IEEE Open Journal of the Industrial Electronics Society最新文献

{"title":"Design and Optimization of a Planar Transformer for LLC Converters in Electric Vehicle Charging Applications","authors":"Pedro Loureiro;Saghir Amin;João L. Afonso;Vítor Monteiro","doi":"10.1109/OJIES.2026.3712192","DOIUrl":"https://doi.org/10.1109/OJIES.2026.3712192","url":null,"abstract":"In recent years, the demand for more sustainable and energy-efficient mobility solutions has accelerated the development of electric vehicles (EVs) and their charging infrastructure. This article proposes a customized high-frequency planar transformer (HFPT) for integration into an <italic>LLC</i> resonant converter intended for EV fast-charging applications. The main contribution of this work is the design and optimization of the HFPT based on a modular segmented ferrite core structure composed of top and bottom tiles, side bars, and a central core piece. When assembled, these components form a compact planar magnetic structure suitable for high-power <italic>LLC</i> converter applications. The <italic>LLC</i> converter is designed to support a wide input voltage range of 550 to 800 V and an output voltage range of 240 to 460 V. The HFPT design methodology carefully considers the effects of the segmented core structure, including magnetizing and leakage inductances, flux density distribution, core and winding losses, and parasitic capacitances. The magnetic design was validated through finite element method electromagnetic simulations performed in ANSYS Maxwell. The overall converter performance was analyzed in PSIM and subsequently verified through the implementation of a 10 kW laboratory prototype. Experimental results were obtained for a 3 kW prototype, confirming zero-voltage and zero-current switching over the entire tested output power and voltage ranges. The prototype achieved a maximum efficiency of 96.1%, with a flat efficiency curve maintained across the full tested operating voltage range, demonstrating that the customized HFPT is a practical, scalable, and cost-effective solution for high-power EV fast-charging applications.","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"1198-1219"},"PeriodicalIF":4.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11603441","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148627947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Physics-Informed ITSC Diagnosis in Induction Motors via CPT-Based V–I Trajectories","authors":"José Pedro B. Neto;Murillo G. Gentil;Mariana Pelisson;Alessandro Goedtel;Marcelo F. Castoldi;Augusto M. S. Alonso;Wesley A. Souza","doi":"10.1109/OJIES.2026.3716567","DOIUrl":"https://doi.org/10.1109/OJIES.2026.3716567","url":null,"abstract":"This article presents a physics-informed representation approach for supervised classification of predefined stator interturn short-circuit (ITSC) severity classes in three-phase induction motors. The term physics-informed is used here in a representation-level sense: physical knowledge is introduced through deterministic input construction, not through physics-constrained neural-network losses or architecture modifications. The proposed formulation constructs voltage–current trajectories from the measured phase currents (&lt;inline-formula&gt;&lt;tex-math&gt;$v$&lt;/tex-math&gt;&lt;/inline-formula&gt;–&lt;inline-formula&gt;&lt;tex-math&gt;$i_{phi }$&lt;/tex-math&gt;&lt;/inline-formula&gt;) and from current components obtained with the conservative power theory (CPT). CPT provides a time-domain, projection-based decomposition of the stator-current vector into active (&lt;inline-formula&gt;&lt;tex-math&gt;$boldsymbol{i}_{mathrm{a}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;), reactive (&lt;inline-formula&gt;&lt;tex-math&gt;$boldsymbol{i}_{mathrm{r}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;), unbalanced (&lt;inline-formula&gt;&lt;tex-math&gt;$boldsymbol{i}_{mathrm{u}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;), and void (&lt;inline-formula&gt;&lt;tex-math&gt;$boldsymbol{i}_{mathrm{v}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;) components in &lt;inline-formula&gt;&lt;tex-math&gt;$L^{2}[0,T]^{3}$&lt;/tex-math&gt;&lt;/inline-formula&gt;. Pairing phase voltages with each CPT component yields componentwise trajectory families whose geometry is associated, respectively, with average power transfer, magnetic energy exchange, interphase asymmetry, and residual distortion. At the terminal-variable level, ITSC progression is interpreted as a redistribution of current energy from the balanced active–reactive subspace toward imbalance and distortion subspaces, quantified through &lt;inline-formula&gt;&lt;tex-math&gt;$mathcal {F}_{text{ITSC}}$&lt;/tex-math&gt;&lt;/inline-formula&gt; and expressed in trajectory deformation, slope variation, phase asymmetry, and loop thickening. The approach is experimentally assessed on a 1 horsepower (HP) induction motor under torque variation, supply–voltage imbalance, and multiple ITSC severity levels. Five CPT-based trajectory representations, namely, &lt;inline-formula&gt;&lt;tex-math&gt;$v$&lt;/tex-math&gt;&lt;/inline-formula&gt;–&lt;inline-formula&gt;&lt;tex-math&gt;$i_{mathrm{a}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;, &lt;inline-formula&gt;&lt;tex-math&gt;$v$&lt;/tex-math&gt;&lt;/inline-formula&gt;–&lt;inline-formula&gt;&lt;tex-math&gt;$i_{mathrm{r}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;, &lt;inline-formula&gt;&lt;tex-math&gt;$v$&lt;/tex-math&gt;&lt;/inline-formula&gt;–&lt;inline-formula&gt;&lt;tex-math&gt;$i_{mathrm{u}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;, &lt;inline-formula&gt;&lt;tex-math&gt;$v$&lt;/tex-math&gt;&lt;/inline-formula&gt;–&lt;inline-formula&gt;&lt;tex-math&gt;$i_{mathrm{v}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;, and the hybrid reactive-void representation &lt;inline-formula&gt;&lt;tex-math&gt;$v$&lt;/tex-math&gt;&lt;/inline-formula&gt;–&lt;inline-formula&gt;&lt;tex-math&gt;$i_{text{rv}}$&lt;/tex-math&gt;&lt;/inline-formula&gt;, are compared with the conventional raw &lt;inline-formula&gt;&lt;tex-math&gt;$v$&lt;/tex-math&gt;&lt;/inline-formula&gt;–&lt;inline-formula&gt;&lt;tex-math&gt;$i_{phi }$&lt;/tex-math&gt;&lt;/inline-formula&gt; representation using a manually ","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"1234-1265"},"PeriodicalIF":4.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11619550","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148747860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Practical Prescribed-Time Parameter Estimation for Underactuated Uncrewed Surface Vessel Based on DREM Method and LOS Control","authors":"Jawhar Ghommam;Maarouf Saad;Ibrahim Al-Naimi;Majdi Mansouri;Faisal Mnif","doi":"10.1109/OJIES.2026.3721426","DOIUrl":"https://doi.org/10.1109/OJIES.2026.3721426","url":null,"abstract":"Accurate control-oriented composite hydrodynamic parameter knowledge is essential for high-performance control of uncrewed surface vessels, yet conventional identification methods rely on persistent excitation (PE), which is often impractical in real operations. This article proposes an integrated framework combining online parameter estimation with prescribed-time line-of-sight cascade guidance control. The estimation scheme uses dynamic regressor extension and mixing to decouple a 14-parameter composite dynamic model into independent scalar regressions, followed by a prescribed-time least-squares estimator that guarantees practical prescribed-time convergence under bounded disturbances and exact prescribed-time convergence in the disturbance-free case of the estimation errors at a user-defined time. The approach requires only finite excitation from standard maneuvers, eliminating the need for PE. Leveraging these estimates, a prescribed-time cascade guidance controller is developed to achieve precise velocity tracking, with adaptive compensation for unknown disturbances. The framework ensures practical prescribed-time stability of both the composite parameter estimation errors and the tracking errors, independent of initial conditions. Theoretical analysis and experimental results validate the approach, demonstrating accurate composite parameter identification and high-precision tracking under realistic operating conditions.","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"1349-1365"},"PeriodicalIF":4.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11644400","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148836853","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"False Recovery? The Impact of Irregular Rest Periods on Lithium-Ion Battery Capacity and State-of-Health Estimation in Intermittent Use Applications","authors":"Thomas R. McKinney;Daniel T. Gladwin;Matthew J. Smith;Christopher Longbottom","doi":"10.1109/OJIES.2026.3718055","DOIUrl":"https://doi.org/10.1109/OJIES.2026.3718055","url":null,"abstract":"Accurate state-of-health (SOH) estimation underpins maintenance planning, warranty assessment, and safe operation of lithium-ion batteries, yet it rests on an assumption that fails in real-world intermittent use: that capacity fades smoothly and monotonically. This article shows that irregular rest periods, characteristic of leisure vehicles, backup power, off-grid storage, and standby EV or grid assets, can trigger capacity recovery far larger and more persistent than the sub-1% transients reported in prior work. Drawing on a longitudinal dataset spanning over two years, thousands of cycles, and four lithium-ion cell types, we analyze the joint influence of C-rate, state-of-charge during rest, and rest duration across the full ageing trajectory from fresh to advanced degradation. We observe recovery events large enough to alter the degradation pathway for hundreds of subsequent cycles, effectively masking a cell's true state and distorting cycle-life prediction. Critically, recovery magnitude is governed less by rest duration than by prior electrochemical imbalance and is most pronounced at the mid-range SOCs where systems are typically stored, a regime underexamined in existing literature. We demonstrate the consequences for SOH estimation using a Kalman-filter estimator applied to two contrasting cells: under sparse-update conditions typical of real-world BMS operation, tracking errors exceed 6% for a recovery-affected cell while remaining below 1.5% for one degrading monotonically. These findings establish rest-induced recovery as a first-order factor in degradation behavior, not a measurement artefact, and motivate rest-aware or adaptive estimation methods for intermittently operated energy storage systems.","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"1311-1326"},"PeriodicalIF":4.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11627162","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148781353","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Estimation of Apparent Inductances in Electrically Excited Synchronous Machines Using Enhanced Particle Swarm Optimization","authors":"Bowen Jiang;Junfei Tang;Yujing Liu","doi":"10.1109/OJIES.2026.3671359","DOIUrl":"https://doi.org/10.1109/OJIES.2026.3671359","url":null,"abstract":"Electrically excited synchronous machines (EESMs) are becoming increasingly preferred in electric vehicles due to their nonreliance on rare-earth materials. In traction EESMs, accurate estimation of apparent inductances is crucial for designing current controllers, calculating electromagnetic torques, and determining operation maps. Existing inductance estimation methods, developed based on the permanent magnet synchronous machines, only consider self-inductances of the stator windings. However, due to the presence of both stator and field windings in EESMs, it is necessary to estimate both self and mutual inductances, which can greatly increase the estimation difficulty due to more unknown parameters. Hence, in this article, a new EESM apparent inductance estimation method is proposed. In the proposed method, self and mutual inductances are simultaneously estimated using an enhanced particle swarm optimization (PSO) algorithm. The effectiveness of the proposed method is experimentally verified. Experimental results show that, compared to the standard PSO and dynamic PSO, the enhanced PSO exhibits higher estimation accuracy and better estimation stability. Additionally, by employing the estimated apparent inductances in the EESM current controller, precise current control can be achieved during dynamic processes.","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"551-562"},"PeriodicalIF":4.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11423888","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147557594","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Hybrid Observer for Cell Voltage Estimation of Modular Multilevel Converters Without Internal Voltage Measurements","authors":"Oliver Kalmbach;Christoph M. Hackl","doi":"10.1109/OJIES.2026.3673007","DOIUrl":"https://doi.org/10.1109/OJIES.2026.3673007","url":null,"abstract":"The modular multilevel converter (MMC) has become a famous topology for high-power applications due to its modularity, scalability, and superior output waveform quality. For the MMC to work, all cell voltages must be known. Therefore, voltage sensors are installed in each cell. This results in a high number of voltage sensors, which complicates the system, increases costs, and reduces reliability. Observers can be used to estimate the required voltages, thereby reducing the number of sensors. This article proposes a hybrid observer (HO) approach for estimating cell voltages in MMCs without the use of internal voltage sensors in each cell. Instead, solely the internal cluster currents, the DC-link voltage, and the grid voltages are measured. The design is based on six HOs running in parallel, implemented on a field-programmable gate array to achieve fast execution times of 1 <inline-formula><tex-math>$mu$</tex-math></inline-formula>s. An adaptive observer, implemented on a digital signal processor, completes the design and estimates the sum capacitor voltages required for the outer energy control loop of the MMC, and additionally provides a common-mode voltage estimate to the HOs. The observer design is theoretically derived, and the proper functionality is validated experimentally under steady-state and transient conditions. Also, parameter mismatches are considered to demonstrate the robustness of the proposed observer approach.","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"671-688"},"PeriodicalIF":4.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11430589","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147606281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automatic Mapping of AutomationML Files to Ontologies for Graph Queries and Validation","authors":"Tom Westermann;Malte Ramonat;Johannes Hujer;Felix Gehlhoff;Alexander Fay","doi":"10.1109/OJIES.2026.3672995","DOIUrl":"https://doi.org/10.1109/OJIES.2026.3672995","url":null,"abstract":"AutomationML has seen widespread adoption as an open data exchange format in the automation domain. It is an open and vendor-neutral standard based on the extensible markup language (XML). However, AutomationML extends XML with additional semantics that limit the applicability of common XML-tools for applications, such as querying and data validation. This article demonstrates how the transformation of AutomationML into web ontology language (OWL) enables new use cases in querying with SPARQL and validation with shapes constraint language. To support this, it provides practitioners with, first, an up-to-date ontology of the concepts defined in the AutomationML standard and, second, a declarative mapping to automatically transform any AutomationML model into resource description framework triples. A study on examples from the automation domain concludes that transforming AutomationML to OWL opens up new powerful ways for querying and validation that would have been impossible without this transformation.","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"589-607"},"PeriodicalIF":4.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11430567","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147606323","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IEEE Open Journal of the Industrial Electronics Society Publication Information","authors":"","doi":"10.1109/OJIES.2025.3646948","DOIUrl":"https://doi.org/10.1109/OJIES.2025.3646948","url":null,"abstract":"","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"C2-C2"},"PeriodicalIF":4.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11450486","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148491457","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IEEE Industrial Electronics Society Information","authors":"","doi":"10.1109/OJIES.2025.3646950","DOIUrl":"https://doi.org/10.1109/OJIES.2025.3646950","url":null,"abstract":"","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"C3-C3"},"PeriodicalIF":4.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11456477","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148489435","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Digital Twin-Augmented Multifeature Fusion Ensemble for Accurate and Uncertain PV Power Forecasting","authors":"Majdi Mansouri;Abdelmalek Kouadri;Rami Al-Hmouz;Abdullah Al-Badi;Abdulnasir Hossen","doi":"10.1109/OJIES.2026.3706363","DOIUrl":"https://doi.org/10.1109/OJIES.2026.3706363","url":null,"abstract":"This article presents a short-term (1-h ahead) physics-guided multirepresentation feature fusion framework for photovoltaic (PV) power prediction with uncertainty quantification based on quantile regression for interval forecasting. The proposed hybrid methodology integrates feature-representation fusion, deep learning architectures, including long short-term memory (LSTM) and gated recurrent unit (GRU), and machine learning models, such as support vector regression (SVR) and random forest (RF), along with dimensionality reduction techniques, namely principal component analysis (PCA) and autoencoder (AE), and a fully defined physics-based digital twin (DT) model with explicit irradiance-to-power equations and parameter calibration. The DT residual, computed causally using only training data, is incorporated as a physics-informed feature to capture unmodeled nonlinearities. PIs are learned using the pinball (quantile) loss function, replacing heuristic assumptions, and feature-representation fusion reduces uncertainty while enhancing robustness. Validation on real meteorological datasets from Izki and Manah stations (2017–2023) shows that conventional feature sets achieve root mean squared error (RMSE) values between 0.14–0.17 with coefficient of determination (&lt;inline-formula&gt;&lt;tex-math&gt;$R^{2}$&lt;/tex-math&gt;&lt;/inline-formula&gt;) ranging from 0.47–0.61. Incorporating DT residuals significantly improves performance, while maintaining realistic generalization performance, with LSTM achieving RMSE &lt;inline-formula&gt;&lt;tex-math&gt;$approx$&lt;/tex-math&gt;&lt;/inline-formula&gt; 0.0105 and &lt;inline-formula&gt;&lt;tex-math&gt;$R^{2}$&lt;/tex-math&gt;&lt;/inline-formula&gt; up to 0.9979. SVR and RF models also benefit, achieving RMSE &lt;inline-formula&gt;&lt;tex-math&gt;$leq 0.0525$&lt;/tex-math&gt;&lt;/inline-formula&gt; and &lt;inline-formula&gt;&lt;tex-math&gt;$R^{2} geq 0.9467$&lt;/tex-math&gt;&lt;/inline-formula&gt;. Comprehensive evaluation including ensemble comparisons, ablation studies, and uncertainty metrics, namely prediction interval coverage probability and mean prediction interval width, confirms high coverage and sharpness, while feature-representation fusion ensures consistent, low-uncertainty predictions. Further experiments on a physically inspired synthetic dataset with explicitly defined generation process and temporal resolution of 1000 samples demonstrate similar trends. Using DT residual features, LSTM and GRU achieve &lt;inline-formula&gt;&lt;tex-math&gt;$R^{2} &gt; 0.998$&lt;/tex-math&gt;&lt;/inline-formula&gt; with RMSE &lt;inline-formula&gt;&lt;tex-math&gt;$approx 0.042$&lt;/tex-math&gt;&lt;/inline-formula&gt;, while RF reaches &lt;inline-formula&gt;&lt;tex-math&gt;$R^{2} = 0.897$&lt;/tex-math&gt;&lt;/inline-formula&gt; and a relatively higher RMSE &lt;inline-formula&gt;&lt;tex-math&gt;$approx 0.326$&lt;/tex-math&gt;&lt;/inline-formula&gt;. PCA and AE representations improve computational efficiency but provide comparatively lower predictive accuracy. Direct comparison between individual models and the proposed ensemble highlights the robustness and stability of the fusion strategy, particularly in high-noise con","PeriodicalId":52675,"journal":{"name":"IEEE Open Journal of the Industrial Electronics Society","volume":"7 ","pages":"1037-1050"},"PeriodicalIF":4.6,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11575094","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148489189","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}