancing the energy locally and offering ancillary services to the wider transmission and distribution system operators. Maximizing the grid impact from this flexibility calls for novel data analytics and artificial intelligence techniques to enhance the system’s security and reduce the energy costs of local prosumers. At the same time, however, relying on data-based approaches increases the risk of cyberattacks, and robust countermeasures are, therefore, needed as an integral aspect of digitalization efforts. This article discusses the key role of centralized data analytics to fully benefit from the advantages of LEMs in terms of system’s security enhancement and energy costs’ reduction. Data-driven paradigms are investigated that allow for flexibility from decentralized markets, mitigate the physical security risks, and devise defensive strategies shielding the system from cyber threats.

Recently, moving target defence (MTD) has been proposed to thwart false data injection (FDI) attacks in power system state estimation by proactively triggering the distributed flexible AC transmission system (D-FACTS) devices. One of the key challenges for MTD in power grid is to design its real-time implementation with performance guarantees against unknown attacks. Converting from the noiseless assumptions in the literature, this paper investigates the MTD design problem in a noisy environment and proposes, for the first time, the concept of robust MTD to guarantee the worst-case detection rate against all unknown attacks. We theoretically prove that, for any given MTD strategy, the minimal principal angle between the Jacobian subspaces corresponds to the worst-case performance against all potential attacks. Based on this finding, robust MTD algorithms are formulated for the systems with both complete and incomplete configurations. Extensive simulations using standard IEEE benchmark systems demonstrate the improved average and worst-case performances of the proposed robust MTD against state-of-the-art algorithms.

This paper presents a problem in power networks that creates an exciting and yet challenging real-world scenario for application of multi-agent reinforcement learning (MARL). The emerging trend of decarbonisation is placing excessive stress on power distribution networks. Active voltage control is seen as a promising solution to relieve power congestion and improve voltage quality without extra hardware investment, taking advantage of the controllable apparatuses in the network, such as roof-top photovoltaics (PVs) and static var compensators (SVCs). These controllable apparatuses appear in a vast number and are distributed in a wide geographic area, making MARL a natural candidate. This paper formulates the active voltage control problem in the framework of Dec-POMDP and establishes an open-source environment. It aims to bridge the gap between the power community and the MARL community and be a drive force towards real-world applications of MARL algorithms. Finally, we analyse the special characteristics of the active voltage control problems that cause challenges (e.g. interpretability) for state-of-the-art MARL approaches, and summarise the potential directions.