Search for: All records

Machine Learning (ML) algorithms have shown quite promising applications in smart meter data analytics enabling intelligent energy management systems for the Advanced Metering Infrastructure (AMI). One of the major challenges in developing ML applications for the AMI is to preserve user privacy while allowing active end-users participation. This paper addresses this challenge and proposes Differential Privacy-enabled AMI with Federated Learning (DP-AMI-FL), framework for ML-based applications in the AMI. This framework provides two layers of privacy protection: first, it keeps the raw data of consumers hosting ML applications at edge devices (smart meters) with Federated Learning (FL), and second, it obfuscates the ML models using Differential Privacy (DP) to avoid privacy leakage threats on the models posed by various inference attacks. The framework is evaluated by analyzing its performance on a use case aimed to improve Short-Term Load Forecasting (STLF) for residential consumers having smart meters and home energy management systems. Extensive experiments demonstrate that the framework when used with Long Short-Term Memory (LSTM) recurrent neural network models, achieves high forecasting accuracy while preserving users data privacy. 

The current centralized model of the electricity market is not efficient in performing distributed energy transactions required for the transactive smart grid. One of the prominent solutions to this issue is to integrate blockchain technologies, which promise transparent, tamper-proof, and secure transaction systems specifically suitable for the decentralized and distributed energy markets. Blockchain has already been shown to successfully operate in a microgrid peer-to-peer (P2P) energy market. The prime determinant of different blockchain implementations is the consensus algorithm they use to reach consensus on which blocks/transactions to accept as valid in a distributed environment. Although different blockchain implementations have been proposed independently for P2P energy market in the microgrid, quantitative experimental analyses and comparison of the consensus algorithms that the different blockchains may use for energy markets, has not been studied. Identifying the right consensus algorithm to use is essential for scalability and operation of the energy market. To this end, we evaluate three popular consensus algorithms: (i) proof of work (PoW), (ii) proof of authority (PoA), and (iii) Istanbul Byzantine fault tolerance (IBFT), running them on a network of nodes set up using a network of docker nodes to form a microgrid energy market. Using a series of double auctions, we assess each algorithm's viability using different metrics, such as time to reach consensus and scalability. The results indicate that PoA is the most efficient and scalable consensus algorithm to hold double auctions in the smart grid. We also identified the minimum hardware specification necessary for devices such as smart meters, which may run these consensus algorithms 

The current centralized model of the electricity market is not efficient in performing distributed energy transactions required for the transactive smart grid. One of the prominent solutions to this issue is to integrate blockchain technologies, which promise transparent, tamper-proof, and secure transaction systems specifically suitable for the decentralized and distributed energy markets. Blockchain has already been shown to successfully operate in a microgrid peer-to-peer (P2P) energy market. The prime determinant of different blockchain implementations is the consensus algorithm they use to reach consensus on which blocks/transactions to accept as valid in a distributed environment. Although different blockchain implementations have been proposed independently for P2P energy market in the microgrid, quantitative experimental analyses and comparison of the consensus algorithms that the different blockchains may use for energy markets, has not been studied. Identifying the right consensus algorithm to use is essential for scalability and operation of the energy market. To this end, we evaluate three popular consensus algorithms: (i) proof of work (PoW), (ii) proof of authority (PoA), and (iii) Istanbul Byzantine fault tolerance (IBFT), running them on a network of nodes set up using a network of docker nodes to form a microgrid energy market. Using a series of double auctions, we assess each algorithm’s viability using different metrics, such as time to reach consensus and scalability. The results indicate that PoA is the most efficient and scalable consensus algorithm to hold double auctions in the smart grid. We also identified the minimum hardware specification necessary for devices such as smart meters, which may run these consensus algorithms. 

The applications for wide area monitoring, protection, and control systems (WAMPC) at the control center, help with providing resilient, efficient, and secure operation of the transmission system of the smart grid. The increased proliferation of phasor measurement units (PMUs) in this space has inspired many prudent applications to assist in the process of decision making in the control centers. Machine learning (ML) based decision support systems have become viable with the availability of abundant high-resolution wide area operational PMU data. We propose a deep neural network (DNN) based supervisory protection and event diagnosis system and demonstrate that it works with very high degree of confidence. The system introduces a supervisory layer that processes the data streams collected from PMUs and detects disturbances in the power systems that may have gone unnoticed by the local monitoring and protection system. Then, we investigate compromise of the insights of this ML based supervisory control by crafting adversaries that corrupt the PMU data via minimal coordinated manipulation and identification of the spatio-temporal regions in the multidimensional PMU data in a way that the DNN classifier makes wrong event predictions. This dataset contains images that represent PMU data described in the reference paper. Each image has a dimension of [300X20X3] comprising of 300 time points, 10 voltage and 10 frequency measurements, and 3 fundamental color intensities. Each of the image represents the instance of a disturbance. We consider a disturbance pattern length of 5s, with 0.5 s before the trigger and 4.5 s after the trigger. Voltage and frequency data streams from 10 PMUs at a sampling rate of 60 frames per second, were aggregated to form these pseudo color images. The data-set consisted of three sub-folders: 1. 344 instances of faults located in the sub-folder “DB_FLT” 2. 140 instances of loss of generation located in the sub-folder “DB_GNL” 3. 21 instances of synchronous motor switching events located in the sub-folder “DB_SMS”.