Review Paper for SOC Estimation Techniques: Challenges and Future Areas for Improvement

May combine two claims.

ML models are most effective in scenarios where the battery operates under condition changes because they can describe relationships between variables in detail.

“ML models are most effective in scenarios where the battery operates under condition changes, because it can describe relationships between variables in detail, [42].”

• cites [42] 42. Y. Dong, K. Chen, G. Zhang, and R. Li, "Joint Estimation of State of Charge and State of Health of Lithium-Ion Batteries Based on Stacking Machine Learning Algorithm," World Electric Vehicle Journal 2024, Vol. 15, Page 75, vol. 15, no. 3, p. 75, Feb. 2024, doi: 10.3390/WEVJ15030075. sourcePartly supported

  The Introduction (page unavailable) says data-driven methods can adapt to uncertainty and nonlinear battery behavior and map measured features to state of charge. It does not establish that ML models are most effective when operating conditions change, or that describing variable relationships in detail is the reason for such superiority. (LLM, conf=0.83)

  Experimental — accuracy not yet measured

May combine two claims.

Deep model training is computationally resource consuming, and deploying models for real-time systems often may call for specialized architecture such as GPUs.

“Model training deeply is computationally resource consuming, and deploying models for real-time systems often may call for specialized architecture, namely GPUs [47].”

• cites [47] 47. M. A. Hannan et al., "Deep learning approach towards accurate state of charge estimation for lithium-ion batteries using self-supervised transformer model," Scientific Reports 2021 11:1, vol. 11, no. 1, pp. 1-13, Oct. 2021, doi: 10.1038/s41598-021-98915- 8. sourcePartly supported

  The Introduction says deep-learning training can take many hours or days, and the Methods (Implementation) report training the models on a GPU. The paper discusses GPU use for training, but does not say that deploying models in real-time systems may require GPUs. The supplied passages have no page numbers. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

May combine two claims.

Kalman filter gives real-time variation to the SOC estimate, making it efficient in ever-changing conditions.

“The algorithm gives real time variation to the SOC estimate, which makes it very efficient when operating in ever changing conditions [57].”

• cites [57] 57. L. Hu, R. Hu, Z. Ma, and W. Jiang, "State of Charge Estimation and Evaluation of Lithium Battery Using Kalman Filter Algorithms," Materials, vol. 15, no. 24, Dec. 2022, doi: 10.3390/MA15248744. sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

SOC estimation methods can be categorized into direct measurement techniques, model-based techniques, and data-driven techniques.

“The paper categorizes various SOC estimation methods into three primary approaches: direct measurement techniques, model-based methods, and datadriven algorithms.”

• Also stated in: § 1. Introduction · hedged

  “In one way or the other, these techniques can be said to be categorized into direct measurement techniques, model-based techniques and data driven techniques.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In EV applications, SOC estimation is crucial for determining travel range, controlling energy consumption, and providing user information.

“Hence, in some applications like EVs, estimation of SOC is crucial in determining the range of travel, controlling the energy consumption, and, most importantly, in providing information to the users at any one time.”

May combine two claims.

Direct measurement techniques, such as Coulomb counting and Open Circuit Voltage (OCV), are straightforward but often lack precision under dynamic conditions.

“Direct measurement techniques, such as Coulomb counting and Open Circuit Voltage (OCV), are straightforward but often lack precision under dynamic conditions.”

• Also stated in: § 4.3.1 Direct Measurement Methods · hedged

  “These methods are very easy to apply, but in most cases, they may not be accurate enough for use in for instance a complex or dynamic environment.”

May combine two claims.

Data-driven approaches utilizing machine learning algorithms exhibit promising adaptability and accuracy, albeit relying heavily on large datasets.

“In contrast, data-driven approaches utilizing machine learning algorithms exhibit promising adaptability and accuracy, albeit relying heavily on large datasets.”

• Also stated in: § 4.2.2 Based on Nature of Technique · hedged

  “These methods may perform well for quantitative and even nonlinear relationships between measurable parameters and SOC but for this very reason, their true power and accuracy is highly dependent on the volumes of available training datasets.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Model-based methods can capture a large spectrum of internal and external influences on battery performance but can also be resource consuming.

“These kinds of models can capture a large spectrum of internal and external influences of battery performance but can also be resource consuming.”

May combine two claims.

Data-driven approaches use past and current data to update SOC and can develop models that accurately predict SOC.

“As these approaches use past and current data to update SOC, they can also develop models that accurately predict SOC.”

Probably not a substantial claim: may not pass S3 (load-bearing).

SOC estimation techniques can be categorized based on the dominating parameter used, including voltage, current, temperature, and internal resistance.

“SOC estimation techniques can be categorized based on the dominating parameter used in estimating the battery SOC, including voltage, current, temperature and internal resistances [23].”

• cites [23] 23. B. Yang et al., "Classification, summarization and perspectives on state-of-charge estimation of lithium-ion batteries used in electric vehicles: A critical comprehensive survey," J Energy Storage, vol. 39, p. 102572, Jul. 2021, doi: 10.1016/J.EST.2021.102572. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The characteristics of Li-ion batteries make them the most suitable battery for SOC estimation techniques.

“High energy density, long cycle life, fast charging capabilities and low self-discharge rates that characterize Li-ion batteries make it the most suitable battery for SOC estimation techniques.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Model-based methods, including Equivalent Circuit Models (ECM) and electrochemical models, provide detailed insights into battery behavior but can be computationally intensive.

“Model-based methods, including Equivalent Circuit Models (ECM) and electrochemical models, provide detailed insights into battery behavior but can be computationally intensive.”

• Also stated in: § 1. Introduction · plain

  “These kinds of models can capture a large spectrum of internal and external influences of battery performance but can also be resource consuming.”

Probably not a substantial claim: may not pass S4 (based).

Current SOC estimation methods are hindered by their dependence on accurate battery models and quality data, which can degrade over time.

“Current SOC estimation methods are hindered by their dependence on accurate battery models and quality data, which can degrade over time.”

Classified as probably not a substantial claim.

Poor SOC estimation can lead to unexpected battery depletion, potentially causing range anxiety among users.

“Poor SOC estimation can lead to unexpected battery depletion, potentially causing range anxiety among users, making it a significant area of research in battery management systems (BMS).”

Classified as probably not a substantial claim. May combine two claims.

Accurate SOC estimation is essential for optimizing battery performance, ensuring safety, and prolonging battery life.

“accurate SOC estimation becomes essential for optimizing battery performance, enhancing safety, and extending battery life.”

• Also stated in: § 4.1 Overview of SOC · plain

  “Accurate SOC estimation is essential for optimizing battery performance, ensuring safety, and prolonging battery life.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Li-ion batteries are characterized by high energy density, long cycle life, fast charging capabilities, and low self-discharge rates.

“High energy density, long cycle life, fast charging capabilities and low self-discharge rates that characterize Li-ion batteries make it the most suitable battery for SOC estimation techniques.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Machine learning techniques use algorithms fed with big data to identify dependency between input parameters and SOC.

“Machine learning (ML) techniques entail use of algorithms that are fed with big data so as to identify dependency between input parameters and SOC, [38].”

• cites [38] 38. P. Eleftheriadis, S. Giazitzis, S. Leva, and E. Ogliari, "Data-Driven Methods for the State of Charge Estimation of Lithium-Ion Batteries: An Overview," Forecasting 2023, Vol. 5, Pages 576-599, vol. 5, no. 3, pp. 576-599, Sep. 2023, doi: 10.3390/FORECAST5030032. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims. The quote only partly states this claim.

ECMs allow for the incorporation of temperature within reaction kinetics, which increases performance in environments of varying temperature.

“ECMs also allow for the incorporation of temperature within the reaction kinetics which increases the performance of the bioreactors in environments of varying temperature.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Data-driven SOC estimation schemes build the relationship between measurable parameters such as voltage, current and temperature, and SOC using historical data and machine learning methods.

“data-driven SOC estimation schemes build the relationship between demonstrable parameters such as voltage, current and temperature, and SOC using historical data and machine learning methods [37].”

• cites [37] 37. M. H. Sulaiman and Z. Mustaffa, "State of charge estimation for electric vehicles using random forest," Green Energy and Intelligent Transportation, vol. 3, no. 5, p. 100177, Oct. 2024, doi: 10.1016/J.GEITS.2024.100177. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Accounting for all variables in parameter-based classification is challenging, especially when external conditions such as temperature vary significantly during operation.

“However, the challenge lies in the complexity of accounting for all variables, especially when external conditions, such as temperature, vary significantly during operation.”

Probably not a substantial claim: may not pass S3 (load-bearing).

SOC estimation techniques can be classified by their inherent nature into direct measurements, analytical estimations, and data analysis methods.

“SOC estimation techniques can also be classified by their inherent nature, falling into three broad categories [24]: infrared diagnostics, direct measurements, analytical estimations, and data analysis methods.”

• cites [24] 24. W.-Y. Chang, : M Brünig, and E. Di Nardo, "The State of Charge Estimating Methods for Battery: A Review," Int Sch Res Notices, vol. 2013, no. 1, p. 953792, Jan. 2013, doi: 10.1155/2013/953792. sourceNot checked — no full text of the source

May combine two claims.

The OCV method gives a highly accurate reading when the battery is calmly charged or discharged but has a low chance of success when the battery is actively charging or discharging.

“It gives a highly accurate reading when the battery is calmly charged or discharges but has a low chance of success when the battery is charging or discharging actively.”

May combine two claims.

ECMs only represent battery electrochemical behavior and therefore can be off in reproducing battery internal mechanisms.

“Another issue is that ECMs only represent the battery electrochemical behavior and therefore can be off in the reproduction of battery internal mechanisms.”

May combine two claims.

Electrochemical models are accurate with high precision and can interpret the internal state of a battery including SOC, SOH, and capacity fade.

“Battery electrochemical models are accurate with high precision, which gives them an added advantage of interpreting the internal state of a battery including SOC, SOH and capacity fade.”

May combine two claims.

Coulomb counting depends on current sampling and initial SOC estimation, so slight discrepancies such as inability to predict capacity loss can significantly alter estimated SOC values.

“Because they depend on current sampling and/or initial SOC estimation, various slight discrepancies, such as the inability to predict capacity loss, can significantly alter estimated SOC values.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Coulomb counting is an open-loop technique because measurements cannot be corrected using feedback.

“In Coulomb counting, for instance, the algorithm is an open-loop technique since the measurements cannot be corrected using feedback.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Kalman filtering and other observer-based methods are closed-loop techniques.

“Such approaches include Kalman filtering and other observer-based methods have been categorized under this.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Coulomb counting and the Open Circuit Voltage (OCV) method are widely used direct measurement approaches.

“From the direct measurement approaches, Coulomb counting and the Open Circuit Voltage (OCV) method are widely used.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Closed-loop techniques rely on a feedback system that constantly updates the SOC estimate based on the real value and the estimate.

“Closed-loop techniques, in contrast, rely on feedback system that constantly updates the SOC estimate depending on the real value and estimate one.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Two popular model-based approaches are Equivalent Circuit Models (ECMs) and electrochemical models.

“Two popular model-based approaches are Equivalent Circuit Models (ECMs) and electrochemical models.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

ECMs can be easily incorporated in real-time systems and are commonly utilized in EV applications.

“They can be easily incorporated in real-time systems and are commonly utilized in EV applications as they are able to simulate both one-time and temporary cycles of the battery.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Temperature-based SOC estimation techniques are less common, though temperature can influence voltage and resistance, indirectly affecting SOC estimates.

“Temperature-based techniques are less common, though temperature can influence both voltage and resistance, thereby indirectly affecting SOC estimates.”

May combine two claims.

ECMs are computationally efficient and estimate SOC under dynamic operating conditions with reasonable accuracy.

“As compared to other methods ECMs are computationally efficient and estimates, SOC under dynamic operating conditions with reasonable accuracy.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

Direct measurement techniques are relatively easy to deploy but can be less accurate, particularly in dynamic operating environments.

“Though these techniques are relatively easier to deploy they can be less accurate particularly in dynamic operating environment.”

Probably not a substantial claim: may not pass S3 (load-bearing).

SOC estimation techniques can also be categorized by their open-loop or closed-loop functionality.

“As with other SOC estimation procedures, techniques can also be categorized by their openloop or closed-loop functionality [25].”

• cites [25] 25. K. Das and R. Kumar, "A Comprehensive Review of Categorization and Perspectives on State-of-Charge Estimation Using Deep Learning Methods for Electric Transportation," Wirel Pers Commun, vol. 133, no. 3, pp. 1599-1618, Dec. 2023, doi: 10.1007/S11277-023-10830-5/METRICS.Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Closed-loop implementation is often more computationally complex and requires more design effort than open-loop implementation.

“Nonetheless, the need for special computational methods for the implementation of a closedloop is often more complicated than in the open-loop case, as well as the design effort needed to incorporate the former.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Direct measurement techniques are quick and easy to deploy but often need to be complemented with model-based or data-driven approaches to compensate for their inherent limitations in real-time, dynamic environments.

“Direct measurement techniques are quick and easy to deploy, but they often need to be complemented with model-based or data-driven approaches to compensate for their inherent limitations in real-time, dynamic environments.”

May combine two claims.

Coulomb counting is not very accurate at determining SOC and is usually applied with other algorithms such as Kalman filtering.

“However, since Coulomb counting is not very accurate at determining the state of charge it is usually applied with other algorithms, such as Kalman filtering [27].”

• cites [27] 27. B. E. Lebrouhi, Y. Khattari, B. Lamrani, M. Maaroufi, Y. Zeraouli, and T. Kousksou, "Key challenges for a large-scale development of battery electric vehicles: A comprehensive review," J Energy Storage, vol. 44, p. 103273, Dec. 2021, doi: 10.1016/J.EST.2021.103273. sourceNot checked — no full text of the source

May combine two claims.

Data-driven SOC estimation techniques do not require a precise physical representation of the battery system and are relatively versatile with respect to battery chemistry and conditions.

“these non-invasive techniques do not need a precise physical representation of the battery system and are therefore relatively versatile with respect to battery chemistry and conditions.”

May combine two claims.

Coulomb counting is relatively simple but suffers from error accumulation over time.

“Since the amount of charge entering or leaving the battery directly affects SOC, this method is relatively simple but suffers from error accumulation over time.”

• Also stated in: § 4.3.1.1 Coulomb Counting · plain

  “The main disadvantage of Coulomb counting is that the calculation of the SOC increases with time cause of cumulative errors.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Coulomb counting is easy to calculate and can be easily embedded in real-time systems because of its simplicity.

“Coulomb counting is easy to calculate and can be easily embedded in real-time systems because of its simplicity.”

May combine two claims.

Temperature variations affect the OCV reading, requiring temperature compensation to improve accuracy.

“Temperature variations also affect the OCV reading, requiring temperature compensation to improve accuracy.”

• AssumptionRests on an assumption: Temperature variations affect OCV readings.

May combine two claims.

The OCV-SOC relationship is nonlinear, especially at the extremes of SOC near 0% and 100%, making the OCV method less accurate in those regions.

“Additionally, the OCV-SOC relationship is nonlinear, as shown in Figure 4, [28], especially at the extremes of SOC (near 0% and 100%), making the method less accurate in those regions.”

• cites [28] 28. "What Is BMS? -News." Accessed: Sep. 30, 2024. [Online]. Available: https://www.fangpusun.com/news/what-is-bms-76816591.htmlNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Coulomb counting assumes that the nominal capacity does not vary with time.

“This method involves determination of SOC at the start of battery operation and further assumes that the nominal capacity does not vary with time.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Classifying SOC techniques by the parameters they use offers flexibility in practical applications.

“The advantage of classifying SOC techniques by the parameters they use is the flexibility it offers in practical applications.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Internal resistance can be used with model-based techniques to provide deeper understanding of battery state.

“Internal resistance is another popular method for measuring battery SOC as it can be used in conjunction with model-based techniques, where the dynamic behavior of the battery's impedance provides a deeper understanding of the battery's state.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The DFN model developed by Doyle et al. accounts for diffusion of lithium-ion concentration in the solid and liquid phase.

“One such electrochemical model is the DFN model developed by Doyle et al [34], which accounts for diffusion of lithium-ion concentration in the solid and liquid phase.”

• cites [34] 34. R. Drummond and S. R. Duncan, "Observer design for the Doyle-Fuller-Newman Li- ion battery model without electrolyte dynamics," J Energy Storage, vol. 23, pp. 250- 257, Jun. 2019, doi: 10.1016/J.EST.2019.02.012. sourceNot checked — no full text of the source

May combine two claims.

SVM is considered inflexible compared to other data mining techniques like neural networks and may lack efficacy in intricate or fluctuating environments.

“analytically, SVM is considered an inflexible model compared to some other approaches of data mining techniques like neural networks; it maybe lacks efficacy specifically in intricate or fluctuating environments [54].”

• cites [54] 54. G. F. Guo, L. Shui, X. L. Wu, and B. G. Cao, "SOC Estimation for Li-Ion Battery Using SVM Based on Particle Swarm Optimization," Adv Mat Res, vol. 1051, pp. 1004-1008, 2014, doi: 10.4028/WWW.SCIENTIFIC.NET/AMR.1051.1004. sourceNot checked — no full text of the source

May combine two claims.

Neural networks have real-time evaluation characteristics; real SOC predictions are possible once training is complete.

“Neural Networks also have real-time evaluation characteristics, one can have real SOC predictions as soon as the training is complete.”

May combine two claims.

Combining data-driven and model-based techniques combines characteristics and achieves advantages of both approaches.

“The data-driven and model-based techniques are used together and in this the characteristics of both the approaches are combined and thus advantageous of both approaches are achieved [69].”

• cites [69] 69. X. Lai, S. Wang, L. He, L. Zhou, and Y. Zheng, "A hybrid state-of-charge estimation method based on credible increment for electric vehicle applications with large sensor and model errors," J Energy Storage, vol. 27, p. 101106, Feb. 2020, doi: 10.1016/J.EST.2019.101106. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Adaptive filtering is an improved form of traditional filtering where model parameters adjust over time due to changes in battery operating conditions.

“Adaptive filtering stands as an improved form of traditional filtering like Kalman filtering whereby the model parameters are adjusted with time due to changes in the battery operating conditions [73].”

• cites [73] 73. X. Renxin, H. Yanwen, Z. Wei, and C. Zhaohui, "A novel approach to estimate the state of charge for lithium-ion battery under different temperatures incorporating open circuit voltage online identification," J Energy Storage, vol. 67, p. 107509, Sep. 2023, doi: 10.1016/J.EST.2023.107509. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Coulomb Counting is significantly less complex and does not drain battery power, enabling use in low-powered systems.

“Coulomb Counting, for example, is significantly less complex and will not drain battery power in a system, thus making it possible to work on low-powered systems.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Fuzzy logic SOC estimation methods incorporate linguistic variables to manage vagueness and imprecision of battery SOC.

“The methods employing Fuzzy logic technique of SOC estimation incorporates linguistic variables to manage vagueness and imprecision of the SOC of the battery [48].”

• cites [48] 48. D. Jiani, L. Zhitao, W. Youyi, and W. Changyun, "A fuzzy logic-based model for Li- ion battery with SOC and temperature effect," IEEE International Conference on Control and Automation, ICCA, pp. 1333-1338, 2014, doi: 10.1109/ICCA.2014.6871117. sourceNot checked — no full text of the source

May combine two claims.

Particle filtering requires more computational effort than Kalman filtering and is especially suitable for strongly nonlinear systems.

“The above approaches of particle filtering require more computational effort as compared to the Kalman filtering; it is especially suitable for strongly nonlinear systems [63]”

• cites [63] 63. Y. He, Z. Hong, Z. Zhang, and S. Huang, "Particle Filter Based Robust State and Parameter Estimation for Estimating SOC and Discharge Current of Lithium Batteries," Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022, pp. 244-249, 2022, doi: 10.1109/DDCLS55054.2022.9858503. sourceNot checked — no full text of the source

May combine two claims.

H-infinity filtering requires detailed understanding of the battery and is demanding and complex in real-time systems.

“The algorithm lends itself to a detailed understanding of the battery and can be very demanding and complex in real time systems [66].”

• cites [66] 66. Y. Chen, R. Li, Z. Sun, L. Zhao, and X. Guo, "SOC estimation of retired lithium-ion batteries for electric vehicle with improved particle filter by H-infinity filter," Energy Reports, vol. 9, pp. 1937-1947, Dec. 2023, doi: 10.1016/J.EGYR.2023.01.018. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Combining model-based and data-driven approaches achieves higher accuracy than purely model-based methods and is less computationally expensive, emerging in EV applications.

“the combination of the model-based approach and data-driven approaches that allows achieving higher accuracy than purely model-based methods and yet being less computationally expensive is emerging in EV applications.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Adaptive filtering techniques include extended Kalman filters (EKF), Unscented Kalman Filter (UKF), and particle filters.

“These are the extended Kalman filters (EKF) [75], [76], [77], [78], Unscented Kalman Filter (UKF), [79], [80] and particle filters [81] of the adaptive filtering techniques.”

• cites [75] 75. P. Takyi-Aninakwa, S. Wang, H. Zhang, X. Yang, and C. Fernandez, "An optimized long short-term memory-weighted fading extended Kalman filtering model with wide temperature adaptation for the state of charge estimation of lithium-ion batteries," Appl Energy, vol. 326, p. 120043, Nov. 2022, doi: 10.1016/J.APENERGY.2022.120043. sourceNot checked — no full text of the source
• cites [76] 76. E. P. Sangeetha, N. Subashini, T. K. Santhosh, A. L. S., and D. Uma, "Validation of EKF based SoC estimation using vehicle dynamic modelling for range prediction," Electric Power Systems Research, vol. 226, p. 109905, Jan. 2024, doi: 10.1016/J.EPSR.2023.109905. sourceNot checked — no full text of the source
• cites [77] 77. F. Ciortea, C. Rusu, M. Nemes, and C. Gatea, "Extended kalman filter for state-of- charge estimation in electric vehicles battery packs," Proceedings -2017 International Conference on Optimization of Electrical and Electronic Equipment, OPTIM 2017 and 2017 Intl Aegean Conference on Electrical Machines and Power Electronics, ACEMP 2017, pp. 611-616, Jul. 2017, doi: 10.1109/OPTIM.2017.7975036. sourceNot checked — no full text of the source

May combine two claims.

Particle filters can give real-time estimation but are constrained by high computational burden.

“Particle Filters are also capable of giving real-time estimation but are constrained by high computational burden.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

AI and machine learning are crucial components of state-of-the-art BMS because they allow improved real-time SOC computation.

“AI and machine learning have become crucial components of the state-of-the-art BMS since they allow for improved computation of the SOC of the battery in real time.”

Classified as probably not a substantial claim. May combine two claims.

Battery capacity decays over time, impacting SOC estimation accuracy.

“the battery capacity decays over time hence impacts the accuracy of SOC estimation when using this kind of battery.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Model-based techniques such as equivalent circuit models are reasonably accurate in most cases, but are not highly accurate under dynamic conditions because of parameter changes.

“The model-based techniques like equivalent circuit models (ECMs) are reasonable accurate in most cases although they are not highly accurate under dynamic conditions in view of parameter changes.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Coulomb counting is prone to cumulative errors and requires an accurate initial SOC.

“Prone to cumulativ e errors, requires accurate initial SOC”

May combine two claims.

No single SOC estimation method suits all kinds of batteries; each method has its own characteristic advantages and drawbacks.

“Either of the methods has its characteristic advantages and drawbacks, and, in general, none of the methods suits all kinds of batteries.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Cloud BMS can enable online supervision and regulation of battery systems at various sites.

“Another new frontier is cloud BMS that can enable online supervising and regulating of battery systems at various sites.”

May combine two claims.

Hybrid methods need tuning and calibration to balance model-based and data-driven approaches, which can be complex for real-time systems.

“These methods need to be tuned and calibrated with an aim of rationing the strengths between the model based and the data driven based approaches, which can be complex when designing real time systems.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

Accurate current sensing is complex and expensive.

“-Accurate current sensing is complex and expensive”

May combine two claims.

Data-driven techniques demand high computational cost and may not be feasible in low-power devices such as portable or embedded devices.

“these techniques demand high computational cost and may not be feasible when used in a low power device such as portable or embedded devices.”

May combine two claims.

Filter performance depends on the quality of the initial model and the calibration of adaptive parameters, which poses problems in some processes.

“Moreover, the filter's filter performance also depends on the quality of the first model and the calibration of adaptable parameters, which poses problems in some processes.”

• cites [82] 82. H. M. Fahmy, R. A. Swief, H. M. Hasanien, M. Alharbi, J. L. Maldonado, and F. Jurado, "Hybrid State of Charge Estimation of Lithium-Ion Battery Using the Coulomb Counting Method and an Adaptive Unscented Kalman Filter," Energies 2023, Vol. 16, Page 5558, vol. 16, no. 14, p. 5558, Jul. 2023, doi: 10.3390/EN16145558. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S4 (based).

Data-driven methods including ML and NN highly rely on big, diverse data.

“It must be understood that data-driven methods including machine learning and neural networks highly rely on big, diverse data.”

May combine two claims.

Hybrid methods provide better accuracy and flexibility but have more implementation problems and require more computational resources.

“the hybrid methods, which provide better accuracy and flexibility, include more problems to implement and require more computational resources.”

May combine two claims.

SOC estimation methods are optimal in certain conditions but decline in others, preventing comparison of efficiency.

“Sometimes they are optimal in certain conditions but decline in other conditions and therefore does not allow one to compare the efficiency of the various methods.”

May combine two claims.

As batteries age, charge capacity reduces and internal resistance rises, causing poor SOC measurement if not corrected.

“As batteries are utilized and age, their ability to hold charge reduces and the internal resistance rises if not corrected cause poor SOC measurement.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Open-circuit-voltage-based SOC estimation is only accurate at rest and cannot be used during charging or discharging.

“Only accurate at rest, cannot be used during charging/ dischargin g”