Article
Engineering, Electrical & Electronic
Zhao Wu, Chao Wang, Huaiqing Zhang, Wenxiong Peng, Weihua Liu
Summary: This paper introduces the application of Hidden Markov Model (HMM) and its variations in Non-Intrusive Load Monitoring (NILM). By proposing a time-efficient Factorial Hidden Semi-Markov Model (TE-FHSMM), the paper achieves a reduction in time consumption while maintaining performance when dealing with datasets with different numbers of appliances. Additionally, experiments show that TE-FHSMM outperforms six state-of-the-art algorithms in terms of Accuracy and F1 score in real-world scenarios and publicly available datasets.
ELECTRIC POWER SYSTEMS RESEARCH
(2021)
Article
Engineering, Electrical & Electronic
Dong Hua, Fanqi Huang, Longjun Wang, Wutao Chen
Summary: This paper studies the event-based NILM approach using a mixed linear integer programming (MILP) model, which can disaggregate power consumption of multiple appliances switched simultaneously. By utilizing non-dominated sorting genetic algorithm II and cumulative sum (CUSUM) technique, a tradeoff between precision and recall can be achieved.
ELECTRIC POWER SYSTEMS RESEARCH
(2021)
Article
Engineering, Electrical & Electronic
Partik Kumar, Abhijit R. Abhyankar
Summary: In this paper, a Modified Factorial Hidden Markov Model (MFHMM) based Non-Intrusive Load Monitoring (NILM) framework is proposed, which enables the assessment of energy consumption behavior through modeling appliance activities and segmenting operating states.
IEEE TRANSACTIONS ON SMART GRID
(2023)
Article
Energy & Fuels
Soumyajit Ghosh, Debashis Chatterjee
Summary: Smart meter technology is crucial in the context of smart grid connected residential load system, with non-intrusive load monitoring being a well-known method to assess power consumption and operating behavior of individual loads. The issue of identifying harmonic polluting loads has arisen due to modern household appliances injecting unwanted harmonics into the system. An improved technique using input aggregated voltage-current data and an Artificial Bee Colony algorithm has been proposed for load monitoring without heavy training mechanisms, showing promising results in comparison to state-of-the-art techniques.
SUSTAINABLE ENERGY GRIDS & NETWORKS
(2021)
Review
Engineering, Electrical & Electronic
Pascal A. Schirmer, Iosif Mporas
Summary: The rapid development of technology in the electrical energy sector has led to increased electric power needs. This has resulted in the adoption of smart-meters and smart-grids, as well as the rise of Load Monitoring (LM) using Non-Intrusive Load Monitoring (NILM) for appliance-specific energy monitoring. The article provides a review of NILM approaches, groups previously published results, and includes a software implementation of the described NILM approaches.
IEEE TRANSACTIONS ON SMART GRID
(2023)
Article
Energy & Fuels
Krzysztof Dowalla, Piotr Bilski, Robert Lukaszewski, Augustyn Wojcik, Ryszard Kowalik
Summary: The paper presents a novel non-intrusive method for identifying appliances by analyzing changes in the common supply current signal. The method achieves high accuracy by processing the signal in the time domain, preserving the information in the original signal. Experimental results demonstrate the effectiveness of the approach in real-world scenarios.
Article
Energy & Fuels
Xiaochao Guo, Chao Wang, Tao Wu, Ruiheng Li, Houyi Zhu, Huaiqing Zhang
Summary: Results from Non-Intrusive Load Monitoring can be used for energy decomposition and load identification, facilitating effective energy consumption management. However, existing studies often focus on fixed settings, which limits the practical application in real-world scenarios. In this study, we propose a novel appliance detection method that analyzes pattern variations of aggregated power sequences and can determine new appliances without prior information, establishing a basis for load monitoring in scenarios with dynamic changes in load topology. Experimental results show significant improvement in F1 metric compared to event-based approaches.
Article
Construction & Building Technology
Zhao Wu, Chao Wang, Wenxiong Peng, Weihua Liu, Huaiqing Zhang
Summary: This paper introduces an Adaptive Density Peak Clustering-Factorial Hidden Markov Model (ADPC-FHMM) that can automatically determine the working states of appliances with reduced dependency on prior information. Case studies show that the proposed model outperforms its counterparts on metrics such as Accuracy, F-measure, and MAE.
ENERGY AND BUILDINGS
(2021)
Article
Engineering, Electrical & Electronic
Sarantis Kotsilitis, Emmanouil Kalligeros, Eftychia C. Marcoulaki, Irene G. Karybali
Summary: Non-intrusive load monitoring (NILM) is used to determine individual-appliance energy consumption by decomposing aggregated electricity measurements. To address the challenges of high-frequency sampling rates and computational resources, a lightweight event-detection algorithm is proposed for on-site implementation. The algorithm utilizes simple features, multiple criteria, and slope-coefficient inspection to detect events accurately and efficiently.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Construction & Building Technology
Alan Meier, Dan Cautley
Summary: Non-intrusive load monitoring system (NILM) in commercial buildings faces challenges in load disaggregation due to the number and complexity of loads, difficulty in interpreting small changes in power consumption, and inability to identify continuously operating loads. Obtaining data sets for evaluation of NILM technologies in actual buildings is hindered by disruptions to occupants, misidentification errors, measurement errors, and expense. Enhancements to basic NILM approaches include tagging key devices, hybrid or supplemental metering, and applying insights from engineering knowledge and audits.
ENERGY AND BUILDINGS
(2021)
Article
Computer Science, Information Systems
Azkario Rizky Pratama, Frank Johan Blaauw, Alexander Lazovik, Marco Aiello
Summary: Precision fine-grained office occupancy detection can be utilized for energy savings in buildings by utilizing power monitoring systems at the level of room circuit breakers. This approach, based on statistical methods, contributes to building context awareness, crucial for achieving energy-efficient buildings. The proposed method is non-intrusive, precise, and can be implemented using machine learning approaches such as nearest neighbors and neural networks.
Review
Engineering, Electrical & Electronic
Hafiz Khurram Iqbal, Farhan Hassan Malik, Aoun Muhammad, Muhammad Ali Qureshi, Muhammad Nawaz Abbasi, Abdul Rehman Chishti
Summary: NILM is a hot topic among researchers, with energy disaggregation datasets used as benchmarks for algorithm validation. This paper provides a comprehensive review of 42 NILM datasets, highlighting strengths, limitations, and future research directions.
ELECTRIC POWER SYSTEMS RESEARCH
(2021)
Article
Energy & Fuels
Tamara Todic, Vladimir Stankovic, Lina Stankovic
Summary: With the widespread deployment of smart meters, Non-Intrusive Load Monitoring (NILM) has emerged as a promising application for informing energy management within buildings. However, existing deep learning NILM models have limitations in terms of flexibility and scalability. In this study, an active learning framework is proposed to improve transferability and reduce the cost of labeling, achieving optimal accuracy-labelling effort trade-off. The results demonstrate the potential of this approach in improving the performance of NILM models and reducing computational resources needed.
Article
Engineering, Electrical & Electronic
Chao Wang, Zhao Wu, Wenxiong Peng, Weihua Liu, Linyun Xiong, Tao Wu, Lili Yu, Huaiqing Zhang
Summary: NILM is an energy-saving technology that estimates energy consumption information from aggregated power data. This paper presents an Adaptive Factorial Hidden Markov Model (Adaptive-FHMM) to characterize appliance working states transition and predict power consumption. Experimental results demonstrate that the proposed model outperforms state-of-the-art models in various metrics.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2022)
Article
Chemistry, Analytical
Nasrin Kianpoor, Bjarte Hoff, Trond Ostrem
Summary: This study proposes an adaptive ensemble filtering framework integrated with long short-term memory (LSTM) for identifying flexible loads. The framework, called AEFLSTM, combines filtering techniques and LSTM to filter noise and learn the long-term dependencies of flexible loads. It also searches ensemble filtering techniques to find the best method for disaggregating different loads.
Article
Computer Science, Information Systems
Claudia De Vizia, Edoardo Patti, Enrico Macii, Lorenzo Bottaccioli
Summary: This study examines the effectiveness of residential demand side management (DSM) strategies in smart grids, emphasizing the importance of user participation. By simulating and analyzing a centralized residential DSM program, the results highlight the significance of user satisfaction in determining participation.
IEEE SYSTEMS JOURNAL
(2022)
Article
Automation & Control Systems
Enrico Tabanelli, Davide Brunelli, Andrea Acquaviva, Luca Benini
Summary: This article addresses the optimization of feature spaces and the reduction of computational and storage costs for running low-latency NILM on low-cost MCU-based meters. The experimental results demonstrate that optimizing the feature space enables edge MCU-based NILM with high accuracy, achieving a significant reduction in cost.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Engineering, Multidisciplinary
Abouzar Estebsari, Pietro Rando Mazzarino, Lorenzo Bottaccioli, Edoardo Patti
Summary: This article presents a developed real-time management schema based on Internet of things solutions, which facilitates interactions between system operators and aggregators for ancillary services. The authors have also developed corresponding algorithms and demonstrated the applicability of the schema through a real-time simulation-based test bed.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2022)
Article
Computer Science, Hardware & Architecture
Emanuele Parisi, Francesco Barchi, Andrea Bartolini, Andrea Acquaviva
Summary: This article discusses the application of deep learning in source code analysis, introducing techniques that utilize different networks and input information. The impact on the accuracy of DL methods and how to extract effective information are also explored. By studying, it was found that normalizing auxiliary information can improve accuracy, and a new method is proposed to enhance mapping accuracy by increasing the dataset's cardinality.
IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS
(2022)
Article
Engineering, Multidisciplinary
Luca Barbierato, Enrico Pons, Andrea Mazza, Ettore Francesco Bompard, Vetrievel Subramaniam Rajkumar, Peter Palensky, Enrico Macii, Lorenzo Bottaccioli, Edoardo Patti
Summary: This article presents an innovative digital real-time cosimulation infrastructure that reduces communication latency and respects real-time constraints by using the Aurora 8B/10B protocol and the IEEE 1588 precision time protocol (PTP) standard. The experimental results demonstrate the stability and accuracy of the proposed infrastructure, making it suitable for scaling future smart grid real-time simulations.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Luca Barbierato, Enrico Pons, Ettore Francesco Bompard, Vetrivel S. Rajkumar, Peter Palensky, Lorenzo Bottaccioli, Edoardo Patti
Summary: This study presents a co-simulation method based on digital real-time simulators connected via Aurora 8B/10B protocol, which allows for the analysis of complex and hybrid system setups while maintaining numerical stability. With the increased share of inverter-based renewable power generation, larger-scale interconnected EMT system studies are needed. This work contributes to a better understanding of the phenomena associated with advanced co-simulation setups using DRTS.
IEEE SYSTEMS JOURNAL
(2023)
Article
Engineering, Multidisciplinary
Simone Eiraudo, Luca Barbierato, Roberta Giannantonio, Alessandro Porta, Andrea Lanzini, Romano Borchiellini, Enrico Macii, Edoardo Patti, Lorenzo Bottaccioli
Summary: Benchmarking buildings based on their electric profiles is an important step in identifying, evaluating, and implementing energy efficiency actions. Temporal data clustering is an effective tool for this purpose, and we propose a novel machine learning methodology that combines decomposition and clustering algorithms. The proposed framework achieved high accuracy in classifying buildings based on their usage category and provides reference key performance indicator values for each cluster to understand energy behavior and possible inefficiencies.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2023)
Article
Energy & Fuels
Marco Castangia, Riccardo Sappa, Awet Abraha Girmay, Christian Camarda, Enrico Macii, Edoardo Patti
Summary: This paper introduces a novel anomaly detection method based on unsupervised deep learning techniques for detecting electrical faults in household appliances. By training a variational autoencoder to analyze the power signatures of appliances, the method shows higher classification accuracy compared to traditional algorithms, allowing for better characterization of normal cycles and more precise alerts.
SUSTAINABLE ENERGY GRIDS & NETWORKS
(2022)
Article
Automation & Control Systems
Marco Castangia, Nicola Barletta, Christian Camarda, Stefano Quer, Enrico Macii, Edoardo Patti
Summary: In smart grids, consumers can participate in demand response programs to reduce household power consumption during peak hours. However, utility companies face challenges in implementing these programs due to their negative impact on user comfort. This article uses neural networks and clustering algorithms to analyze power signatures of appliances and identify different operational modes. By studying washing machines and dishwashers, the analysis reveals distinct working programs based on energy consumption and duration, providing insights for improving demand response programs and reducing overall energy usage by promoting lighter operation modes.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Proceedings Paper
Computer Science, Hardware & Architecture
Mohsen Seyedkazemi Ardebili, Andrea Bartolini, Andrea Acquaviva, Luca Benini
Summary: This paper investigates thermal anomaly detection task in Marconi100, one of the most powerful HPC systems in the world, and successfully validates the suggested method against real thermal hazard events in production.
HIGH PERFORMANCE COMPUTING, ISC HIGH PERFORMANCE 2022 INTERNATIONAL WORKSHOPS
(2022)
Article
Computer Science, Information Systems
Giovanni Chiarion, Francesco Ponzio, Stefano Terna, Cristina Moglia, Edoardo Patti, Santa Di Cataldo, Silvestro Roatta
Summary: This paper introduces a low-cost, stand-alone, portable, and smart device called e-Pupil that uses the pupillary accommodative response as a communication tool. The device utilizes the Internet of Things for portability, accessibility, and usability, and offers two different routes for communication with the external world. The experimental validation confirms the reliability and simplicity of use of the system.
Article
Computer Science, Information Systems
Mulugeta Weldezgina Asres, Luca Ardito, Edoardo Patti
Summary: This study analyzes the energy spent on executing online non-intrusive load monitoring algorithms and proposes a generic framework for large-scale deployment of such algorithms in cloud computing systems. The prediction models developed using statistical and machine learning tools demonstrate the promising applicability of the data-driven approach.
IEEE TRANSACTIONS ON CLOUD COMPUTING
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Nicola Elia, Francesco Barchi, Emanuele Parisi, Livio Pompianu, Salvatore Carta, Andrea Bartolini, Andrea Acquaviva
Summary: Monitoring applications are becoming increasingly important in industrial and civil safety-critical infrastructures. This research proposes a blockchain-based framework for continuous monitoring applications, enabling the certified removal of IoT data in safety-critical databases. The use of smart contracts for data evaluation policies ensures trustworthy operations.
ADVANCES IN DATABASES AND INFORMATION SYSTEMS, ADBIS 2022
(2022)
Review
Engineering, Electrical & Electronic
Francesco Barchi, Emanuele Parisi, Andrea Bartolini, Andrea Acquaviva
Summary: To cope with the complexity of programming digital systems, deep learning techniques have been proposed for source code analysis, particularly for efficient kernel mapping on heterogeneous platforms. However, it is challenging to determine which techniques are most suitable for cyber-physical systems. This paper discusses recent developments in deep learning for source code analysis and highlights the opportunities and challenges for their application to this class of systems.
JOURNAL OF LOW POWER ELECTRONICS AND APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Davide Cannizzaro, Antonio Giuseppe Varrella, Stefano Paradiso, Roberta Sampieri, Yukai Chen, Alberto Macii, Edoardo Patti, Santa Di Cataldo
Summary: Metal Additive Manufacturing (AM) is an important aspect of Industry 4.0, offering several advantages over traditional subtractive fabrication techniques. However, quality issues can hinder mass production. This article proposes a solution that utilizes computer vision and machine learning algorithms for real-time defect monitoring, with the generation of synthetic images using Generative Adversarial Network (GAN) for data augmentation.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTING
(2022)