The concept of smart grid has been introduced as a new vision of the conventional power grid to figure out an efficient way of integrating green and renewable energy technologies. In this way, Internet-connected smart grid, also called energy Internet, is also emerging as an innovative approach to ensure the energy from anywhere at any time. The ultimate goal of these developments is to build a sustainable society. However, integrating and coordinating a large number of growing connections can be a challenging issue for the traditional centralized grid system. Consequently, the smart grid is undergoing a transformation to the decentralized topology from its centralized form. On the other hand, blockchain has some excellent features which make it a promising application for the smart grid paradigm. In this article, we aim to provide a comprehensive survey on the application of blockchain in smart grid. As such, we identify the significant security challenges of smart grid scenarios that can be addressed by blockchain. Then, we present a number of blockchain-based recent research works presented in different literature addressing security issues in the area of smart grid. We also summarize several related practical projects, trials, and products that have emerged recently. Finally, we discuss essential research challenges and future directions of applying blockchain to smart grid security issues.

Recent advances in artificial intelligence have driven increasing intelligent applications at the network edge, such as smart home, smart factory, and smart city. To deploy computationally intensive Deep Neural Networks (DNNs) on resource-constrained edge devices, traditional approaches have relied on either offloading workload to the remote cloud or optimizing computation at the end device locally. However, the cloud-assisted approaches suffer from the unreliable and delay-significant wide-area network, and the local computing approaches are limited by the constrained computing capability. Towards high-performance edge intelligence, the cooperative execution mechanism offers a new paradigm, which has attracted growing research interest recently. In this paper, we propose CoEdge, a distributed DNN computing system that orchestrates cooperative DNN inference over heterogeneous edge devices. CoEdge utilizes available computation and communication resources at the edge and dynamically partitions the DNN inference workload adaptive to devices' computing capabilities and network conditions. Experimental evaluations based on a realistic prototype show that CoEdge outperforms status-quo approaches in saving energy with close inference latency, achieving up to 25.5% ~ 66.9% energy reduction for four widely-adopted CNN models.

Wind ramps introduce significant uncertainty into wind power generation. Reliable system operation, however, requires accurate detection and forecast of wind ramps, especially at high penetration levels. In this paper, to deal with the wind ramp dynamics, a support vector machine (SVM)-enhanced Markov model is developed for short-term wind power forecast, based on one key observation from the measurement data that wind ramps often occur with specific patterns. Specifically, using the historical data of the wind turbine power outputs recorded at an actual wind farm, data analytics-based finite-state Markov models are first developed to model the normal fluctuations of wind generation, while taking into account the diurnal nonstationarity and the seasonality of wind generation. Next, the forecast by the SVM is integrated cohesively into the finite-state Markov models. Based on the SVM-enhanced Markov model, both short-term distributional forecasts and point forecasts are then derived. Numerical test results, using real wind generation data traces, demonstrate the significantly improved accuracy of the proposed forecast approach.

Smart meters, designed for information collection and system monitoring in smart grid, report fine-grained power consumption to utility providers. With these highly accurate profiles of energy usage, however, it is possible to identify consumers' specific activities or behavior patterns, thereby giving rise to serious privacy concerns. This paper addresses these concerns by designing a cost-effective and privacy-preserving energy management technique that uses a rechargeable battery. From a holistic perspective, a dynamic programming framework is designed for consumers to strike a tradeoff between smart meter data privacy and the cost of electricity. In general, a major challenge in solving dynamic programming problems lies in the need for the knowledge of future electricity consumption events. By exploring the underlying structure of the original problem, an equivalent problem is derived, which can be solved by using only the current observations. An online control algorithm is then developed to solve the equivalent problem based on the Lyapunov optimization technique. It is shown that without the knowledge of the statistics of the time-varying load requirements and the electricity price processes, the proposed online control algorithm, parametrized by a positive value V , is within O (1/{V) of the optimal solution to the original problem, where the maximum value of V is limited by the battery capacity. The efficacy of the proposed algorithm is demonstrated through extensive numerical analysis using real data.

This paper studies risk-aware day-ahead scheduling and real-time dispatch for electric vehicle (EV) charging, aiming to jointly optimize the EV charging cost and the risk of the load mismatch between the forecast and the actual EV loads, due to the random driving activities of EVs. It turns out that the consideration of the load mismatch risk in the objective function significantly complicates the risk-aware day-ahead scheduling problem (indeed it involves nonconvex optimization). A key step taken here is to utilize a hidden convexity structure to recast this problem as a two-stage stochastic linear program, and then solve it by using the L-shaped method. Since the computational complexity grows exponentially in the number of EVs, an estimation algorithm is developed based on importance sampling to mitigate the computational complexity. Further, a distributed risk-aware real-time dispatch algorithm is developed, in which the aggregator needs to compute only the shadow prices for each EV to optimize its own charging strategy in a distributed manner. It is shown, based on real data, that the proposed risk-aware day-ahead scheduling algorithm using importance sampling can significantly reduce the overall charging cost with a small number of samples.

Fog computing, as a new paradigm, has many characteristics that are different from cloud computing. Due to the resources being limited, fog nodes/MEC hosts are vulnerable to cyberattacks. Lightweight intrusion detection system (IDS) is a key technique to solve the problem. Because extreme learning machine (ELM) has the characteristics of fast training speed and good generalization ability, we present a new lightweight IDS called sample selected extreme learning machine (SS‐ELM). The reason why we propose “sample selected extreme learning machine” is that fog nodes/MEC hosts do not have the ability to store extremely large amounts of training data sets. Accordingly, they are stored, computed, and sampled by the cloud servers. Then, the selected sample is given to the fog nodes/MEC hosts for training. This design can bring down the training time and increase the detection accuracy. Experimental simulation verifies that SS‐ELM performs well in intrusion detection in terms of accuracy, training time, and the receiver operating characteristic (ROC) value.