Received 26 June 2024, accepted 26 July 2024, date of publication 7 August 2024, date of current version 19 August 2024.

Digital Object Identifier 10.1109/ACCESS.2024.3440067

AI Technologies and Their Applications in

Small-Scale Electric Power Systems
ARQUM SHAHID 1 , (Graduate Student Member, IEEE), FREDDY PLAUM 1 ,
TARMO KORÕTKO 1 , (Member, IEEE), AND ARGO ROSIN 1,2 , (Senior Member, IEEE)
1 Department of Electrical Power Engineering and Mechatronics, Tallinn University of Technology, 12616 Tallinn, Estonia
2 FinEst Centre for Smart Cities, Tallinn University of Technology, 12616 Tallinn, Estonia
Corresponding authors: Freddy Plaum ([email protected]) and Arqum Shahid ([email protected])
This work was supported by the project ‘‘Increasing the Knowledge Intensity of Ida-Viru Entrepreneurship’’ co-funded by the European
Union under Grant 2021-2027.6.01.23-0034.

ABSTRACT As the landscape of electric power systems is transforming towards decentralization, small-
scale electric power systems have garnered increased attention. Meanwhile, the proliferation of artificial
intelligence (AI) technologies has provided new opportunities for power system management. Thus, this
review paper examines AI technology applications and their range of uses in small-scale electrical power
systems. First, a brief overview of the evolution of small-scale electric power systems and the importance
of AI integration is given. The background section explains the principles of small-scale electric power
systems, including stand-alone systems, grid-interactive systems, microgrids, hybrid systems, and virtual
power plants. A thorough analysis is conducted on the effects of AI technologies on power system aspects
such as energy consumption, demand response, grid management, operation, energy generation, and storage.
Based on this foundation, AI Acceleration Performance Indicators (AAPIs) for small-scale electric power
systems are developed to establish a standardized framework for evaluating and comparing different studies.
AAPI framework considers a binary scoring for five quantitative Key Performance Indicators (KPIs) and
five qualitative KPIs examined through a three-tiered scale – established, evolved, and emerging.

INDEX TERMS Artificial intelligence, electric power systems, performance indicators.

I. INTRODUCTION wind and water to produce electricity. These systems were

A. BRIEF OVERVIEW OF THE EVOLUTION OF distinguished by their independence from large-scale grids,
SMALL-SCALE ELECTRIC POWER SYSTEMS and their location frequently served rural populations. There
Significant developments in societal expectations, regulatory was a technological innovation boom in the late 20th century,
frameworks, and technology paradigms have shaped the especially in the area of renewable energy. Photovoltaic cells,
evolution of small-scale electric power systems. Small-scale wind turbines, and other clean energy technologies grew more
systems have historically served isolated locations or sectors, efficient and affordable, as demonstrated by the increase in
taking on a supporting role to centralized power grids. Due solar energy output from 30 GW to 118 GW and wind energy
to technological breakthroughs, renewable energy sources production from 78 GW to 167 GW within the European
have become more prevalent over time, and power gen- Union between 2010 and 2019 [2].
eration equipment has become more affordable, propelling The need to switch to carbon-neutral energy sources
small-scale systems to become an essential component of has become more pressing due to growing worries about
modern-day sustainable energy solutions [1]. climate change and environmental pollution. As essential
Decentralized energy production emerged in the early 20th parts of the broader energy infrastructure, small-scale electric
century when small-scale systems used local resources like power systems are crucial in reducing the carbon footprint
of conventional energy sources [3]. Nations all across the
The associate editor coordinating the review of this manuscript and globe have pledged to cut greenhouse gas emissions and
approving it for publication was Md. Rabiul Islam . move toward sustainable energy practices under international
2024 The Authors. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
109984 For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/ VOLUME 12, 2024
A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

agreements like the Paris Agreement [4]. Since small-scale Because renewable energy resources, such as wind and
electric power systems allow for localized, clean energy solar power, are unpredictable, sophisticated forecasting
production with lower transmission losses, they provide methods are required. To generate reliable renewable energy
a reasonable and practical solution to meet these global generation forecasts, AI algorithms analyze meteorological
sustainability targets. data, historical trends, and current conditions [13]. This
Although small-scale systems have evolved promisingly, makes it possible for grid operators to effectively incorporate
there are still challenges, especially when incorporating renewable energy into small-scale power systems and manage
fluctuating renewable energy sources. Advanced solutions fluctuations proactively.
like aggregated energy flexibility are required for efficient For small-scale electric power systems to balance supply
grid management because of the operational problems and demand, energy storage systems need to be optimized
posed by the intermittent and variable nature of renew- - AI technologies are key to this process. Optimizing
able energy sources like solar and wind [5]. In addition energy storage devices’ charging and discharging processes
to benefiting the environment, small-scale electric power enhances their lifespan and efficiency, which is achieved via
systems empower nearby communities by promoting energy machine learning algorithms that analyze demand patterns,
independence, creating job opportunities, and boosting weather forecasts, and grid conditions [14].
the local economy [6]. Decentralized energy resources Demand response programs powered by AI enable users
in small-scale systems improve community sustainability to actively participate in energy-saving activities. These
overall and increase resilience to disruptions from the main technologies help to increase overall energy efficiency and
grid. sustainability by allowing users to modify their energy
consumption according to grid conditions through intelligent
B. IMPORTANCE OF INTEGRATING AI TECHNOLOGIES IN automation and real-time communication [15].
MODERN ENERGY SYSTEMS The paper is organized as follows: Section II provides
The incorporation of artificial intelligence (AI) technology background information on small-scale electric power sys-
presents unique potential for enhancing the performance tems. Section III is dedicated to an in-depth analysis of the
and reliability of small-scale electric power infrastructures. existing literature related to AI applications in small-scale
By employing machine learning, predictive maintenance electric power systems. Section IV discusses the find-
algorithms can evaluate past data, identify patterns, and ings and proposes AI Acceleration Performance Indicators
anticipate equipment breakdowns before they happen [7]. (AAIPs) that enable evaluating and comparing different
This lowers total maintenance costs by extending the lifespan studies. Section V concludes the review paper with relevant
of crucial components and minimizing downtime [8]. findings.
AI-powered load forecasting models make real-time
energy demand forecasts possible, making grid management C. RELATED WORK AND MOTIVATION
and resource allocation more effective [9]. These models The deployment of AI in power systems has become topical in
improve the flexibility of small-scale systems by analyzing the scientific literature as the number of publications related
variables like user behavior, weather patterns, and past to deep learning and electric power systems in the ScienceDi-
consumption data, guaranteeing that supply and demand rect database has grown from around 20 in 2015 to 200 in
are balanced [10]. Given that renewable energy sources 2019 [16]. Review articles related to this paper primarily
are naturally uncertain, this capability becomes even more focus on AI’s applications in power systems [17], [18]. For
essential. example, the research status in the operation, optimization,
The introduction of AI-powered smart grid technologies control, dispatching, and management of Smart Grid and
is revolutionizing energy transmission, distribution, and Energy Internet fields using AI has been reviewed in [19],
usage. Fig. 1 demonstrates the components of small-scale where it was found that the bottlenecks for future develop-
power systems, which are the scope of this review paper. ment include the lack of training datasets, the interpretability
AI algorithms make real-time grid monitoring and control and reliability of models, and semantic reasoning issues
possible, allowing for automatic response to changing of language models. Machine learning algorithms, such as
conditions. In addition to improving grid stability, it regulates Support Vector Machines (SVMs) and Gradient Boosting
fluctuations and keeps a steady supply of power, which Machines (GBMs), have been utilized to predict energy
facilitates the integration of various energy sources, including consumption patterns with high accuracy, enabling more
renewables [11]. efficient demand response and load forecasting [20], [21].
AI plays a crucial role in coordination and control as AI supports VPPs by optimizing the utilization of renewable
small-scale electric power systems adopt increasingly decen- resources based on their availability and demand predic-
tralized energy resources. Decentralized energy management tions [22]. Deep learning models, particularly Convolutional
systems use AI to balance loads, optimize power flows, and Neural Networks (CNNs) and Long Short-Term Memory
coordinate the use of various energy sources. This raises (LSTM) networks, have demonstrated exceptional capability
the system’s overall efficiency and strengthens the grid’s in identifying and diagnosing faults within microgrids,
resistance to disturbances [12]. thus reducing downtime and maintenance costs [23], [24].

VOLUME 12, 2024 109985

 A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

FIGURE 1. Illustration of the structure of a power system.

Furthermore, reinforcement learning approaches, including The inclusion criteria for selecting relevant studies include
Deep Q-Networks (DQNs) and Proximal Policy Optimization peer-reviewed journal articles and conference papers that
(PPO), have been utilized to optimize microgrid operations specifically address the focus of this study, namely the impact
and manage distributed energy resources more effectively, of AI technologies in small-scale electric power systems.
thereby enhancing overall system performance and sustain- Studies that either provided insufficient information about
ability [25], [26]. The authors of [27] reviewed Explainable the uses of AI technology or did not explore the relationship
Artificial Intelligence techniques for energy and power between AI and small-scale electric power networks were
systems. The application of resilience enhancement of power excluded from the analysis.
systems using AI was reviewed by the authors of [28], The key themes of the literature were identified through
who concluded that supervised deep learning is particularly systematic data extraction. The data was compiled into tables
suited for anomaly detection, classification, and damage based on which the objectives, methodologies, AI models,
detection. In contrast, unsupervised deep learning methods key findings, and limitations of existing research can be
are suitable for defending against cyber-attacks. Thus, the analyzed.
research has primarily focused on AI applications in power
systems; however, to the authors’ best knowledge, there II. FUNDAMENTALS OF SMALL-SCALE ELECTRIC POWER
is a lack of research in evaluating and benchmarking the SYSTEMS
efficacy of AI implementations in electric power systems. The effective operation of small-scale electric power systems
Therefore, the motivation of this research paper is not is essential in meeting the changing energy demands. The
only to give a comprehensive review of AI applications in basic concepts of these systems, which include standalone,
small-scale electric power systems but also to provide a grid-interactive, microgrid, hybrid, and other configurations,
framework for evaluating and benchmarking the efficacy of including Virtual Power Plants (VPPs), are examined in this
AI implementations using the AAPI framework developed in section.
this paper.
A. STANDALONE SYSTEMS
D. REVIEW METHODOLOGY Reliable electricity supply in isolated or off-grid places
A thorough literature search was conducted across major relies heavily on small-scale electric power systems, mainly
academic databases such as Scopus, ScienceDirect, and standalone designs. These systems have become essential in
IEEE Xplore. The search strategy included a combination addressing issues related to energy access because of their
of the following keywords and many more: ‘‘AI tech- independence from the main grid [29]. Standalone systems
nologies,’’ ‘‘small-scale electric power systems,’’ ‘‘micro- include devices such as production units, energy storage, and
grids,’’ ‘‘energy consumption,’’ ‘‘demand response,’’ ‘‘grid loads. Energy is produced with diesel generators, combined
management,’’ ‘‘energy generation,’’ and ‘‘energy storage.’’ heat and power units, or renewable energy sources like
The focus was on newer studies conducted from 2019 solar or wind. At the same time, the storage, which usually
to 2024. takes the form of batteries, guarantees a steady supply of

109986 VOLUME 12, 2024

A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

electricity at times when production is low. Optimizing the D. HYBRID SYSTEMS

performance of standalone systems requires understanding Hybrid systems are a complex solution to small-scale electric
how these components interact [30]. power systems since they integrate multiple energy sources.
Although standalone systems provide energy indepen- These systems combine the benefits of many technologies
dence, they have maintenance, fuel supply, and reliability by integrating renewable energy sources with conventional
issues. Despite these challenges, standalone systems are generators [40]. The viability and versatility of this technique
viable for some applications due to their flexibility, auton- are demonstrated by examples of hybrid systems, such
omy, and lower environmental impact [31]. Examples from as wind-hydro or solar-diesel combinations. Intermittency-
real life demonstrate the adaptability and efficiency of related issues are resolved by combining renewable and
standalone systems [32]. Applications for standalone systems conventional sources to ensure a more steady power
are diverse; they can be used to power distant communication output [41]. Other benefits include better environmental
stations or provide electricity in rural areas or areas affected sustainability, decreased dependency on fossil fuels, and
by disasters. In situations when grid access is difficult or enhanced efficiency. However, the challenges in designing
economically unreasonable, these systems showcase their and integrating complex systems call for both careful
importance in meeting energy demands [33]. planning and innovative technologies [42].

B. GRID-INTERACTIVE SYSTEMS
E. VIRTUAL POWER PLANTS
Grid-interactive systems are a type of small-scale electric
power systems that integrate with the main grid. By enabling The concept of Virtual Power Plants (VPPs) is new in the
bidirectional power flow, these systems allow an interchange world of small-scale electric power systems. These designs
of electricity between the main grid and the local power provide a scalable and adaptable solution by combining
distributed energy resources through the use of modern
sources [34].
technologies. Due to their ability to coordinate operations
Integrating grid-interactive equipment with the main grid is
a crucial component that makes a consistent and dependable centrally, VPPs are essential for optimizing the usage
power supply possible. Grid compatibility and control of distributed resources [43]. Beyond conventional power
methods are subject to additional challenges in the context of generation, VPPs are also applicable for energy storage and
bidirectional power flow, allowing electricity to be provided demand-side control, which improves system efficiency as a
to and consumed from the grid [35]. whole [44]. Improved stability of the grid, effective resource
Grid-interactive systems have several advantages, such as use, and a lower carbon footprint are just a few of the
improved energy efficiency and higher reliability, thanks to economic and environmental advantages that come with the
grid assistance. However, for deployment to be effective, deployment of VPPs [45]. It is anticipated that as technology
obstacles to maintaining grid stability and resolving regula- develops, VPPs will have an even more significant impact on
tory concerns must be carefully considered [36]. small-scale electric power systems.

C. MICROGRIDS III. AI APPLICATIONS IN SMALL-SCALE ELECTRIC

Microgrids represent a significant shift in small-scale POWER SYSTEMS
electrical power systems, offering localized control and Integrating Artificial Intelligence into small-scale electric
independence. Microgrids are characterized by having the power systems presents a promising opportunity for man-
ability to function both autonomously and alongside the aging and optimizing energy resources distinct from those
main grid, i.e., in off-grid or on-grid modes. These attributes encountered in large-scale systems. While AI’s applications
are among the major reasons for their increasing appeal. in both contexts aim to enhance efficiency, reliability, and
There are several use cases of microgrids, each designed to optimization, the scale of operation significantly influences
meet specific requirements, e.g., community, campus, and the nature and impact of these applications.
remote microgrids [37]. It is necessary to understand these In small-scale systems, the applications of AI range from
distinctions to develop microgrids that meet the particular enhancing the efficiency and reliability of distributed energy
needs of various settings. resources, such as through predictive maintenance, optimal
Microgrid management is greatly aided by advanced segmentation of renewable sources, and accurate forecasting,
control systems, which ensure optimal performance and to optimizing battery energy storage and consumption by
coordination between various energy sources. The respon- predicting remaining useful life (RUL), SoC patterns, and
siveness and flexibility of microgrid systems are improved by charging and discharging times. AI in these settings is focused
integrating intelligent technologies such as optimization and on enhancing local grid stability, managing dynamic load,
machine learning algorithms [38]. Enhancing power supply and integrating a higher proportion of renewable energy
reliability is one of microgrids’ distinguishing features. sources. Due to the smaller scale, AI-driven strategies are
Microgrids play an important role in attaining energy security more agile, tailored to local conditions, and responsive to
and contribute to grid stability by offering localized solutions rapid changes in demand and supply. It plays a pivotal
to energy-related problems [39]. role in intelligent load management and demand response,

VOLUME 12, 2024 109987

 A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

providing dynamic pricing strategies and balancing sup- B. GRID MANAGEMENT AND OPERATIONS
ply and demand while predicting consumer behavior for Energy fluctuations from intermittent renewable energy
optimized energy distribution. Furthermore, AI significantly generations introduce vulnerability in grid operations [69].
boosts small-scale grid management by enabling real-time ML plays a crucial role in transforming grid management,
anomaly detection, predictive maintenance, and dynamic particularly in enhancing the capabilities for on-grid sys-
reconfiguration in microgrids to enhance grid stability and tem optimizations, dynamic reconfiguration in microgrids,
resilience and maintain continuous and efficient power and anomaly detection in power systems. For example,
delivery. the LSTM-based reinforcement learning model improved
On the other hand, AI applications in large-scale power renewable energy integration and load balancing optimization
systems typically deal with the complexity of interconnected in a smart grid with 92% accuracy as compared to other
networks and centralized generation facilities, focusing more ML algorithms [70]. On-grid system optimization involves
on high-level grid management, large-scale energy trading, interactions between various microgrid components such as
and maintaining the reliability and security of supply across consumers, renewable energy producers, electricity suppliers,
vast geographical areas. and storage systems. This interaction is characterized by
The scope of AI in small systems extends to sophisticated dynamic reconfiguration, adapting microgrid operations to
applications such as coordinating VPP and community varying factors like renewable energy production, con-
energy systems, aggregating, and intelligently managing sumption patterns, and storage capacities [71]. A techno-
diverse energy resources. Following the overview, the environmental-economic strategy using multi-agent DRL for
subsequent sections will thoroughly discuss the specifics of microgrid planning and optimization is presented in [72].
each area, exploring the enhancement of energy generation, Effective grid management requires improved prediction
storage, consumption, grid management, and advanced appli- stability of microgrids. This includes load-shifting, demand
cations within small-scale electric power systems through AI offsetting, decision-making in virtual power plants, and
technologies. providing ancillary services, thereby focusing on urban
scales and their inherent complexities [73]. To ensure
grid reliability and security, federated learning techniques
A. ENERGY CONSUMPTION AND DEMAND RESPONSE allow for on-device model training and parameter updat-
In the domain of small-scale electric power systems, the ing, significantly enhancing privacy and reducing data
application of artificial intelligence in energy consump- transmission requirements. These approaches, secured with
tion and demand response offers model-free solutions as SSL/TLS protocols, effectively mitigate challenges related
compared to traditional mathematical models to analyze con- to bandwidth, latency, and security, aligning with stringent
sumption patterns, predict demand peaks, exploit consumer privacy regulations [74]. To provide security to client data
energy flexibility, and implement dynamic load adjustments, in microgrids from being compromised, a CNN-BiLSTM
perform real-time pricing and offering innovative solu- categorization criterion for cyber-attacks has shown a success
tions for intelligent energy management at both household rate of 99% compared with traditional approaches [75].
and building scales either with residential or community Table 2 summarizes the research on AI applications in grid
settings. management and operations of small-scale electric power
A DRL algorithm to schedule ESS and HVAC loads systems.
in a smart home without building thermal dynamics is
proposed in [46]. The results indicate 8.10%–15.21% cost C. ENERGY GENERATION
minimization compared to rule-based control approaches. Smart grid technology has enabled the potential benefits of
In [47], the Temporal Convolutional Networks (TCNs) are RES for consumers in small-scale electric power systems.
utilized for community energy management using PV and In this context, the application of AI becomes instrumental
ESS. Energy consumption optimization includes data-driven in enhancing energy generation capabilities by optimally
models for occupant behavior, user comfort, and RES positioning and controlling RES to maximize the efficiency
management using Random Forest [48], NARX ANN [49], of these installations, specifically for the task of maximum
DNN [50], and Q-Learning [51]. Similarly, energy demand power point tracking (MPPT) and adaptive power manage-
prediction for economic and energy savings is also a key ment [76]. By analyzing historical data from various sensors,
aspect of DR strategies. In the literature, authors employed AI algorithms predict potential failures, remaining useful
different machine-learning techniques for short-term [52], life (RUL), and schedule timely maintenance of equipment,
[53], [54] and day-ahead load forecasting [55], [56], thus minimizing downtime and extending the lifespan of the
[57] to enhance consumer engagement in energy trading, generation equipment. Furthermore, accurate solar irradiance
renewable energy integration, and dynamic tariff schemes. and wind speed forecasts enable proper load scheduling and
Table 1 comprehensively examines AI-driven strategies grid power allocation, ensuring a steady and reliable energy
for enhancing energy consumption patterns and refining supply [77]. A SHAP cat-boost algorithm improves MPPT
demand response mechanisms for small-scale electric power control in PV systems by minimizing steady-state error
systems. during low irradiance and partial shading conditions [78].

109988 VOLUME 12, 2024

A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

TABLE 1. Summary of AI studies for energy consumption and demand response.

VOLUME 12, 2024 109989

 A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

TABLE 1. (Continued.) Summary of AI studies for energy consumption and demand response.

A Q-learning-based control strategy has identified optimal solutions, predicting battery health, optimizing charging
equilibrium policies for various power system operating cycles based on real-time energy demands, and identifying
conditions and improved control performance by around 10% degradation patterns [103]. A predictive control mechanism
compared to other ML algorithms [79]. Similarly, for predict- has demonstrated an 84% overall efficiency in microgrid
ing the RUL of rotating machines, a DNN-based model is peak shaving by managing the flow rate of energy storage
utilized that considers time–frequency-wavelet joint features systems for stable power generation [104]. To improve the
to effectively represent the degradation of bearings [80]. real-time charging/discharging decision-making of ESS, RL-
A deep learning-based RNN model is designed to forecast based actor-critic agents are used to optimize the power flow
short-term intra-hour solar irradiance by using infrared sky while minimizing the energy cost [105]. Battery state of
images, resulting in reduced algorithm computational cost health is determined with a mean absolute error of 1.39%
and grid operational cost with high participation of solar by using a simple ANN with a small amount of data. This
energy [81]. Table 3 provides a detailed overview of AI helps optimize the operation and management of energy
applications for optimizing energy generation in small-scale storage systems [106]. A degradation model of lithium
electric power systems. batteries is developed to predict the remaining useful life
using ensemble learning methods for fault diagnosis during
D. ENERGY STORAGE the equipment operation service period to ensure an effective
Efficient energy storage management is essential for the energy supply [107].
effectiveness and reliability of small-scale electric power Table 4 analyzes AI-based techniques to improve the
systems that rely on intermittent renewable energy sources, operation of energy storage systems in small-scale electric
such as solar and wind [101]. The development of energy power systems.
storage system (ESS) technologies such as compressed air,
flywheel, pumped hydro storage, and batteries can increase IV. DISCUSSION
the ESS capacity to store energy from power grids. This A. AI ACCELERATION PERFORMANCE INDICATORS
stored energy can then be used when needed. The advance- (AAPIS) FOR SMALL-SCALE ELECTRIC POWER SYSTEMS
ment of ESS technologies with microgrid utilization has In the rapidly evolving field of small-scale electric power
created a large market for ESS to offer bulk energy storage, systems, the integration of AI has shown promising potential.
transmission and distribution support, ancillary services, However, a critical gap exists in the standardization of
and energy management solutions [102]. AI technologies evaluating and comparing the diverse AI methodologies
significantly enhance the capabilities and functionalities of being employed for similar tasks. This requires a set of
ESS by providing battery-based control and monitoring baseline assessment parameters to establish a standardized

109990 VOLUME 12, 2024

A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

TABLE 2. Summary of AI studies for grid management and operations.

VOLUME 12, 2024 109991

 A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

TABLE 2. (Continued.) Summary of AI studies for grid management and operations.

framework that enables evaluating and comparing different management, and prediction accuracy, along with qualitative
studies. Considering the variations in methodologies and aspects like innovation level and practical applicability.
outcomes in energy sector research, the AI-Acceleration AAPIs framework employs a binary scoring system for
Performance Indicators (AAPIs) are proposed as an initial quantitative KPIs to highlight which aspects are clearly
proposition to provide a consistent benchmark for AI- covered in the studies. In contrast, the qualitative aspects are
accelerated approaches. It involves identifying key per- examined through a three-tiered scale – established, evolved,
formance indicators (KPIs) crucial for evaluating AI in and emerging, where established indicator shows real-world,
small-scale electric power systems. The process is guided by data-driven, and validated AI solutions with reliable results
the dual objectives of ensuring technological viability and in different operating scenarios, evolving parameter shows
enhancing user-centric outcomes. AAPI framework serves the ongoing development and incremental improvements in
as a starting point for standardization in the field, with the the research with simulated analysis to enhance practical
main purpose of establishing a foundation upon which further viability. In contrast, emerging shows new machine learning
research and validation can be built. concepts and early-stage AI solutions that are yet to be
The accelerated KPIs are designed to speed up the commer- extensively tested but point to new directions that could
cialization of AI technologies in energy systems by ensuring drive future advancements. This assessment approach guides
user comfort, scalability, and practical applicability while the field towards practical, user-oriented, and commercially
enhancing user engagement. The framework categorizes sustainable AI solutions.
KPIs into quantitative and qualitative measures, as outlined These indicators are applicable across various types
in Table 5. of AI applications, be they computational, experimental,
Key performance areas critical to AI applications in energy or integrative. Researchers can track the evolution and
systems are identified, such as cost-effectiveness, demand performance enhancements of these systems by consistently

109992 VOLUME 12, 2024

A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

TABLE 3. Summary of AI studies for optimizing energy generation for small-scale electric power systems.

VOLUME 12, 2024 109993

 A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

TABLE 4. Summary of AI studies for energy storage in small scale power systems.

109994 VOLUME 12, 2024

A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

TABLE 4. (Continued.) Summary of AI studies for energy storage in small scale power systems.

TABLE 5. Overview of assessment criterion for AI applications. indicates enhanced AI capability in managing energy demand
by improving load forecasting and performing complex oper-
ational decisions by executing real-time analytics to modulate
energy supply in correspondence with consumption patterns.
However, energy generation and energy storage register a less
pronounced engagement with AI for demand management,
implying that current research has not fully exploited
the potential of AI in this regard. Fig. 2 demonstrates
the comparative performance of key indicators by scoring
AI-based reviewed articles in various application areas – such
as energy storage, grid management and operations, energy
generation, and energy consumption and demand response
against the quantitative KPIs of cost-effectiveness, demand
management, prediction accuracy, computational simplicity,
and user comfort to highlight emerging research trends.
The qualitative assessment of the reviewed articles pro-
vides information about the advancement and maturity
of AI-accelerated solutions within diverse domains of
small-scale electric power systems. In the case of energy
storage, AI applications are mainly in the evolving phase
as methods are being developed for more accurate battery
RUL and SOC predictions. In the same way, AI applications
applying AAPIs in the development and assessment of seem more established for grid management and operations
new AI-based energy platforms. This standardized approach due to the proven effectiveness of reinforcement learning in
enables comparing varied AI algorithms, from traditional optimizing the decision-making process of integrating and
algorithm-based systems to more advanced, innovative maximizing the use of renewables in microgrids and virtual
applications such as deep neural network models. To pro- power plants. Similarly, the higher innovation level in the
vide context and demonstrate the potential application of case of energy consumption and demand response indicates
the AAPIs, the paper applies the framework to various real-world implementation of most of the AI applications
AI-driven studies in the realm of small-scale electric power in forecasting demand, optimizing energy consumption,
systems. As a demonstration, these KPIs are applied to and scheduling controllable appliances with more research
various studies to evaluate their accelerated performance for focused on improving the already developed solutions for
different applications of electric power systems, as shown in better grid operational efficiency. The scalability potential for
Table 6. energy generation shows the dynamic phase of AI solutions,
The quantitative assessment of the reviewed literature such as federated and transfer learning, in improving
highlights distinct trajectories in the application of AI across renewable energy generation and predicting the maintenance
various domains of small-scale electric power systems. of generation equipment while delving into expanding the
Regarding energy generation and energy storage, AI plays impact of AI in larger systems. Fig. 3 highlights the
a significant role in grid management and operations, qualitative spectrum of small-scale electric power systems
energy consumption, and demand response with respect to across multiple operational domains ranging from established
optimizing renewable energy production in smart grids and practices to emerging innovations within energy storage,
managing energy flexibility for demand response and other grid management and operations, energy consumption, and
processes. The advancement in AI technologies, especially energy generation.
deep learning, makes the prediction accuracy more accurate Similarly, the research trends are more oriented towards
but at the cost of higher computational demand and complex microgrids with household-level energy management to
ML algorithms. The assessment of demand management address various objectives related to renewable energy

VOLUME 12, 2024 109995

 A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

TABLE 6. Demonstration of the use of KPIs in evaluating the performance of AI-driven research in electric power systems.

FIGURE 2. Comparative radar charts illustrating the performance of reviewed articles across key quantitative indicators for small-scale
electric power systems.

FIGURE 3. Qualitative analysis of AI application across different domains in small-scale electric power systems.

optimization, energy efficiency, and load forecasting. indicating the significant emphasis on AI in the residential
Microgrids-related studies represent 29.4% of the literature, sector.
while smart grids and VPPs account for 14.7% and 8.8%, AI applications show promising results in several
respectively, thus pointing towards a trend of decentralized, small-scale power system domains. However, based on
consumer-focused energy solutions. Compared to building- AAPI analysis, certain application limitations and areas
level studies, which comprise 13.3% of the studies, require further research to fully exploit AI’s potential in
household-level studies make up 26.5% of the research, this sector. Many AI applications are data-dependent and

109996 VOLUME 12, 2024

A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

are confined to specific scenarios. For instance, studies on computational simplicity and only 10% considering
using federated learning or interactive learning require user comfort in energy storage methodologies. Conversely,
large, diversified datasets for training and validation that research related to grid management and operations has
impact the performance and generalizability of AI models shown a robust interest in prediction accuracy and demand
to different environments, operational conditions, and grid management, with 80% of articles emphasizing these aspects.
configurations. Most of the studies focus on the specific area Qualitatively, innovation in energy generation has emerged
of energy management systems, such as only consideration as a critical area with approximately 60% of research
of shiftable appliances while lacking integration of thermal marked as ‘Emerging’, indicating a promising frontier for
models with HVAC control systems, which hinders practical future research. The AAPIs framework serves not only as a
applicability. Computational complexity, data synchroniza- benchmarking tool for current research performance but also
tion, and resource demands of advanced AI systems pose guides future AI applications toward achieving user-centric
significant challenges to scalability and real-time application. and economically viable solutions.
AI applications in power systems also pose security and
privacy concerns, such as vulnerability to eavesdropping REFERENCES
attacks and IoT integration in microgrids that affect reliability [1] L. Strezoski, ‘‘Distributed energy resource management systems—
in the residential sector. In the case of energy storage, DERMS: State of the art and how to move forward,’’ WIREs Energy
Environ., vol. 12, no. 1, p. e460, Jan. 2023, doi: 10.1002/wene.460.
AI-based battery energy storage system shows limited [2] Electrical Capacity for Wind and Solar Photovoltaic Power–
focus on different battery chemistries and aging factors. Statistics. Accessed: Jan. 17, 2024. [Online]. Available: https://ec.
By incorporating a broader range of datasets, enhancing the europa.eu/eurostat/statistics-explained/index.php?title=Electrical
_capacity_for_wind_and_solar_photovoltaic_power_-_statistics
processing capability of AI models in dynamic environments, [3] M. A. Hannan, M. Faisal, P. Jern Ker, R. A. Begum, Z. Y. Dong, and
implementing robust security protocols, and exploring C. Zhang, ‘‘Review of optimal methods and algorithms for sizing energy
unified AI models that can adapt to various power system storage systems to achieve decarbonization in microgrid applications,’’
Renew. Sustain. Energy Rev., vol. 131, Oct. 2020, Art. no. 110022, doi:
scenarios related to energy storage and grid management will 10.1016/j.rser.2020.110022.
significantly contribute to the user-centric and feasible AI [4] (2015). Paris Agreement (adopted 12 December 2015, Entered Into
solution in small-scale electric power systems. Force 4 November 2016) United Nations Treaty Collection, Chapter
XXVII 7 D. Accessed: Jan. 15, 2024. [Online]. Available: https://treaties.
un.org/Pages/ViewDetails.aspx?src=TREATY&mtdsg_no=XXVII-7-
V. CONCLUSION d&chapter=27&clang=_en
Small-scale electric power systems have been instrumental in [5] F. Plaum, R. Ahmadiahangar, A. Rosin, and J. Kilter, ‘‘Aggregated
enhancing energy resilience and sustainability. These systems demand-side energy flexibility: A comprehensive review on charac-
terization, forecasting and market prospects,’’ Energy Rep., vol. 8,
allow for a more flexible and efficient energy management pp. 9344–9362, Nov. 2022, doi: 10.1016/j.egyr.2022.07.038.
approach, facilitating the local generation, storage, and [6] T. Korõtko, ‘‘Assessment of power system asset dispatch under different
distribution of energy, thereby mitigating the challenges local energy community business models,’’ Energies, vol. 16, no. 8,
Jan. 2023, Art. no. 8, doi: 10.3390/en16083476.
associated with the integration of renewables. This review [7] M. Y. Arafat, M. J. Hossain, and M. M. Alam, ‘‘Machine learning scopes
paper presents an extensive analysis of AI applications within on microgrid predictive maintenance: Potential frameworks, challenges,
these systems, highlighting the transformative role AI plays and prospects,’’ Renew. Sustain. Energy Rev., vol. 190, Feb. 2024,
Art. no. 114088, doi: 10.1016/j.rser.2023.114088.
across various aspects of energy generation, storage, and [8] J. De La Cruz, E. Gómez-Luna, M. Ali, J. C. Vasquez, and J. M. Guerrero,
consumption, offering a unique perspective on the future ‘‘Fault location for distribution smart grids: Literature overview, chal-
trajectory of AI in enhancing the efficiency and reliability of lenges, solutions, and future trends,’’ Energies, vol. 16, no. 5, Jan. 2023,
Art. no. 5, doi: 10.3390/en16052280.
small-scale electric power systems. [9] R. Ahmadiahangar, T. Häring, A. Rosin, T. Korõtko, and J. Martins,
Firstly, a brief overview of small-scale electric power ‘‘Residential load forecasting for flexibility prediction using machine
systems’ evolution is presented. Subsequently, the review learning-based regression model,’’ in Proc. IEEE Int. Conf. Environ.
Electr. Eng. IEEE Ind. Commercial Power Syst. Eur. (EEEIC / ICPS
explores their key role in enhancing the resilience and Europe), Jun. 2019, pp. 1–4, doi: 10.1109/EEEIC.2019.8783634.
efficiency of modern energy distribution. A detailed analysis [10] J. Shi, C. Li, and X. Yan, ‘‘Artificial intelligence for load fore-
of AI across various domains of power systems is presented, casting: A stacking learning approach based on ensemble diversity
regularization,’’ Energy, vol. 262, Jan. 2023, Art. no. 125295, doi:
from optimizing energy consumption and demand response 10.1016/j.energy.2022.125295.
through smart load management and dynamic pricing to [11] S. M. Nosratabadi, R.-A. Hooshmand, and E. Gholipour, ‘‘A com-
enhancing grid operations with real-time anomaly detection prehensive review on microgrid and virtual power plant concepts
employed for distributed energy resources scheduling in power systems,’’
and predictive maintenance. The discussion converges on
Renew. Sustain. Energy Rev., vol. 67, pp. 341–363, Jan. 2017, doi:
the AAPIs framework, representing an initial step towards 10.1016/j.rser.2016.09.025.
establishing a standardized evaluative framework for AI [12] N. M. Kumar, A. A. Chand, M. Malvoni, K. A. Prasad, K. A. Mamun,
applications in small-scale electric power systems. This F. R. Islam, and S. S. Chopra, ‘‘Distributed energy resources and the
application of AI, IoT, and blockchain in smart grids,’’ Energies, vol. 13,
framework incorporates both quantitative and qualitative no. 21, p. 5739, Nov. 2020, doi: 10.3390/en13215739.
KPIs, such as cost-effectiveness, prediction accuracy, and [13] S. E. Haupt, T. C. McCandless, S. Dettling, S. Alessandrini, J. A. Lee,
innovation level, providing a comprehensive metric for S. Linden, W. Petzke, T. Brummet, N. Nguyen, B. Kosovic, G. Wiener,
T. Hussain, and M. Al-Rasheedi, ‘‘Combining artificial intelligence with
assessing AI technologies. The AAPIs framework reveals physics-based methods for probabilistic renewable energy forecasting,’’
significant research trends, with 70% of studies focusing Energies, vol. 13, no. 8, p. 1979, Apr. 2020, doi: 10.3390/en13081979.

VOLUME 12, 2024 109997

 A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

[14] J. Cao, D. Harrold, Z. Fan, T. Morstyn, D. Healey, and K. Li, ‘‘Deep [33] L. Ali and F. Shahnia, ‘‘Determination of an economically-suitable and
reinforcement learning-based energy storage arbitrage with accurate sustainable standalone power system for an off-grid town in western
lithium-ion battery degradation model,’’ IEEE Trans. Smart Grid, vol. 11, Australia,’’ Renew. Energy, vol. 106, pp. 243–254, Jun. 2017, doi:
no. 5, pp. 4513–4521, Sep. 2020, doi: 10.1109/tsg.2020.2986333. 10.1016/j.renene.2016.12.088.
[15] L. A. Arias, E. Rivas, F. Santamaria, and V. Hernandez, ‘‘A review and [34] A. Satchwell, ‘‘A national roadmap for grid-interactive efficient build-
analysis of trends related to demand response,’’ Energies, vol. 11, no. 7, ings,’’ Lawrence Berkeley Nat. Lab. (LBNL), Berkeley, CA, USA,
p. 1617, Jun. 2018, doi: 10.3390/en11071617. May 2021, doi: 10.2172/1784302.
[16] A. K. Ozcanli, F. Yaprakdal, and M. Baysal, ‘‘Deep learning methods [35] C. J. Jena and P. K. Ray, ‘‘Power allocation scheme for grid interactive
and applications for electrical power systems: A comprehensive review,’’ microgrid with hybrid energy storage system using model predictive
Int. J. Energy Res., vol. 44, no. 9, pp. 7136–7157, Jul. 2020, doi: control,’’ J. Energy Storage, vol. 81, Mar. 2024, Art. no. 110401, doi:
10.1002/er.5331. 10.1016/j.est.2023.110401.
[17] M. S. Ibrahim, W. Dong, and Q. Yang, ‘‘Machine learning driven [36] M. Colak, I. Cetinbas, and M. Demirtas, ‘‘Fuzzy logic and artificial neural
smart electric power systems: Current trends and new perspec- network based grid-interactive systems for renewable energy sources: A
tives,’’ Appl. Energy, vol. 272, Aug. 2020, Art. no. 115237, doi: review,’’ in Proc. 9th Int. Conf. Smart Grid (icSmartGrid), Jun. 2021,
10.1016/j.apenergy.2020.115237. pp. 186–191, doi: 10.1109/icSmartGrid52357.2021.9551219.
[18] N. Voropai, ‘‘Electric power system transformations: A review of main [37] M. Ahmed, L. Meegahapola, A. Vahidnia, and M. Datta, ‘‘Stabil-
prospects and challenges,’’ Energies, vol. 13, no. 21, p. 5639, Oct. 2020, ity and control aspects of microgrid architectures—A comprehen-
doi: 10.3390/en13215639. sive review,’’ IEEE Access, vol. 8, pp. 144730–144766, 2020, doi:
[19] L. Cheng and T. Yu, ‘‘A new generation of AI: A review and perspective 10.1109/ACCESS.2020.3014977.
on machine learning technologies applied to smart energy and electric [38] M. Sedighizadeh, S. S. Fazlhashemi, H. Javadi, and M. Tagh-
power systems,’’ Int. J. Energy Res., vol. 43, no. 6, pp. 1928–1973, vaei, ‘‘Multi-objective day-ahead energy management of a microgrid
May 2019, doi: 10.1002/er.4333. considering responsive loads and uncertainty of the electric vehi-
[20] V. K. Saini, R. Kumar, A. S. Al-Sumaiti, S. A., and cles,’’ J. Cleaner Prod., vol. 267, Sep. 2020, Art. no. 121562, doi:
E. Heydarian-Forushani, ‘‘Learning based short term wind speed 10.1016/j.jclepro.2020.121562.
forecasting models for smart grid applications: An extensive review [39] A. Muhtadi, D. Pandit, N. Nguyen, and J. Mitra, ‘‘Distributed energy
and case study,’’ Electric Power Syst. Res., vol. 222, Sep. 2023, resources based microgrid: Review of architecture, control, and reliabil-
Art. no. 109502, doi: 10.1016/j.epsr.2023.109502. ity,’’ IEEE Trans. Ind. Appl., vol. 57, no. 3, pp. 2223–2235, May 2021,
[21] P. Vrablecová, A. B. Ezzeddine, V. Rozinajová, S. Šárik, and doi: 10.1109/TIA.2021.3065329.
A. K. Sangaiah, ‘‘Smart grid load forecasting using online support vector [40] E. I. Come Zebra, H. J. van der Windt, G. Nhumaio, and A. P. C. Faaij,
regression,’’ Comput. Electr. Eng., vol. 65, pp. 102–117, Jan. 2018, doi: ‘‘A review of hybrid renewable energy systems in mini-grids for off-
10.1016/j.compeleceng.2017.07.006. grid electrification in developing countries,’’ Renew. Sustain. Energy Rev.,
[22] H. M. Rouzbahani, H. Karimipour, and L. Lei, ‘‘A review on virtual power vol. 144, Jul. 2021, Art. no. 111036, doi: 10.1016/j.rser.2021.111036.
plant for energy management,’’ Sustain. Energy Technol. Assessments, [41] B. S. Sami, ‘‘Intelligent energy management for off-grid renewable
vol. 47, Oct. 2021, Art. no. 101370, doi: 10.1016/j.seta.2021.101370. hybrid system using multi-agent approach,’’ IEEE Access, vol. 8,
[23] A. K. Ozcanli and M. Baysal, ‘‘Islanding detection in microgrid pp. 8681–8696, 2020, doi: 10.1109/ACCESS.2019.2963584.
using deep learning based on 1D CNN and CNN-LSTM networks,’’ [42] P. Malik, M. Awasthi, and S. Sinha, ‘‘A techno-economic investiga-
Sustain. Energy, Grids Netw., vol. 32, Dec. 2022, Art. no. 100839, doi: tion of grid integrated hybrid renewable energy systems,’’ Sustain.
10.1016/j.segan.2022.100839. Energy Technol. Assessments, vol. 51, Jun. 2022, Art. no. 101976, doi:
[24] S. Karan and H.-G. Yeh, ‘‘Fault classification in microgrids using deep 10.1016/j.seta.2022.101976.
learning,’’ in Proc. IEEE Green Energy Smart Syst. Conf. (IGESSC), [43] J. Liu, H. Hu, S. S. Yu, and H. Trinh, ‘‘Virtual power plant with renewable
Nov. 2020, pp. 1–7, doi: 10.1109/IGESSC50231.2020.9285101. energy sources and energy storage systems for sustainable power grid-
[25] H. Shengren, P. P. Vergara, E. M. Salazar Duque, and P. Palensky, formation, control techniques and demand response,’’ Energies, vol. 16,
‘‘Optimal energy system scheduling using a constraint-aware reinforce- no. 9, p. 3705, Apr. 2023, doi: 10.3390/en16093705.
ment learning algorithm,’’ Int. J. Electr. Power Energy Syst., vol. 152, [44] N. Naval and J. M. Yusta, ‘‘Virtual power plant models and electricity
Oct. 2023, Art. no. 109230, doi: 10.1016/j.ijepes.2023.109230. markets—A review,’’ Renew. Sustain. Energy Rev., vol. 149, Oct. 2021,
[26] C. Wang, J. Zhang, A. Wang, Z. Wang, N. Yang, Z. Zhao, C. S. Lai, and L. Art. no. 111393, doi: 10.1016/j.rser.2021.111393.
L. Lai, ‘‘Prioritized sum-tree experience replay TD3 DRL-based online [45] E. A. Bhuiyan, M. Z. Hossain, S. M. Muyeen, S. R. Fahim,
energy management of a residential microgrid,’’ Appl. Energy, vol. 368, S. K. Sarker, and S. K. Das, ‘‘Towards next generation virtual power
Aug. 2024, Art. no. 123471, doi: 10.1016/j.apenergy.2024.123471. plant: Technology review and frameworks,’’ Renew. Sustain. Energy
[27] R. Machlev, L. Heistrene, M. Perl, K. Y. Levy, J. Belikov, Rev., vol. 150, Oct. 2021, Art. no. 111358, doi: 10.1016/j.rser.2021.
S. Mannor, and Y. Levron, ‘‘Explainable artificial intelligence (XAI) 111358.
techniques for energy and power systems: Review, challenges and [46] L. Yu, W. Xie, D. Xie, Y. Zou, D. Zhang, Z. Sun, L. Zhang, Y. Zhang,
opportunities,’’ Energy AI, vol. 9, Aug. 2022, Art. no. 100169, doi: and T. Jiang, ‘‘Deep reinforcement learning for smart home energy
10.1016/j.egyai.2022.100169. management,’’ IEEE Internet Things J., vol. 7, no. 4, pp. 2751–2762,
[28] M. M. Hosseini and M. Parvania, ‘‘Artificial intelligence for resilience Apr. 2020, doi: 10.1109/JIOT.2019.2957289.
enhancement of power distribution systems,’’ Electr. J., vol. 34, no. 1, [47] E. Giglio, G. Luzzani, V. Terranova, G. Trivigno, A. Niccolai, and
Jan. 2021, Art. no. 106880, doi: 10.1016/j.tej.2020.106880. F. Grimaccia, ‘‘An efficient artificial intelligence energy management
[29] A. Razmjoo, R. Shirmohammadi, A. Davarpanah, F. Pourfayaz, and system for urban building integrating photovoltaic and storage,’’ IEEE
A. Aslani, ‘‘Stand-alone hybrid energy systems for remote area Access, vol. 11, pp. 18673–18688, 2023, doi: 10.1109/ACCESS.2023.
power generation,’’ Energy Rep., vol. 5, pp. 231–241, Nov. 2019, doi: 3247636.
10.1016/j.egyr.2019.01.010. [48] F. Smarra, A. Jain, T. de Rubeis, D. Ambrosini, A. D’Innocenzo,
[30] L. W. Chong, Y. W. Wong, R. K. Rajkumar, R. K. Rajkumar, and D. Isa, and R. Mangharam, ‘‘Data-driven model predictive control using
‘‘Hybrid energy storage systems and control strategies for stand-alone random forests for building energy optimization and climate
renewable energy power systems,’’ Renew. Sustain. Energy Rev., vol. 66, control,’’ Appl. Energy, vol. 226, pp. 1252–1272, Sep. 2018, doi:
pp. 174–189, Dec. 2016, doi: 10.1016/j.rser.2016.07.059. 10.1016/j.apenergy.2018.02.126.
[31] S. Vosen, ‘‘Hybrid energy storage systems for stand-alone electric power [49] S. Yang, M. P. Wan, W. Chen, B. F. Ng, and S. Dubey, ‘‘Model predictive
systems: Optimization of system performance and cost through control control with adaptive machine-learning-based model for building energy
strategies,’’ Int. J. Hydrogen Energy, vol. 24, no. 12, pp. 1139–1156, efficiency and comfort optimization,’’ Appl. Energy, vol. 271, Aug. 2020,
Dec. 1999, doi: 10.1016/s0360-3199(98)00175-x. Art. no. 115147, doi: 10.1016/j.apenergy.2020.115147.
[32] H. Rezk and G. M. Dousoky, ‘‘Technical and economic analysis of dif- [50] S. Ikeda and T. Nagai, ‘‘A novel optimization method combining
ferent configurations of stand-alone hybrid renewable power systems— metaheuristics and machine learning for daily optimal operations in
A case study,’’ Renew. Sustain. Energy Rev., vol. 62, pp. 941–953, building energy and storage systems,’’ Appl. Energy, vol. 289, May 2021,
Sep. 2016, doi: 10.1016/j.rser.2016.05.023. Art. no. 116716, doi: 10.1016/j.apenergy.2021.116716.

109998 VOLUME 12, 2024

A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

[51] A. Mathew, M. J. Jolly, and J. Mathew, ‘‘Improved residential [68] X. C. Miow, Y. S. Lim, L. C. Hau, J. Wong, and H. Patsios, ‘‘Demand
energy management system using priority double deep Q-learning,’’ response for frequency regulation with neural network load controller
Sustain. Cities Soc., vol. 69, Jun. 2021, Art. no. 102812, doi: under high intermittency photovoltaic systems,’’ Energy Rep., vol. 9,
10.1016/j.scs.2021.102812. pp. 2869–2880, Dec. 2023, doi: 10.1016/j.egyr.2023.01.099.
[52] F. Pallonetto, C. Jin, and E. Mangina, ‘‘Forecast electricity demand in [69] S. Basnet, K. Deschinkel, L. Le Moyne, and M. C. Péra, ‘‘A review on
commercial building with machine learning models to enable demand recent standalone and grid integrated hybrid renewable energy systems:
response programs,’’ Energy AI, vol. 7, Jan. 2022, Art. no. 100121, doi: System optimization and energy management strategies,’’ Renew. Energy
10.1016/j.egyai.2021.100121. Focus, vol. 46, pp. 103–125, Sep. 2023, doi: 10.1016/j.ref.2023.06.001.
[53] M. Mansoor, F. Grimaccia, S. Leva, and M. Mussetta, ‘‘Comparison [70] S. Sankarananth, M. Karthiga, and D. P. Bavirisetti, ‘‘AI-enabled
of echo state network and feed-forward neural networks in electrical metaheuristic optimization for predictive management of renewable
load forecasting for demand response programs,’’ Math. Comput. Simul., energy production in smart grids,’’ Energy Rep., vol. 10, pp. 1299–1312,
vol. 184, pp. 282–293, Jun. 2021, doi: 10.1016/j.matcom.2020.07.011. Nov. 2023, doi: 10.1016/j.egyr.2023.08.005.
[54] X. J. Luo and L. O. Oyedele, ‘‘Forecasting building energy consumption: [71] M. Thirunavukkarasu, Y. Sawle, and H. Lala, ‘‘A comprehensive
Adaptive long-short term memory neural networks driven by genetic review on optimization of hybrid renewable energy systems using
algorithm,’’ Adv. Eng. Informat., vol. 50, Oct. 2021, Art. no. 101357, doi: various optimization techniques,’’ Renew. Sustain. Energy Rev., vol. 176,
10.1016/j.aei.2021.101357. Apr. 2023, Art. no. 113192, doi: 10.1016/j.rser.2023.113192.
[55] N. Mughees, S. A. Mohsin, A. Mughees, and A. Mughees, ‘‘Deep [72] F. Monfaredi, H. Shayeghi, and P. Siano, ‘‘Multi-agent deep reinforce-
sequence to sequence bi-LSTM neural networks for day-ahead peak load ment learning-based optimal energy management for grid-connected
forecasting,’’ Expert Syst. Appl., vol. 175, Aug. 2021, Art. no. 114844, multiple energy carrier microgrids,’’ Int. J. Electr. Power Energy Syst.,
doi: 10.1016/j.eswa.2021.114844. vol. 153, Nov. 2023, Art. no. 109292, doi: 10.1016/j.ijepes.2023.109292.
[56] F. Z. Abera and V. Khedkar, ‘‘Machine learning approach electric [73] H. Gao, T. Jin, C. Feng, C. Li, Q. Chen, and C. Kang, ‘‘Review of virtual
appliance consumption and peak demand forecasting of residential power plant operations: Resource coordination and multidimensional
customers using smart meter data,’’ Wireless Pers. Commun., vol. 111, interaction,’’ Appl. Energy, vol. 357, Mar. 2024, Art. no. 122284, doi:
no. 1, pp. 65–82, Mar. 2020, doi: 10.1007/s11277-019-06845-6. 10.1016/j.apenergy.2023.122284.
[57] Z. Zhu, M. Zhou, F. Hu, S. Wang, J. Ma, B. Gao, K. Bian, and W. Lai, [74] M. Ghiasi, T. Niknam, Z. Wang, M. Mehrandezh, M. Dehghani, and
‘‘A day-ahead industrial load forecasting model using load change rate N. Ghadimi, ‘‘A comprehensive review of cyber-attacks and defense
features and combining FA-ELM and the AdaBoost algorithm,’’ Energy mechanisms for improving security in smart grid energy systems: Past,
Rep., vol. 9, pp. 971–981, Dec. 2023, doi: 10.1016/j.egyr.2022.12.044. present and future,’’ Electric Power Syst. Res., vol. 215, Feb. 2023,
Art. no. 108975, doi: 10.1016/j.epsr.2022.108975.
[58] M. Nakıp, O. Çopur, E. Biyik, and C. Güzeliş, ‘‘Renewable
[75] Y. Duan and Y. Zhang, ‘‘Enhancing smart grid security: A novel approach
energy management in smart home environment via forecast
for efficient attack detection using SMART framework,’’ Meas. Sensors,
embedded scheduling based on recurrent trend predictive neural
vol. 32, Apr. 2024, Art. no. 101015, doi: 10.1016/j.measen.2023.101015.
network,’’ Appl. Energy, vol. 340, Jun. 2023, Art. no. 121014, doi:
10.1016/j.apenergy.2023.121014. [76] O. Majeed Butt, M. Zulqarnain, and T. Majeed Butt, ‘‘Recent advance-
ment in smart grid technology: Future prospects in the electrical power
[59] A. Petrucci, G. Barone, A. Buonomano, and A. Athienitis, ‘‘Modelling
network,’’ Ain Shams Eng. J., vol. 12, no. 1, pp. 687–695, Mar. 2021, doi:
of a multi-stage energy management control routine for energy demand
10.1016/j.asej.2020.05.004.
forecasting, flexibility, and optimization of smart communities using
[77] V. Prema, M. S. Bhaskar, D. Almakhles, N. Gowtham, and K. U. Rao,
a recurrent neural network,’’ Energy Convers. Manage., vol. 268,
‘‘Critical review of data, models and performance metrics for wind and
Sep. 2022, Art. no. 115995, doi: 10.1016/j.enconman.2022.115995.
solar power forecast,’’ IEEE Access, vol. 10, pp. 667–688, 2022, doi:
[60] S. Singh and A. Yassine, ‘‘Mining energy consumption behavior patterns 10.1109/ACCESS.2021.3137419.
for households in smart grid,’’ IEEE Trans. Emerg. Topics Comput., vol. 7,
[78] Z. M. Omer and H. Shareef, ‘‘An adjustable machine learning gradient
no. 3, pp. 404–419, Jul. 2019, doi: 10.1109/TETC.2017.2692098.
boosting-based controller for PV applications,’’ Intell. Syst. Appl., vol. 19,
[61] M. Amayri, C. S. Silva, H. Pombeiro, and S. Ploix, ‘‘Flexibility Sep. 2023, Art. no. 200261, doi: 10.1016/j.iswa.2023.200261.
characterization of residential electricity consumption: A machine [79] T. Yu, H. Z. Wang, B. Zhou, K. W. Chan, and J. Tang, ‘‘Multi-agent
learning approach,’’ Sustain. Energy, Grids Netw., vol. 32, Dec. 2022, correlated equilibrium Q(λ) learning for coordinated smart generation
Art. no. 100801, doi: 10.1016/j.segan.2022.100801. control of interconnected power grids,’’ IEEE Trans. Power Syst., vol. 30,
[62] A. Haghighatseresht, R. MansouriBidekani, S. Razavi, A. Aslani, and no. 4, pp. 1669–1679, Jul. 2015, doi: 10.1109/TPWRS.2014.2357079.
R. Zahedi, ‘‘Investigating the impact of building local photovoltaic [80] L. Ren, Y. Sun, J. Cui, and L. Zhang, ‘‘Bearing remaining useful life pre-
power plants on the national grid, an artificial intelligence approach,’’ diction based on deep autoencoder and deep neural networks,’’ J. Manuf.
Ain Shams Eng. J., vol. 14, no. 11, Nov. 2023, Art. no. 102518, doi: Syst., vol. 48, pp. 71–77, Jul. 2018, doi: 10.1016/j.jmsy.2018.04.008.
10.1016/j.asej.2023.102518. [81] G. Terrén-Serrano and M. Martínez-Ramón, ‘‘Deep learning for
[63] Y. He, F. Luo, M. Sun, and G. Ranzi, ‘‘Privacy-preserving and hierarchi- intra-hour solar forecasting with fusion of features extracted from
cally federated framework for short-term residential load forecasting,’’ infrared sky images,’’ Inf. Fusion, vol. 95, pp. 42–61, Jul. 2023, doi:
IEEE Trans. Smart Grid, vol. 14, no. 6, pp. 4409–4423, Nov. 2023, doi: 10.1016/j.inffus.2023.02.006.
10.1109/TSG.2023.3268633. [82] A. Dridi, H. Afifi, H. Moungla, and J. Badosa, ‘‘A novel deep reinforce-
[64] L. Hao and Y. Xu, ‘‘Semisupervised learning-based occupancy esti- ment approach for IIoT microgrid energy management systems,’’ IEEE
mation for real-time energy management using ambient data,’’ IEEE Trans. Green Commun. Netw., vol. 6, no. 1, pp. 148–159, Mar. 2022, doi:
Internet Things J., vol. 10, no. 20, pp. 18426–18437, Oct. 2023, doi: 10.1109/TGCN.2021.3112043.
10.1109/jiot.2023.3280361. [83] M. H. Zafar, S. M. S. Bukhari, M. Abou Houran, S. K. R. Moosavi,
[65] M. Khan, B. N. Silva, O. Khattab, B. Alothman, and C. Joumaa, M. Mansoor, N. Al-Tawalbeh, and F. Sanfilippo, ‘‘Step towards
‘‘A transfer reinforcement learning framework for smart home energy secure and reliable smart grids in industry 5.0: A federated learning
management systems,’’ IEEE Sensors J., vol. 23, no. 4, pp. 4060–4068, assisted hybrid deep learning model for electricity theft detection using
Feb. 2023, doi: 10.1109/JSEN.2022.3218840. smart meters,’’ Energy Rep., vol. 10, pp. 3001–3019, Nov. 2023, doi:
[66] A. A. Amer, K. Shaban, and A. M. Massoud, ‘‘DRL-HEMS: Deep 10.1016/j.egyr.2023.09.100.
reinforcement learning agent for demand response in home energy [84] F. Gallego, C. Martín, M. Díaz, and D. Garrido, ‘‘Maintaining flexibility
management systems considering customers and operators perspectives,’’ in smart grid consumption through deep learning and deep reinforce-
IEEE Trans. Smart Grid, vol. 14, no. 1, pp. 239–250, Jan. 2023, doi: ment learning,’’ Energy AI, vol. 13, Jul. 2023, Art. no. 100241, doi:
10.1109/TSG.2022.3198401. 10.1016/j.egyai.2023.100241.
[67] S. Pelekis, A. Pipergias, E. Karakolis, S. Mouzakitis, F. Santori, [85] U. Arul, R. Gnanajeyaraman, A. Selvakumar, S. Ramesh, T. Manikandan,
M. Ghoreishi, and D. Askounis, ‘‘Targeted demand response for flexible and G. Michael, ‘‘Integration of IoT and edge cloud computing
energy communities using clustering techniques,’’ Sustain. Energy, Grids for smart microgrid energy management in VANET using machine
Netw., vol. 36, Dec. 2023, Art. no. 101134, doi: 10.1016/j.segan.2023. learning,’’ Comput. Electr. Eng., vol. 110, Sep. 2023, Art. no. 108905,
101134. doi: 10.1016/j.compeleceng.2023.108905.

VOLUME 12, 2024 109999

 A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

[86] M. Önder, M. U. Dogan, and K. Polat, ‘‘Classification of smart grid [101] W. H. Tee, C. K. Gan, and J. Sardi, ‘‘Benefits of energy storage
stability prediction using cascade machine learning methods and the systems and its potential applications in Malaysia: A review,’’ Renew.
Internet of Things in smart grid,’’ Neural Comput. Appl., vol. 35, no. 24, Sustain. Energy Rev., vol. 192, Mar. 2024, Art. no. 114216, doi:
pp. 17851–17869, Aug. 2023, doi: 10.1007/s00521-023-08605-x. 10.1016/j.rser.2023.114216.
[87] S. Banerjee and P. S. Bhowmik, ‘‘A machine learning approach based [102] J. Sardi, N. Mithulananthan, M. M. Islam, and C. K. Gan, ‘‘Framework
on decision tree algorithm for classification of transient events in of virtual microgrids formation using community energy storage in
microgrid,’’ Electr. Eng., vol. 105, no. 4, pp. 2083–2093, Aug. 2023, doi: residential networks with rooftop photovoltaic units,’’ J. Energy Storage,
10.1007/s00202-023-01796-5. vol. 35, Mar. 2021, Art. no. 102250, doi: 10.1016/j.est.2021.102250.
[88] B. Feng, Z. Liu, G. Huang, and C. Guo, ‘‘Robust federated deep [103] A. G. Olabi, ‘‘Application of artificial intelligence for prediction,
reinforcement learning for optimal control in multiple virtual power optimization, and control of thermal energy storage systems,’’ Thermal
plants with electric vehicles,’’ Appl. Energy, vol. 349, Nov. 2023, Sci. Eng. Prog., vol. 39, Mar. 2024, Art. no. 101730. [Online]. Available:
Art. no. 121615, doi: 10.1016/j.apenergy.2023.121615. https://www.sciencedirect.com/science/article/pii/S2451904923000835
[89] C. Liu, R. J. Yang, X. Yu, C. Sun, and G. Rosengarten, ‘‘Supporting [104] T. Ouyang, M. Zhang, P. Qin, and X. Tan, ‘‘Flow battery energy
virtual power plants decision-making in complex urban environments storage system for microgrid peak shaving based on predictive control
using reinforcement learning,’’ Sustain. Cities Soc., vol. 99, Dec. 2023, algorithm,’’ Appl. Energy, vol. 356, Feb. 2024, Art. no. 122448, doi:
Art. no. 104915. Accessed: Jan. 15, 2024. [Online]. Available: 10.1016/j.apenergy.2023.122448.
https://www.sciencedirect.com/science/article/pii/S2210670723005267? [105] K. Bio Gassi and M. Baysal, ‘‘Improving real-time energy decision-
via%3Dihub making model with an actor-critic agent in modern microgrids with
[90] J. Maeng, D. Min, and Y. Kang, ‘‘Intelligent charging and energy storage devices,’’ Energy, vol. 263, Jan. 2023, Art. no. 126105,
discharging of electric vehicles in a vehicle-to-grid system using doi: 10.1016/j.energy.2022.126105.
a reinforcement learning-based approach,’’ Sustain. Energy, Grids [106] L. Driscoll, S. de la Torre, and J. A. Gomez-Ruiz, ‘‘Feature-based
Netw., vol. 36, Dec. 2023, Art. no. 101224, doi: 10.1016/j.segan.2023. lithium-ion battery state of health estimation with artificial neural
101224. networks,’’ J. Energy Storage, vol. 50, Jun. 2022, Art. no. 104584, doi:
[91] J. Jithish, B. Alangot, N. Mahalingam, and K. S. Yeo, ‘‘Dis- 10.1016/j.est.2022.104584.
tributed anomaly detection in smart grids: A federated learning- [107] J. Wu, L. Kong, Z. Cheng, Y. Yang, and H. Zuo, ‘‘RUL prediction for
based approach,’’ IEEE Access, vol. 11, pp. 7157–7179, 2023, doi: lithium batteries using a novel ensemble learning method,’’ Energy Rep.,
10.1109/ACCESS.2023.3237554. vol. 8, pp. 313–326, Nov. 2022, doi: 10.1016/j.egyr.2022.10.298.
[92] K. Aurangzeb, S. Aslam, S. I. Haider, S. M. Mohsin, S. U. Islam, [108] J. Li, X. Huang, X. Tang, J. Guo, Q. Shen, Y. Chai, W. Lu,
H. A. Khattak, and S. Shah, ‘‘Energy forecasting using multiheaded T. Wang, and Y. Liu, ‘‘The state-of-charge predication of lithium-ion
convolutional neural networks in efficient renewable energy resources battery energy storage system using data-driven machine learning,’’
equipped with energy storage system,’’ Trans. Emerg. Telecommun. Sustain. Energy, Grids Netw., vol. 34, Jun. 2023, Art. no. 101020, doi:
Technol., vol. 33, no. 2, p. e3837, Feb. 2022, doi: 10.1002/ett.3837. 10.1016/j.segan.2023.101020.
[93] M. A. Houran, ‘‘COA-CNN-LSTM: Coati optimization algorithm- [109] Z. Ni, B. Li, and Y. Yang, ‘‘Deep domain adaptation network for transfer
based hybrid deep learning model for PV/wind power forecasting learning of state of charge estimation among batteries,’’ J. Energy Stor-
in smart grid applications,’’ Appl. Energy, vol. 349, Nov. 2023, age, vol. 61, May 2023, Art. no. 106812, doi: 10.1016/j.est.2023.106812.
Art. no. 121638. Accessed: Jan. 16, 2024. [Online]. Available:
[110] Z. Wen, P. Fang, Y. Yin, G. Królczyk, P. Gardoni, and Z. Li, ‘‘A novel
https://www.sciencedirect.com/science/article/pii/S0306261923010024?
machine learning model for safety risk analysis in flywheel-battery
via%3Dihub
hybrid energy storage system,’’ J. Energy Storage, vol. 49, May 2022,
[94] S. N. Pichika, G. Meganaa, S. G. Rajasekharan, and A. Malapati,
Art. no. 104072, doi: 10.1016/j.est.2022.104072.
‘‘Multi-component fault classification of a wind turbine gearbox
using integrated condition monitoring and hybrid ensemble method [111] H. Kang, S. Jung, H. Kim, J. Jeoung, and T. Hong, ‘‘Reinforcement
approach,’’ Appl. Acoust., vol. 195, Jun. 2022, Art. no. 108814, doi: learning-based optimal scheduling model of battery energy storage
10.1016/j.apacoust.2022.108814. system at the building level,’’ Renew. Sustain. Energy Rev., vol. 190,
Feb. 2024, Art. no. 114054, doi: 10.1016/j.rser.2023.114054.
[95] Y. Lu, ‘‘Predicting surface solar radiation using a hybrid radiative
transfer–machine learning model,’’ Renew. Sustain. Energy Rev., vol. 173, [112] D. Kang, D. Kang, S. Hwangbo, H. Niaz, W. B. Lee, J. J. Liu, and
Mar. 2023, Art. no. 113105. Accessed: Jan. 16, 2024. [Online]. Available: J. Na, ‘‘Optimal planning of hybrid energy storage systems using cur-
https://www.sciencedirect.com/science/article/pii/S1364032122009868? tailed renewable energy through deep reinforcement learning,’’ Energy,
via%3Dihub vol. 284, Dec. 2023, Art. no. 128623, doi: 10.1016/j.energy.2023.128623.
[96] M. M. Elymany and M. A. Enany, ‘‘Hybrid optimized-ANFIS [113] Y. Liu, J. Sun, Y. Shang, X. Zhang, S. Ren, and D. Wang, ‘‘A novel
based MPPT for hybrid microgrid using zebra optimization remaining useful life prediction method for lithium-ion battery based on
algorithm and artificial gorilla troops optimizer,’’ Energy Convers. long short-term memory network optimized by improved sparrow search
Manag., vol. 299, Jan. 2024, Art. no. 117809. [Online]. Available: algorithm,’’ J. Energy Storage, vol. 61, May 2023, Art. no. 106645, doi:
https://www.sciencedirect.com/science/article/pii/S019689042301155X? 10.1016/j.est.2023.106645.
via%3Dihub [114] J. Lee, H. Sun, Y. Liu, and X. Li, ‘‘A machine learning framework for
[97] Z. Guo and Z. Zhuang, ‘‘Accurate and generalizable photovoltaic remaining useful lifetime prediction of Li-ion batteries using diverse
panel segmentation using deep learning for imbalanced datasets,’’ neural networks,’’ Energy AI, vol. 15, Jan. 2024, Art. no. 100319, doi:
Renew. Energy, vol. 219, Dec. 2023, Art. no. 119471, doi: 10.1016/j.egyai.2023.100319.
10.1016/j.renene.2023.119471. [115] M. Li, C. Li, Q. Zhang, W. Liao, and Z. Rao, ‘‘State of charge estimation
[98] Y. Elomari, C. Mateu, M. Marín-Genescà, and D. Boer, ‘‘A data- of Li-ion batteries based on deep learning methods and particle-swarm-
driven framework for designing a renewable energy community optimized Kalman filter,’’ J. Energy Storage, vol. 64, Aug. 2023,
based on the integration of machine learning model with life Art. no. 107191, doi: 10.1016/j.est.2023.107191.
cycle assessment and life cycle cost parameters,’’ Appl. Energy, [116] M. Karthikeyan, D. Manimegalai, and K. RajaGopal, ‘‘Power control
vol. 358, Mar. 2024, Art. no. 122619, doi: 10.1016/j.apenergy.2024. of hybrid grid-connected renewable energy system using machine
122619. learning,’’ Energy Rep., vol. 11, pp. 1079–1087, Jun. 2024, doi:
[99] A. R. D. A. V. Filho, D. F. Moraes, M. V. B. de Aguiar Vallim, L. S. Da 10.1016/j.egyr.2023.12.060.
Silva, and L. A. da Silva, ‘‘A machine learning modeling framework for [117] L. Pinciroli, P. Baraldi, M. Compare, and E. Zio, ‘‘Optimal operation
predictive maintenance based on equipment load cycle: An application and maintenance of energy storage systems in grid-connected microgrids
in a real world case,’’ Energies, vol. 15, no. 10, p. 3724, May 2022, doi: by deep reinforcement learning,’’ Appl. Energy, vol. 352, Dec. 2023,
10.3390/en15103724. Art. no. 121947, doi: 10.1016/j.apenergy.2023.121947.
[100] J. Darville, T. Runsewe, A. Yavuz, and N. Celik, ‘‘Machine [118] Z. Xiao, B. Gao, X. Huang, Z. Chen, C. Li, and Y. Tai, ‘‘An interpretable
learning based simulation for fault detection in microgrids,’’ in horizontal federated deep learning approach to improve short-term
Proc. Winter Simul. Conf. (WSC), Dec. 2022, pp. 701–712, doi: solar irradiance forecasting,’’ J. Cleaner Prod., vol. 436, Jan. 2024,
10.1109/WSC57314.2022.10015473. Art. no. 140585, doi: 10.1016/j.jclepro.2024.140585.

110000 VOLUME 12, 2024

A. Shahid et al.: AI Technologies and Their Applications in Small-Scale Electric Power Systems

[119] A. Bampoulas, F. Pallonetto, E. Mangina, and D. P. Finn, ‘‘A Bayesian TARMO KORÕTKO (Member, IEEE) received
deep-learning framework for assessing the energy flexibility of residential the B.Sc. and M.Sc. degrees in mechatronics
buildings with multicomponent energy systems,’’ Appl. Energy, vol. 348, and the Ph.D. degree in energy and geotech-
Oct. 2023, Art. no. 121576, doi: 10.1016/j.apenergy.2023.121576. nology from Tallinn University of Technology
[120] K. Zhou, N. Peng, H. Yin, and R. Hu, ‘‘Urban virtual power plant (TalTech), Tallinn, Estonia, in 2007, 2010, and
operation optimization with incentive-based demand response,’’ Energy, 2019, respectively. In 2018, he became a member
vol. 282, Nov. 2023, Art. no. 128700, doi: 10.1016/j.energy.2023.128700. of the Microgrids and Metrology Research Group,
[121] Y. Zhang, M. Zhao, and R. Xiong, ‘‘Online data-driven battery life Department of Electrical Power Engineering and
prediction and quick classification based on partial charging data within
Mechatronics, TalTech, where he is currently
10 min,’’ J. Power Sources, vol. 594, Feb. 2024, Art. no. 234007, doi:
employed as a Senior Researcher. He has pub-
10.1016/j.jpowsour.2023.234007.
lished more than 20 articles on the topics of energy management and
control of microgrids, power system digitalization, local energy markets,
energy storage systems, and machine learning applications in electric power
systems. His research interests include microgrids, local energy markets,
prosumers, power system digitalization, and artificial intelligence in electric
power systems.
ARQUM SHAHID (Graduate Student Member,
IEEE) received the B.Sc. degree in electrical
engineering from the University of Engineering
and Technology, Lahore, Pakistan, and the M.Sc.
degree in power systems from Mirpur University ARGO ROSIN (Senior Member, IEEE) received
of Science and Technology, Pakistan. He is the Dipl.Eng., M.Sc., and Dr.Sc.Eng. degrees in
currently pursuing the Ph.D. degree in electrical electrical engineering from Tallinn University of
engineering with Tallinn University of Technology Technology (TUT), Tallinn, Estonia, in 1996,
(TalTech), Estonia. His research interests include 1998, and 2005, respectively. He became a
the role of artificial intelligence in demand-side Professor of power supply, in 2018. He is cur-
management and energy flexibility, crucial areas for the future of energy rently the Head of the Microgrids and Metrology
systems. Group, Department of Electrical Power Engineer-
ing and Mechatronics, TUT. He is the Co-Founder
and a Leading Researcher with the first global
cross-border Smart City Center of Excellence—Finest Twins (2019–2026),
which focuses on developing research and cross-border innovation networks
and capabilities in five domains—data, governance, mobility, energy, and
FREDDY PLAUM received the B.Sc. degree built environment. He is also the Co-Founder of the Estonian Centre
in mechatronics and the M.Sc. degree (cum of Excellence for zero energy and resource-efficient smart buildings and
laude) in energy conversion and control systems districts. He has supervised project-related fields of six Ph.D. theses and over
from Tallinn University of Technology (TalTech), 30 master’s theses. He has published several international (Springer and
Tallinn, Estonia, in 2017 and 2019, respectively. IET) books and more than 80 articles on smart grids and microgrids, energy
He is currently pursuing the Ph.D. degree in elec- management, and control and diagnostic systems. His research interests
trical power engineering and mechatronics with include the energy efficiency of household and industrial systems, energy
TalTech. He has co-authored two book chapters demand-side management, microgrids, and industrial control systems.
related to machine learning teaching materials and He has received several national awards, including the Golden Badge of
authored or co-authored seven published articles the Estonian Ministry of Defence for productive cooperation and The Best
related to his research field. His research interests include developing Development Work 2005 from Tallinn City Government.
quantification methods of aggregated energy flexibility.

VOLUME 12, 2024 110001