Abstract
Photovoltaic (PV) power is considered as one of the most promising sustainable energy resources in recent years. However, the existing intermittency in the nature of solar energy is a significant problem for the optimization of smart grids. In this paper, to overcome PV generation uncertainty and provide an accurate spatio-temporal (ST) PV forecast, we propose a novel deep generative convolutional graph rough variational autoencoder (CGRVAE) that captures each PV site's probability distribution functions (PDFs) of future PV generation in a modeled weighted graph. Having the learned PDFs enables CGRVAE to accurately generate the future values of PV power time series. To train and evaluate our model, we used the measurements of a set of PV sites in California, US. The sites are modeled as a weighted graph where each node represents PV measurements at each site while edges reflect their correlations. Using graph spectral convolutions the proposed model extracts the most relevant information of the graph to estimate the future PV given the historical time series for each node in the modeled graph. Experimental results show the superiority of CGRVAE over state-of-the-art forecasting approaches in terms of the root mean square error (RMSE) and mean absolute error (MAE) metric.

Abstract
The combination of consumer owned Distributed Energy Resources, new Information and Communication Technologies (ICT), as well as changes to the national electricity regulations have created new opportunities for consumer engagement in the electricity sector. In this paper, this combination of technologies and regulations is examined in the Portuguese context. The new regulations dealing with self-consumption from prosumers are combined with smart contracts and distributed ledger technology to formulate an automated energy trading system for residential end-users in local energy markets. Results show that including prosumers in the local energy market brings significant benefits to all market participants. Additionally, results show that the newly created regulatory role of a Market Facilitator is beneficial to these type of local energy exchanges.

Abstract

Abstract
Geomagnetically induced currents (GICs) are referred to as the quasi-dc current flows in power networks, driven by complex space weather-related phenomena. Such currents are a potential threat to the power delivery capability of electrical grids. To mitigate the detrimental impacts of GICs on critical infrastructures, the GICs should be monitored in power systems. Being inherently dc from the power frequency point of view, the components of GICs are, however, challenging and costly to monitor in ac power grids. This article puts forward a novel methodology for the real-time estimation of GICs in power transformers. Such aim is attained by means of an extended Kalman filter (EKF)-based approach, mounted on the nonlinear state-space model of the transformer, whose parameters can be derived from standard tests. The proposed EKF-based algorithm employs the available measurements for the transformer differential protection. The proposed approach, relying on the differential current, can properly deal with the external sources of interference like harmonic excitation and loading. The EKF-based estimator presented is validated by simulation and experimental data. The results verify the ability of the proposed approach to robustly estimate the GIC level during various operating conditions.

Abstract
Due to recent advances in power electronic systems, direct current (DC) microgrid (MG) topology is considered as a promising solution to unite pollution-free renewable energy sources and DC loads. This paper investigates the issue of finite-time robust adaptive stability and tracking issue of a nonlinear direct current (DC) microgrid (MG) comprising a buck converter, linear resistive loads, and nonlinear constant power loads (CPLs). The developed approach is based on a sliding mode controller (SMC) and a nonlinear and nonsingular sliding surface. It is proved that the tracking error converges to zero in a finite-time in the presence of matched disturbance input and uncertainties. The novel controller manipulates the buck converter of the source side to regulate the DC bus voltage by counteracting the destabilizing effect of CPLs and disturbances. Further, the finite value of the convergence time is presented and the effects of the SMC parameter on the stability and transient performance are evaluated. Lastly, numerical simulations are conducted to illustrate the merits of the developed control approach in the viewpoints of fast reference tracking and robust stability.

Abstract
This paper proposes a cascading failure simulation (CFS) method and a hybrid machine learning method for vulnerability analysis of integrated power-gas systems (IPGSs). The CFS method is designed to study the propagating process of cascading failures between the two systems, generating data for machine learning with initial states randomly sampled. The proposed method considers generator and gas well ramping, transmission line and gas pipeline tripping, island issue handling and load shedding strategies. Then, a hybrid machine learning model with a combined random forest (RF) classification and regression algorithms is proposed to investigate the impact of random initial states on the vulnerability metrics of IPGSs. Extensive case studies are carried out on three test IPGSs to verify the proposed models and algorithms. Simulation results show that the proposed models and algorithms can achieve high accuracy for the vulnerability analysis of IPGSs.