Intelligent Diagnostics and Mitigation of Supraharmonics in Converter-Interfaced Smart Grids: Critical Assessment and Future Direction

Intelligent Diagnostics and Mitigation of Supraharmonics in Converter-Interfaced Smart Grids: Critical Assessment and Future Direction

Idris Umar Usman (1)

M.Tech (RE) Research Scholar Mewar University, Gangrar Chittorgarh, Rajasthan, India

Deepak Kumar Joshi (2)

Assistant Professor (Head of Dept.) Mewar University, Gangrar Chittorgarh, Rajasthan, India

Department of Electrical Engineering

Sadisu Idris Saleh (3)

M.Tech (RE) Research Scholar Mewar University, Gangrar Chittorgarh, Rajasthan, India

Abstract

The drastic shift towards low carbon power networks has been driving the integration of converter interfaced technologies, including Solar Photovoltaic Systems, wind energy converters, Electric vehicle, Battery Energy Storage Systems and charging infrastructure. Therefore, a power quality and electromagnetic compatibility (EMC) issue has become an important problem of high frequency perturbations often referred to as supraharmonics ranging from 2 – 150 kHz. These emissions are produced by semiconductor switching, pulse width modulation (PWM) techniques, interactions between components and complex network impedances. This paper gives a detailed analysis of comparative study of conventional and intelligent techniques for supraharmonic analysis, source identification and

Keywords: Supraharmonic, Renewable Energy, Smart Grids, Power Quality, Artificial Intelligence,

1. Introduction

   Today, the power grid is undergoing revolutionary changes due to the growing number of renewable energy resources and loads connected to the power grid via the power electronics [1]. The main components are solar Photovoltaic generators, wind turbine converters, battery energy storage systems, and electric vehicle chargers. On the other hand, new digital paradigm tools like the Internet of Things (IoT), Edge Computing, Digital Twins and Artificial Intelligence (AI) are being increasingly used to monitor grid health, classify power quality disturbances, and take automatic control actions [1].

   Smart grid features enhance system flexibility, observability and efficiency but the significant number of converter interfaced devices brings in power quality issues which are complex and mainly centred on low order integer multiples of the fundamental power frequency 50 – 60Hz, typically

   mitigation for renewable energy integrated smart grid. Signal processing tools, machine learning (ML) and artificial intelligence (AI) methodologies, active power filters (APF), hybrid architectures, adaptive control, and intelligent inverter modulation strategies are critically analysed with regard to parameters like computational complexity, real time applicability and robustness. Key findings include the fact that current research is fragmented, with numerous gaps in standardized monitoring (2- 150 kHz), dynamic multi source tracking, representative AI training sets and real time coordinated converter control. A single, intelligent end to end solution that integrates advanced signal processing, predictive AI diagnostics, and adaptive mitigation is proposed to address these gaps, improve operational resilience, and provide comprehensive observability in converter dominated power systems.

   Machine Learning, Signal Processing, Active Power Filter, Converter Interaction.

   in the range of the 40th and 50th harmonic up to 2 kHz. Today, however, the switching frequency of the power electronic converters is in the kHz range, from several kilohertz to tens of kilohertz [2].

   The voltage and current deviations in the frequency range of 2150kHz are considered as supraharmonics [3]. Supraharmonic emissions are observed in both the narrowband and broadband noise. Narrowband emission can be seen as discrete switching frequency and sideband emissions, while the broadband noise can be seen in wider spectra [4]. The high frequency emissions have adverse impacts such as thermal stress and premature failure of grid connected equipment, degradation of insulation in cables and transformers, high power losses, acoustic noises and malfunction or disruption of Power Line Communication (PLC) systems [5].

   Although the topics of isolated parts of supraharmonics have been the subject of recent

   research, the picture is fragmented [3]. The synthesis of these sources, measurement needs, signal processing capabilities, AI based source identification and adaptive mitigation of the sources under dynamic multi converter interactions is a highly desirable system that is needed urgently. This review critically analyses traditional and intelligent techniques highlighting accuracy, real time support, computational load, adaptability and viability for implementation in the grid [3].

2. Smart Grid Architecture & Power Electronics Interactions

   Smart grids integrate advanced sensing, communication, automation, and distributed energy resources to enable the optimal operation of the power system [6]. The converter interfaces which are a key component for grid connected operations, however, are the major sources of supraharmonic disturbance [7].

   1. Renewable Energy Integration

      Systems using renewable energy use a variety of converter topologies (AC/DC, DC/DC, and DC/AC) to process power. These power electronic interfaces feature semiconductor switching devices with high frequency emissions that are influenced by switching frequency, control topology, pulse width modulation (PWM) techniques, output filter design and dynamic grid impedance are illustrated [8].

   2. Inverter Topologies and Modulation

      Inverters can be used in the grid following or grid forming [9]. mode Switching pulses for the semiconductor devices are controlled by PWM schemes such as Sinusoidal Pulse Width Modulation (SPWM), Space Vector Pulse Width Modulation (SVPWM), Discontinuous Pulse Width Modulation (DPWM) and Variable Frequency Pulse Width Modulation (VFPWM) [9].

      The high frequency spectrum is directly determined by parameters like carrier frequency, modulation index, dead time, switching pattern, etc. [8]. If several converters are working simultaneously, mutual coupling caused by the grid impedance can lead to resonance amplification and cross talk, making emission characterisation very complicated [10].

3. Fundamentals of Supraharmonics

   The supraharmonic spectra 2-150 kHz are dominated by switching frequency components, harmonic sidebands, interharmonics, resonant peaks and broadband noise [11]. The emissions from the individual device converter, self-generated by its internal operation. Emissions that come from other emissions sources in the network and pass-through a given device because of the low high frequency input impedance.

   Propagation of supraharmonics through network components such as capacitors, cables, transformers, filters is highly non-stationary and depends on local system impedance, system topology, dynamic loading and interactions between parallel converters [3], [12].

4. Standard Frameworks and Measurement Approaches

   1. Assessment Standards

      Standardisation in the 2150 kHz range is evolving. Table 1 summarises key international standards and guidance documents

      Ref	Standar d/ Guideli ne	Frequenc y Focus	Main Purpose	Existing Limitations
      [1 3]	IEC 61000 Series	Power Quality/p> / EMC	Measure ment methods and compatibi lity levels	Limited formal regulator y framewor ks above 9 kHz
      [1 4]	IEEE 519- 2022	Power System Harmo nics	Recomme nded harmonic limits	Focused predomin antly on conventio nal low order harmonic s up to 2 kHz

      Table 1: Comparison of Standards and Guidelines for Supraharmonic Assessment

      [1 5]	CIGR E Techn ical Papers	>2 kHz emissio ns	Technical assessmen t and field measurem ent guidance	Informati ve guidance; not an enforceab le regulator y standard
      [1 6]	CISP R Standa rds	EMC and High Freque ncy	Equipmen t emission limits	Device- level orientatio n; lacks network interactio n perspecti ve

5. Source Classification of supraharmonic

   Modern power networks are associated with many types of supraharmonic sources, which have different mechanisms and impacts, are classified in Table 2 [17].

   Source /Ref	Primary Mechanis m	Typical Impact	Key Research Challeng e
   PV Inverter s, [18]	High frequenc y switching and PWM patterns	Current and voltage supraharmo nics	High dependen ce on variable solar irradiance and operating state
   Wind Energy Convert ers, [19]	Multi stage converter switching and control interactio n	High frequency voltage and current distortion	Dynamic converter grid resonance and impedanc e variation

   Table 2: Comparative Analysis of Supraharmonic Sources

   BESS Convert ers, [20]	Bidirecti onal AC/DC and DC/DC switching	Conducted emissions across modes	Shift in emissions during charge versus discharge modes
   Electric Vehicle Charger [5]	High power AC/DC and DC/DC stage conversio n	High magnitude supraharmo nic current injection	Cumulati ve impacts of simultane ous multi vehicle charging
   Parallel Convert ers [21]	Inter converter resonanc e and mutual coupling	Amplificati on of background high frequency distortion	Complex interactio n through varying network impedanc e
   LEDs & SMPS Loads, [3]	Mass distribute d low power switching supplies	Collective broadband conducted emissions	Aggregat ed impact of massive numbers of distribute d units
   PLC Devices [3]	Intention al high frequenc y carrier signal injection	Discrete narrowband voltage and current emissions	Cross talk, signal attenuatio n, and impedanc e mismatch

6. Intelligent Classification Method

   The Signal Processing and Intelligent Diagnostic Techniques. However, the Fast Fourier Transform (FFT) is not suitable for processing supraharmonic signals, which are often non-stationary, intermittent and dynamic [22]. The use of advanced signal processing in conjunction with machine learning (ML) and deep learning (DL) brings improved diagnostic capability [23].

   Table 3 compares the diagnostic performance of the signal processing, classical ML and deep learning architectures.

   Techniq ue	Principal Strengths	Primary Limitation s	Preferred Grid Applicatio n
   FFT [24]	Simple, low computati onal overhead	Fails on non- stationary signals, no time resolution	Stationary, steady state baseline spectral analysis
   STFT [22]	Provides time frequency localised informatio n	Fixed resolution trade-off determine d by windowin g	Time varying, non- stationary signal analysis
   Wavele t, [25]	Multi resolution analysis for fast transients	High computati onal load and parameter sensitivity	Transient detection and non- stationary profiling
   S- Transfo rm, [26]	Excellent phase and time frequency tracking	High computati onal complexit y	High precision offline and power quality characteris ation
   ANN, [27]	Strong non-linear modelling capability	Sensitive to training data and black box nature	Automated pattern detection and event classificati on
   CNN, [24]	Automatic feature extraction from spectrogra ms	Requires large high quality labelled datasets	Image based automated and time frequency classificati on

   Table 3: Comparative Assessment of Supraharmonic Detection and Diagnostic Techniques

   ANFIS, [28]	Combines neural learning with fuzzy logic reasoning	High training complexit y under high feature dimension s	Adaptive diagnosis under uncertain system models
   Deep Learnin g (ResNet ), [29]	Superior feature extraction and accuracy	Requires high computati onal resources and big training data	High density grid monitoring and automated analytics

7. Supraharmonic Mitigation Methods

   The range of mitigation strategies includes passive filtering, active power filters (APF), hybrid configurations and converter-level control modifications [21]. A comparison of these mitigation strategies is given in Table 4.

   Fig. 1. classification of Power Filters [30].

   Table 4. The Table Below Compares the Different Types of Supraharmonic Mitigation Techniques.

   Topology/ Ref.	Structural Strengths	Key Operation al Limitatio ns	Practical Grid Applicab ility
   Passive Filters, [31]	Low cost, high reliability and simple design	Fixed tuning and potential resonance with grid	Static and predictab le single frequenc y

   		impedanc e	emission s
   Active Power Filters, [32]	Dynamic compensa tion and adaptive response	Higher hardware cost and limited bandwidt h	Variable dynamic emission s and localized mitigatio n
   Hybrid Active Filters, [33]	Reduced APF rating and superior HF attenuatio n	Design complexit y and tuning interactio ns	Medium to high voltage RE grid connecti on points
   Converter Control Mitigation , [34]	Software based, zero added hardware cost	Requires accurate modelling , grid impedanc e dependent	Embedde d inverter modulati on and firmware updates
   Adaptive Control, [34]	Tracks dynamic grid impedanc e variations	Control stability proof required under transients	Variable RE penetrati on microgri ds and weak grids
   AI-Based Control, [23]	Predictive mitigatio n and self- optimisin g	Requires extensive training, Explainab ility challenge s	Next generatio n smart inverters and edge controlle rs

   Techniques	Technical Characteristics and Main Comparison
   Artificial Neural Networks, [29]	Good non-linear pattern recognition, moderate computational load and requires clean training datasets.
   Support Vector Machines [23]	High classification accuracy on small datasets and performance depends on kernel hyperparameter tuning.
   Decision Trees, [24]	Computationally fast and interpretable and prone to overfitting under noisy conditions.
   Random Forest [35]	High resistance to noise and overfitting and increased computational memory footprint.
   Extreme Learning Machine, [36]	Ultrafast training time and highly sensitive to feature selection and input data quality.
   Convolutional Networks, [36]	Superior automated feature extraction from 2D spectral images and high computational demands.
   Transfer Learning, [37]	Excellent generalization across domain shifts and high resource footprint during training.
   Hybrid AI and DSP Frameworks [23]	Combines DSP feature extraction with AI classification and yields an optimal balance of speed and precision.
   AI Inverter Control [38]	Enhances grid stability through adaptive control and harmonic reduction and dynamic response to grid fluctuations.

8. Comparative Synthesis of AI Techniques

   Intelligent algorithms offer distinct performance trade-offs with respect to smart grid deployment parameters [35].

   Table 5: Synthesis of AI Diagnostic and Control Techniques

   Grid Scalability

   Low and localized hardware bound installations

   High and scalable via distributed edge computing

9. Performance Evaluation of Conventional and Intelligent Techniques

   Table 6: show the Performance evaluation comparing conventional and intelligent techniques across operational power system metrics.

   Performance criterion	Conventional Techniques	Intelligent and AI- Driven Techniques
   Accuracy	High for stationary, continuous signals	Superior for complex and time-varying emission patterns
   Computational cost	Low to moderate and deterministic	Moderate to high during training, low during inference
   Real-Time Capability	Well established and low latency	Fast post training, depends on hardware execution
   System Adaptability	Poor and requires manual component redesign	Excellent and continuously retrainable and self- tuning
   Noise Robustness	Susceptible to dynamic grid background noise	High and capable of learning under noisy backgrounds
   Data Requirement	Minimal and relies on direct analytical formulas	Moderate to extensive and requires significant training data
   Interpretability	High and explicit deterministic mathematical basis	Lower and machine learning models present black box challenges

   Table 6: Comparative Evaluation of Conventional and Intelligent

10. Future Research Directions

    To transition supraharmonic management from reactive filtering to proactive, real-time control, future research should prioritize:

    1. Coordinated Multi Inverter Modulation: Implementing distributed reinforcement learning (DRL) to adaptively shift 1. Unified Monitoring Frameworks: Creating consistent, ongoing monitoring guidelines for distribution infrastructures in the 2150 kHz range.

    2. Edge-AI Analytics: For low latency, real time source identification, lightweight machine learning models are directly applied to smart meters and inverter controllers.

    3. Digital Twins for High Frequency Grid Modelling: Creating dynamic digital twins to simulate cross talk between many converters and anticipate resonance events prior to control instability.

    4. carrier frequencies across interconnected PV and storage inverters, preventing cumulative spectral peaks.

11. Conclusion

Supraharmonics 2150 kHz represent an important power quality and EMC challenge in converter- dominated smart grids [17]. Traditional fixed passive filtering approaches are increasingly inadequate for dynamic, non-stationary grid operating states. While advanced signal processing (Wavelets, S-Transform) provides robust feature extraction, integrating AI and ML techniques enables real-time disturbance classification, dynamic source tracking, and predictive mitigation

Authors’ Contribution

All authors contributed to the interpretation of the reviewed literature, critically revised the paper, approved the final version for publication and agreed to be held responsible for the entire contents. Acknowledgments

This publication was the result of research conducted without external funding.

Conflicts of Interest

There are no conflicts of interest for the authors to disclose.

Generative AI Statement

Throughout the creation of this work, the authors used Quillbot and ChatGPT to rephrase the content and correct grammar. The authors then reviewed and edited the text and accept sole responsibility for the content of this publication.

References

1. M. Çolak and E. Irmak, A State-of-the-Art Review on Electric Power Systems and Digital Transformation, Electric Power Components and Systems, vol. 51, no. 11, pp. 10891112, Jul. 2023, doi:

   10.1080/15325008.2023.2189760.

2. J. Kaur and S. K. Bath, Harmonic distortion in power systems due to electronic control and renewable energy integration: a comprehensive review, Discov Electron, vol. 2, no. 1, p. 67, Aug. 2025, doi: 10.1007/s44291-025-00111-9.

3. S. Rajkumar, R. Balasubramanian, and P. Kathirvelu, A Comprehensive Review on SupraharmonicsThe Next Big Power Quality Concern, Smart Grids and Energy, vol. 9, no. 1, p. 15, Feb. 2024, doi: 10.1007/s40866-024-00195-4.

4. S. Li, Y. Xu, S. Wang, H. Wang, and B. Li, Analysis and modeling of supraharmonic coupling among multiple inverters in renewable energy systems, Microelectronics International, vol. 43, no. 2, pp. 6273, Apr.

   2026, doi: 10.1108/MI-07-2025-0122.

5. S. Rajkumar, R. Balasubramanian, and K. P. Kathirvelu, A comparative analysis of supraharmonic emission by DC fast chargers used for electric vehicle charging, Sci Rep, vol. 15, no. 1, p. 7992, Mar. 2025, doi: 10.1038/s4598-025-92049-x.

6. J. Powell, A. McCafferty-Leroux, W. Hilal, and S. A. Gadsden, Smart Grids: A Comprehensive Survey of Challenges, Industry Applications, and Future Trends, 2024, arXiv. doi: 10.48550/ARXIV.2401.13105.

7. M. Khalid, Smart grids and renewable energy systems: Perspectives and grid integration challenges, Energy Strategy Reviews, vol. 51, p. 101299, Jan. 2024, doi: 10.1016/j.esr.2024.101299.

8. B. Sharma, S. Manna, V. Saxena, P. K. Raghuvanshi, M. H. Alsharif, and M.-K. Kim, A comprehensive review of multi-level inverters, modulation, and control for grid- interfaced solar PV systems, Sci Rep, vol. 15, no. 1, p. 661, Jan. 2025, doi: 10.1038/s41598-

   024-84296-1.

9. H. Pishbahar, F. Blaabjerg, and H. Saboori, Emerging grid-forming power converters for renewable energy and storage resources integration A review, Sustainable Energy Technologies and Assessments, vol. 60, p. 103538, Dec. 2023, doi:

   10.1016/j.seta.2023.103538.

10. P. K. Sharma, P. Singh, S. C. Choube, and L.

    S. Titare, Enhancement of Power Quality in Photovoltaic Systems for Weak Grid Connections, Energies, vol. 18, no. 15, p. 4066, Jul. 2025, doi: 10.3390/en18154066.

11. M. S. Osheba, A. M. Aboutaleb, J. Desmet, and J. Knockaert, The Impact of Grid Distortion on the Power Conversion Harmonics of AC/DC Converters in the Supraharmonic Range, Electronics, vol. 13, no. 12, p. 2244, Jun. 2024, doi: 10.3390/electronics13122244.

12. K. Peiris, S. Elphick, and D. Robinson, Characterising Non-Intentional Supraharmonic Emissions from Inverters in Power Grids: Review and Challenges, Energies, vol. 18, no. 11, p. 2980, Jun. 2025, doi: 10.3390/en18112980.

13. A. Taghvaie, T. Warnakulasuriya, D. Kumar,

    F. Zare, R. Sharma, and D. M. Vilathgamuwa, A Comprehensive Review of Harmonic Issues and Estimation Techniques in Power System Networks Based on Traditional and Artificial Intelligence/Machine Learning, IEEE Access, vol. 11, pp. 3141731442,

    2023, doi: 10.1109/ACCESS.2023.3260768.

14. S. Shelar, D. Bankar, and P. K. Singh, Harmonic Measurement and Analysis in Manufacturing Industry: Case Study, in 2024 21st International Conference on Harmonics and Quality of Power (ICHQP), Chengdu, China: IEEE, Oct. 2024, pp. 463467. doi: 10.1109/ICHQP61174.2024.10768848.

15. S. K. Ronnberg, A. G. Castro, M. H. J. Bollen,

    A. Moreno-Munoz, and E. Romero-Cadaval, Supraharmonics from power electronics converters, in 2015 9th International Conference on Compatibility and Power Electronics (CPE), Costa da Caparica, Portugal: IEEE, Jun. 2015, pp. 539544. doi: 10.1109/CPE.2015.7231133.

16. B. Menssen, H. Brech, and H. Garbe, Measurement Validation for the Extension of Emission Measurements in Alternative Test Sites Above 1 GHz, IEEE Trans. Electromagn. Compat., vol. 58, no. 4, pp. 12741281, Aug. 2016, doi:

    10.1109/TEMC.2016.2554463.

17. S. T. Y. Alfalahi et al., Supraharmonics in Power Grid: Identification, Standards, and Measurement Techniques, IEEE Access, vol. 9, pp. 103677103690, 2021, doi:

    10.1109/ACCESS.2021.3099013.

18. J. Pinto, B. Grasel, and J. Baptista, Analysis of Supraharmonics Emission in Power Grids: A Case Study of Photovoltaic Inverters, Electronics, vol. 13, no. 24, p. 4880, Dec.

    2024, doi: 10.3390/electronics13244880.

19. H. Shao, R. Henriques, H. Morais, and E. Tedeschi, Power quality monitoring in electric grid integrating offshore wind energy: A review, Renewable and Sustainable Energy Reviews, vol. 191, p. 114094, Mar. 2024, doi: 10.1016/j.rser.2023.114094.

20. A. B. Ahmad et al., Renewable integration and energy storage management and conversion in grid systems: A comprehensive review, Energy Reports, vol. 13, pp. 2583 2602, Jun. 2025, doi:

    10.1016/j.egyr.2025.02.008.

21. A. M. Aboutaleb, M. S. Osheba, J. Desmet, and J. Knockaert, Assessment of Supraharmonic Emissions of Parallel Grid Connected Inverters Using Symmetric Carrier Phase Shift Concept, 2025, doi: 10.34746/EPE2025-0031.

22. L. Sandrolini and A. Mariscotti, Impact of short-time fourier transform parameters on the accuracy of EMI spectra estimates in the 2-

    150 kHz supraharmonic interval, Electric Power Systems Research, vol. 195, p. 107130, Jun. 2021, doi: 10.1016/j.epsr.2021.107130.

23. I. S. Samanta et al., Artificial intelligence and machine learning techniques for power quality event classification: a focused review and future insights, Results in Engineering, vol. 25, p. 103873, Mar. 2025, doi: 10.1016/j.rineng.2024.103873.

24. J. E. Caicedo, D. Agudelo-Martínez, E. Rivas- Trujillo, and J. Meyer, A systematic review of real-time detection and classification of power quality disturbances, Prot Control Mod Power Syst, vol. 8, no. 1, p. 3, Dec. 2023, doi: 10.1186/s41601-023-00277-y.

25. . Michalec, P. Kostya, and Z. Leonowicz, Supraharmonic Pollution Emitted by Nonlinear Loads in Power Networks Ongoing Worldwide Research and Upcoming Challenges, Energies, vol. 16, no. 1, p. 273, Dec. 2022, doi: 10.3390/en16010273.

26. S. Madgula, V. Veeramsetty, and R. Durgam, Signal processing approaches for power quality disturbance classification: A comprehensive review, Results in Engineering, vol. 25, p. 104569, Mar. 2025, doi: 10.1016/j.rineng.2025.104569.

27. M. Micev, M. alasan, M. Radulovi, S. H. E.

    A. Aleem, H. M. Hasanien, and A. F. Zobaa, Artificial Neural Network-Based Nonlinear Black-Box Modeling of Synchronous Generators, IEEE Trans. Ind. Inf., vol. 19, no. 3, pp. 28262837, Mar. 2023, doi: 10.1109/TII.2022.3187740.

28. K. Wiktorowicz, T2RFIS: type-2 regression- based fuzzy inference system, Neural Comput & Applic, vol. 35, no. 27, pp. 20299

    20317, Sep. 2023, doi: 10.1007/s00521-023-

    08811-7.

29. I. S. Samanta et al., A Comprehensive Review of Deep-Learning Applications to Power Quality Analysis, Energies, vol. 16, no. 11, p. 4406, May 2023, doi: 10.3390/en16114406.

30. W. U. K. Tareen et al., Mitigation of Power Quality Issues Due to High Penetration of Renewable Energy Sources in Electric Grid Systems Using Three-Phase APF/STATCOM Technologies: A Review, Energies, vol. 11, no. 6, p. 1491, Jun. 2018, doi: 10.3390/en11061491.

31. D. K. Dash and P. K. Sadhu, A Review on the Use of Active Power Filter for Grid- Connected Renewable Energy Conversion Systems, Processes, vol. 11, no. 5, p. 1467, May 2023, doi: 10.3390/pr11051467.

32. Y. Khayat et al., Decentralized Optimal Frequency Control in Autonomous Microgrids, IEEE Trans. Power Syst., vol. 34, no. 3, pp. 23452353, May 2019, doi:

    10.1109/TPWRS.2018.2889671.

33. B. Singh, V. Verma, A. Chandra, and K. Al- Haddad, Hybrid filters for power quality improvement, IEE Proc., Gener. Transm. Distrib., vol. 152, no. 3, p. 365, 2005, doi: 10.1049/ip-gtd:20045027.

34. N. B. P., Adaptive grid-connected inverter control schemes for power quality enrichment in microgrid systems: Past, present, and future perspectives, Electric Power Systems Research, vol. 230, p. 110288, May 2024, doi: 10.1016/j.epsr.2024.110288.

35. S. Jain, A. Satsangi, R. Kumar, D. Panwar, and M. Amir, Intelligent assessment of power quality diturbances: A comprehensive review on machine learning and deep learning solutions, Computers and Electrical Engineering, vol. 123, p. 110275, Apr. 2025, doi: 10.1016/j.compeleceng.2025.110275.

36. Y. Zhu, F. Zhong, J. Gao, Y. Cao, X. Wang, and Z. Jiang, Research on Supraharmonic

    detection in renewable energy grid Inte- gration based on improved ResNet18, Results in Engineering, vol. 25, p. 104281, Mar. 2025, doi: 10.1016/j.rineng.2025.104281.

37. H. Herman, Y. Jaya Kumar, S. Yong Wee, and

    V. Kumar Perhakaran, A Systematic Review on Deep Learning Model in Computer-aided Diagnosis forAnterior Cruciate Ligament Injury, CMIR, vol. 20, p. e15734056295157,

    May 2024, doi: 10.2174/0115734056295157240418043624.

38. M. M. Alam, M. J. Hossain, M. A. Habib, M.

Y. Arafat, and M. A. Hannan, Artificial intelligence integrated grid systems: Technologies, potential frameworks, challenges, and research directions, Renewable and Sustainable Energy Reviews, vol. 211, p. 115251, Apr. 2025, doi: 10.1016/j.rser.2024.115251.