Grid Edge Research Articles

Uncertainty Matters: Bayesian Probabilistic Forecasting for Residential Smart Meter Prediction, Segmentation, and Behavioral Measurement and Verification

As new grid edge technologies emerge—such as rooftop solar panels, battery storage, and controllable water heaters—quantifying the uncertainties of building load forecasts is becoming more critical. The recent adoption of smart meter infrastructures provided new granular data streams, largely unavailable just ten years ago, that can be utilized to better forecast building-level demand. This paper uses Bayesian Structural Time Series for probabilistic load forecasting at the residential building level to capture uncertainties in forecasting. We use sub-hourly electrical submeter data from 120 residential apartments in Singapore that were part of a behavioral intervention study. The proposed model addresses several fundamental limitations through its flexibility to handle univariate and multivariate scenarios, perform feature selection, and include either static or dynamic effects, as well as its inherent applicability for measurement and verification. We highlight the benefits of this process in three main application areas: (1) Probabilistic Load Forecasting for Apartment-Level Hourly Loads; (2) Submeter Load Forecasting and Segmentation; (3) Measurement and Verification for Behavioral Demand Response. Results show the model achieves a similar performance to ARIMA, another popular time series model, when predicting individual apartment loads, and superior performance when predicting aggregate loads. Furthermore, we show that the model robustly captures uncertainties in the forecasts while providing interpretable results, indicating the importance of, for example, temperature data in its predictions. Finally, our estimates for a behavioral demand response program indicate that it achieved energy savings; however, the confidence interval provided by the probabilistic model is wide. Overall, this probabilistic forecasting model accurately measures uncertainties in forecasts and provides interpretable results that can support building managers and policymakers with the goal of reducing energy use.

Security Risk Analysis of Grid Edge Computing

The emergence of edge computing technology brings new development direction to the construction of power grid industry. At present, for the edge computing technology of China is still in the exploratory stage, the combination of power grid and edge computing may face new security risks. This paper analysed the security risks that may occur in grid edge computing under the circumstances of identification system, communication technology, equipment protection and resource constraints, and put forward the relevant security risk analysis model and defence, aiming at solving the security problems of grid edge computing and promoting the better development of power grid.

An overview of grid-edge control with the digital transformation

Distribution networks are evolving to become more responsive with increasing integration of distributed energy resources (DERs) and digital transformation at the grid edges. This evolution imposes many challenges to the operation of the network, which then calls for new control and operation paradigms. Among others, a so-called grid-edge control is emerging to harmonise the coexistence of the grid control system and DER’s autonomous control. This paper provides a comprehensive overview of the grid-edge control with various control architectures, layers, and strategies. The challenges and opportunities for such an approach at the grid edge with the integration of DERs and digital transformation are summarised. The potential solutions to support the network operation by using the inherent controllability of DER and the availability of the digital transformation at the grid edges are discussed.

Multi-keyword Cipher-text Retrieval Method for Smart Grid Edge Computing*

With the continuous deepening of China's power grid intelligent upgrading and the rise of edge computing technology, power grid terminals can carry out data storage and intelligent calculation at the edge of the power grid. However, the generated data by users may cause the leakage of user privacy information. The power grid edge system can retrieve the user's electricity data on demand, so as to improve the level of privacy protection and reduce the demand of transmission and computing resources through the localized storage of data. Therefore, how to carry out privacy protection and on-demand retrieval is a key problem to realize the intelligent power grid edge. In this paper, a multi-keyword cipher-text retrieval scheme (MCR- SGEC) for smart grid edge computing is proposed, which can achieve accurate matching of index and trap door, and has low spatial complexity. At the same time, on this basis, a multi-keyword fuzzy cipher-text retrieval method for smart grid edge computing (MFSE-SGEC) is proposed, which allows users to have spelling errors, which makes the user experience better.

WITHDRAWN: IMPROVED RESOURCE MANAGEMENT AND UTILIZATION BASED ON A FOG-CLOUD COMPUTING SYSTEM WITH IOT INCORPORATED WITH CLASSIFIER SYSTEMS

With the rapid rise of computing devices-linked networks, the Internet of Things (IoT) has increased proportionally and speedily. After that, the size of the data generated and significantly increased computational effort. A solution for a very large workload is cloud computing significantly slowed and continues to be a problem in a distributed computer network. Addressing the workload at the network edge can enchase time feedback, and delaying the data load management task imposes performance constraints from the power source to the edge of the network, which does not support only limited battery power. Therefore, the workload must be distributed equally to the clouds and across the network. This document presents the XCS Classification System (LCS), by XCS and BCM-XCS, to ensure energy consumption at the end of the project which decreases the workloads of the processing time. Our test outcome showed the dominance of BCM-XCS XCS according to standard methods. A distributed method of allocating responsibility for reducing the processing time between node cloud and fog. In addition to the significant benefits of handling fluctuating turnaround times, the proposed procedures can simultaneously decrease 42% response times utilizing an average consumption of energy in the grid edge. With the propounded method, rechargeable batteries are used on the edge of the grid also can load approximately 18% more than the best method of the prior art.

Short‐term load demand forecasting through rich features based on recurrent neural networks

With the emerging penetration of renewables and dynamic loads, the understanding of grid edge loading conditions becomes increasingly substantial. Load modelling researches commonly consist of explicitly expressed load models and non-explicitly expressed techniques, of which artificial intelligence approaches turn out to be the major path. This paper reveals the artificial intelligence-based load modelling technique to enhance the knowledge of current and future load information considering geographical and weather dependencies. This paper presents a recurrent neural network based sequence to sequence (Seq2Seq) model to forecast the short-term power loads. Also, a feature attention mechanism, which is along channel and time directions, is developed to improve the efficiency of feature learning. The experiments over three publicly available datasets demonstrate the accuracy and effectiveness of the proposed model.

Multitimescale Three-Tiered Voltage Control Framework for Dispersed Smart Inverters at the Grid Edge

Voltage control schemes are being employed to guarantee that the voltage across the grid stays within the predefined boundaries. Growing number of smart PV-based inverters (SPVIs) necessitates cutting-edge control schemes to ensure addressing voltage fluctuations suitably aligned with the existing grid codes. Designing such control schemes for grid clusters with high penetration of SPVIs is complex and depends on various agencies, such as configuration of the clusters and physical limitations of the SPVIs, themselves. This article proposes a multitimescale three-tiered voltage control framework for dispersed SPVIs at the grid edge. The uppermost tier called the long-term supervisory tier oversees the entire grid cluster to make sure that total generation and consumption on the cluster are balanced while power delivery losses are minimized and voltage across the grid is within the permissible boundaries. In short term, the middle tier of the proposed framework called short-term distributed tier ensures cooperative operation of the SPVIs for establishing short-term voltage stability across the cluster in a subsecond timeframe. The undermost tier, called the model-predictive-based local tier, fulfills the operational commands dictated by the upper layers with fast dynamics. By implementing the proposed control framework, the SPVIs will participate in voltage regulation across the grid in occurrence of rapid voltage fluctuations while keeping the entire distribution grid at the optimal operating condition for a long term. The proposed multitimescale three-tiered control framework is verified by simulation.

A framework for designing and evaluating realistic blockchain-based local energy markets

A growing customer base for solar-plus-storage at the grid edge has resulted in stronger interest at the regulatory level towards energy markets at the distribution level. Local energy markets (LEMs) running on blockchains are being studied as a possible direction, but the relevant literature treats the blockchain component as a black box. We make the case that this approach is flawed because the choices in this layer affect the market’s performance significantly. We explicitly identify the design space that the blockchain layer introduces, and analyze how the design choices made therein affect the performance, governance, and degree of decentralization of these markets. As an exercise, we consider three distinct configurations for a next-generation LEM, and compare their performance on both the blockchain and the market layer via a case study. We demonstrate that simple changes in the data model can decrease the market efficiency by up to 90%. We also show that changes in the way bids get encrypted may result in economic improvements, but they do so at the risk of subverting the proper operation and resilience of the market. The simulations for our case study are conducted via a framework that we developed and open-sourced as part of this work.

High-Frequency Power Electronics at the Grid Edge: A Bottom-Up Approach Toward the Smart Grid

Currently, Solar Energy Systems are supplying 2% of U.S. electricity generation and their use is rapidly growing. Data centers make up 2% of the U.S. total energy consumption. By 2030, 40% of the overall worldwide energy consumption is expected to be in the form of electricity. The future development of transportation electrification and smart grids requires grid-scale energy storage. These emerging and high-impact applications locate in the distribution grid. They represent exciting opportunities in developing fundamentally new principles for high-performance power electronics that are smaller, smarter, more efficient, and capable of performing new functions (Figure 1).

Time and voltage domain load models for appliance-level grid edge simulation and control

Developments in simulation and distributed control of distributed energy resources require increasingly granular characterization and modeling of load behavior. Previous work has demonstrated the viability of voltage-based power control of residential appliances to enable load flexibility without service interruption. However, conventional static and dynamic load models fail to capture the coupled voltage and state dynamics relevant for this type of control.We develop a new class of dynamic load models for residential appliances. The input-output dynamics are learned by varying input voltage, which is enabled by custom hardware capable of controlling single-phase AC voltage and collecting high-resolution measurements. We estimate model parameters using nonlinear least squares regression and particle swarm optimization. The RMSE of power predictions is significantly reduced for loads with coupled time and voltage dynamics relative to traditional models. Using these models for voltage-based power control can help improve the ability of DERs to provide grid services.

Sparsity Based Approaches for Distribution Grid State Estimation - A Comparative Study

The power distribution grid is typically unobservable due to a lack of measurements. While deploying more sensors can alleviate this issue, it also presents new challenges related to data aggregation and the underlying communication infrastructure. Therefore, developing state estimation methods that enhance situational awareness at the grid edge with compressed measurements is critical. For this purpose, a suite of sparsity-based approaches that exploit the correlation among states/measurements in spatial as well as temporal domains have been proposed recently. This article presents a systematic comparison and evaluation of these approaches. Specifically, the performance and complexity of spatial methods (1-D compressive sensing and matrix completion) and spatio-temporal methods (2-D compressive sensing and tensor completion) are compared using the IEEE 37 and IEEE 123 bus test systems. Additionally, new robust formulations of these sparsity-based methods are derived and shown to be robust to bad data and network parameter uncertainties. Among the sparsity-based approaches, compressive sensing methods tend to outperform matrix completion and tensor completion methods in terms of error performance.

Using energy storage to stabilise grids and increase revenues

Electric networks are facing unprecedented challenges. Increasing security concerns, more frequent natural disasters, and growing shares of renewables are adding stress to these networks. There are new solutions based on energy storage, smart automation, and controls that stabilise grids, which can reduce the curtailment of local renewable generators and provide continuous power to network customers. An innovative Hitachi ABB Power Grids grid edge solution in South Australia's lower Yorke Peninsula demonstrates how e-mesh control, management, automation, and the PowerStore Battery Energy Storage System (BESS) improve reliability and resilience in the network. The project unlocks new revenue streams from both the local network and the national energy market. Since beginning operation in late 2018, the 30 MW ESCRI (Energy Storage for Commercial Renewable Integration) Dalrymple BESS drastically reduced outages from 8 h down to 30 min within its first 6 months of operation. The solution allows >100 km of network feeders to island, limiting curtailment from the 91 MW Wattle Point Wind Farm and >2 MW of distributed rooftop solar panels. While islanded, local network customers continue to receive secure, reliable power entirely from renewable sources. The demonstrated functionalities – in particular, fast response and synthetic inertia – unlock new revenue opportunities for BESS deployment globally.

Validity Evaluation of Transformer DGA Online Monitoring Data in Grid Edge Systems

Dissolved gas analysis (DGA) is an important method of predicting transformer faults, and the accuracy of DGA measurements is of great significance to the evaluation of the transformer state in grid edge systems. In actual situations, it is difficult to regularly calibrate online DGA monitoring devices in a uniform way. Therefore, this paper proposes a method based on B-EMD and DBN to evaluate the validity of online DGA monitoring data and optimize the corresponding calibration plan. An analysis of actual DGA signals shows that the method proposed in this paper can effectively diagnose faults and improve the reliability of DGA online monitoring devices.

Design and Simulation of Multi-Function Virtual Grid Edge Intelligent Electronic Device with Standardized Semantics Based on IEC 61850 Standard

The advancement of communication technologies has facilitated the evolution of traditional electric power system into a vast entity called smart grid. Various domains such as substation automation, distributed energy resources, advanced metering infrastructures, energy storage systems and many more serve as parts and parcels of smart grid. Fundamentally, these domains communicate and exchange information within the domain as well as other domains to execute their intended functions. One example is the grid edge domain in smart grid, where devices at customer premise interact with utility grid as well as electricity market for applications such as demand response and energy trading. Generally, devices in grid edge domain exists independently to perform its specific functions. The information model and communication standards in these devices vary depending on the implementation by its manufacturers. This pose challenges for integration works which may require these devices to exchange information with each other. Therefore, this paper aims to design and proposeamulti-function virtual intelligent electronic device (IED) which capable of performing various applications at grid edge domain. The virtual IED utilizes standardized data model and semantics based on IEC 61850 standard. Information exchange based on IEC 61850 is then simulated between the virtual IED and a Client Simulator. The results indicate successful communication and information exchange based on IEC 61850 standard hence demonstrates the potential applications of the standard for grid edge domain.

Artificial Neural Networks for Volt/VAR Control of DER Inverters at the Grid Edge

With increased grid penetration of distributed energy resources (DERs), volt/VAR control (VVC) is becoming increasingly complex and challenging in distribution systems. Conventional VVC approaches in distribution networks reside typically on the primary feeder. However, these approaches are ineffective for handling the voltage variability caused by DERs and load dynamics on the secondary side, termed as the grid edge. With more and more deployments of DERs at the grid edge, advanced VVC technologies through controlling DER inverters are perceived to be the key enabling technologies for distributed voltage control to meet the future smart grid requirements. Nevertheless, a great challenge is that DER inverters would fight each other when operated autonomously in participating grid voltage control using the conventional control methods. This paper proposes a novel VVC strategy based on an artificial neural network (ANN) control. The proposed approach is able to properly handle DER inverter constraints in achieving VVC objectives at the grid edge and overcomes the challenges of conventional DER inverter control techniques. The behaviors of the proposed method are evaluated via simulation and hardware experiments, which demonstrate great advantages of the ANN-based VVC for controlling DER inverters at the grid edge.

Grid Architecture: A Core Discipline for Grid Modernization

By the beginning of the present decade, it had become clear that efforts to improve U.S. electric power systems as well as power systems around the world were being stymied by a problem that was more or less recognized but not often articulated: the overwhelming complexity of what is generally and loosely referred to as the grid had become so great that understanding the implications of changes to the grid, or even determining what changes should be made, was significantly impeding grid modernization. Less obvious, but just as significant, changes to the grid were already occurring that increasingly diverged from the basic principles and assumptions under which the 20th century grid was developed. These changes could lead to severe consequences for grid reliability and functionality if they were not addressed. New developments are largely driven by evolving consumer expectations, the emergence of new technologies, and the change from central economies of scale to network economies. The latter is driven by increased penetration of distributed energy resources (DERs) connected at the distribution level (the grid edge) and by ubiquitous communication connectivity. Additional drivers include deficiencies in resilience and increasing threat of cyberattack.

Pairing based anonymous and secure key agreement protocol for smart grid edge computing infrastructure

The most vital concern in the realization of the Internet of Things (IoT) is to encounter the disparate communication systems and technologies. Interoperability solutions such as standards can help us to integrate plenty of diverse devices and their applications in an interoperable framework. Since Smart Grid is a non-trivial prevalent application of Edge computing under the umbrella of IoT. The Smart Grid furnishes communication through Internet Protocol to enable interoperability. However, IP-Based communication makes it vulnerable to serious security threats. Therefore, the secure information sharing among diverse communicating agents in the smart grid environments has become a vital concern. Specifically, to enable secure communication between the smart meter and utility, key management prior to authentication is the most critical task to do. Nowadays, several mechanisms have been introduced to establish secure communication within the emerging smart grid environment. Although, these protocols do not support smart meter anonymity and fail to offer reasonable security. In this paper, we use the identity-based signature to present an anonymous key agreement protocol for the Smart Grid infrastructure. This protocol enables the smart meters to get connected with utility control anonymously to avail the services provided by them. The smart meters realize this objective with the private key in the absence of trusted authority. The trusted authority is involved only during the registration phase The proposed protocol is verified and validated through random oracle model and automated tool ProVerif. Moreover, performance analysis is also observed to consolidate the reliability and efficiency of the proposed protocol.

Grid Modernization, DER Integration & Utility Business Models - Trends & Challenges

The modernization of the distribution grid addresses the need to build an intelligent, reliable, resilient and secure grid that enables consumers to actively participate and benefit from greater and more diverse market opportunities and services. Building this intelligent grid requires not only a robust foundational grid but also pervasive monitoring, protection, automation and control capabilities (including at grid edge), which are beyond the functionalities traditionally available in distribution grids. Examples of grid modernization technologies include Distribution Automation (DA) [1], Advanced Metering Infrastructure (AMI), Distribution and Outage Management Systems (DMS /OMS), and grid analytics. Successful grid modernization comprises not only the deployment of new infrastructure and technologies, but also the implementation of updated distribution planning, operations, and engineering processes, and the approval of new regulations, policies and business models. In this regard, the capabilities resulting from grid modernization initiatives are also vital for the efficient, seamless, and successful adoption of Distributed Energy Resources (DER), as shown in Figure 1.

Computational fluid dynamics simulation of indoor climate in low energy buildings: Computational set up

In this paper CFD was used for simulation of the indoor climate in a part of a low energy building. The focus of the work was on investigating the computational set up, such as grid size and boundary conditions in order to solve the indoor climate problems in an accurate way. Future work is to model a complete building, with reasonable calculation time and accuracy. A limited number of grid elements and knowledge of boundary settings are therefore essential. An accurate grid edge size of around 0.1 m was enough to predict the climate according to a grid independency study. Different turbulence models were compared with only small differences in the indoor air velocities and temperatures. The models show that radiation between building surfaces has a large impact on the temperature field inside the building, with the largest differences at the floor level. Simplifying the simulations by modelling the radiator as a surface in the outer wall of the room is appropriate for the calculations. The overall indoor climate is finally compared between three different cases for the outdoor air temperature. The results show a good indoor climate for a low energy building all around the year.

A Fast Superpixel Segmentation Algorithm for PolSAR Images Based on Edge Refinement and Revised Wishart Distance.

The superpixel segmentation algorithm, as a preprocessing technique, should show good performance in fast segmentation speed, accurate boundary adherence and homogeneous regularity. A fast superpixel segmentation algorithm by iterative edge refinement (IER) works well on optical images. However, it may generate poor superpixels for Polarimetric synthetic aperture radar (PolSAR) images due to the influence of strong speckle noise and many small-sized or slim regions. To solve these problems, we utilized a fast revised Wishart distance instead of Euclidean distance in the local relabeling of unstable pixels, and initialized unstable pixels as all the pixels substituted for the initial grid edge pixels in the initialization step. Then, postprocessing with the dissimilarity measure is employed to remove the generated small isolated regions as well as to preserve strong point targets. Finally, the superiority of the proposed algorithm is validated with extensive experiments on four simulated and two real-world PolSAR images from Experimental Synthetic Aperture Radar (ESAR) and Airborne Synthetic Aperture Radar (AirSAR) data sets, which demonstrate that the proposed method shows better performance with respect to several commonly used evaluation measures, even with about nine times higher computational efficiency, as well as fine boundary adherence and strong point targets preservation, compared with three state-of-the-art methods.