High-power Lithium-ion Batteries Research Articles

Phase change material based passive battery thermal management system to predict delay effect

In the modern scenario of the automobile industry there is a translation from conventional fuel vehicles to electric vehicles as a set to reduce carbon percentage. High power Lithium-ion batteries are the main power source of this vehicle to drive the powertrain, but due to the natural influence of various hot environmental conditions, there is a significant reduction in mileage of vehicle which is not desirable because a lithium-ion battery pack covers 50 to 60% whole cost of vehicle. To tackle this problem from recent years there are several types of research have been performed over active and passive thermal management techniques in which passive thermal management system is widely growth from recent years due to its numerous working advantages with the low-cost application. In the present study, we performed PCM based passive thermal management study over 60 Cell 18,650 Lithium-ion battery packs which are considered to be working for automobile applications. The high-power energy density battery is submerged with a layer of PCM to investigate the delay effect caused by the battery pack to resists the set limit of the threshold temperature range. The output field results from numerical investigation are predicted using neural network approach in which a multi-objective optimization strategy is proposed between battery pack delay effect for selected paraffin wax and RT-18 PCM against its given C-rate. furthermore, liquid fraction rate is monitored for the fixed temperature limit in varying C-rate condition, for optimization strategy Fixed temperature is taken as a function for time delay by PCM addition with liquid fraction rate of PCM over the varying C-rate condition. Correlations suggest paraffin wax is the best choice for a set temperature limit below 60 °C because paraffin wax is found to be melted faster and acts as a resistance to increase the temperature of the cell. The best fit line from the linear regression model using neural network approach suggest best fit line as the time delay is the function of liquid fraction rate over the selected PCM configuration. The highest time delay is occurred in case of paraffin wax PCM as 11,900 s for 1C rate and lowest time delay is occurred as 10,000 s for 1C rate in case of RT-18.

Engineering the architecture and oxygen deficiency of T-Nb2O5-carbon-graphene composite for high-rate lithium-ion batteries

Developing advanced architectures using a cost-effective synthesis strategy is still a challenge for wide-spread commercial application of Nb 2 O 5 in high-power rechargeable lithium-ion batteries (LIBs). Here we report a new two-dimensional (2D) architecture composed of oxygen-vacancy-rich T -Nb 2 O 5 on reduced graphene oxide nanosheet and carbon (2D Nb 2 O 5 -C-rGO), which is synthesized via a one-pot hydrolysis route followed by a heat-treatment. As an anode for LIBs, the 2D Nb 2 O 5 -C-rGO architecture shows excellent rate capability (achieving a capacity of 114 mAh g −1 at 100 C or 20 A g −1 ) and cycling stability (maintaining a capacity of 147 mAh g −1 at 5 C for 1,500 cycles and 107 mAh g −1 at 50 C for 5,000 cycles). Experimental investigations and density functional theory (DFT)-based calculations reveal that the outstanding Li + storage performance of the 2D Nb 2 O 5 -C-rGO electrode is attributed to the enhanced electronic conductivity facilitated by the C-rGO electronic network and fast Li + migration within small Nb 2 O 5 grains enhanced by in-situ formed lattice oxygen vacancies, which alter the Nb d band structure and Li + interaction. This work results in an anode with advanced architecture for fast Li + storage and provides more insight into the energy storage mechanism in the Nb 2 O 5 -based carbonaceous composite electrodes. • A 2D T -Nb 2 O 5 -C-rGO composite anode was synthesized via a cost-effective route. • The architecture and oxygen deficiency were co-engineered by the synthesis route. • High electronic and ionic conductivities were achieved. • The anode delivered a superior rate-capability and stability for Li + storage.

Ultrafine hollow Fe3O4 anode material modified with reduced graphene oxides for high-power lithium-ion batteries

In this work, a composite material consisting of ultrafine Fe3O4 with (ca. 15 nm) uniformly anchored in reduced graphene oxides (rGO) is synthesized and reported to be applied in anodes for lithium-ion battery. The obtained hollow Fe3O4/rGO composites (H-Fe3O4/rGO) (827.3 mA h g−1) has much superior specific capacity than that of the solid Fe3O4/rGO composites (S-Fe3O4/rGO) (654.9 mA h g−1) after 550 cycles of the coin cell at 0.5 A g−1. Even if the cells are tested at 1 A g−1, H-Fe3O4/rGO displays a remarkable specific capacity of 917.4 mA h g−1 as well as the capacity retention rate of 56.7%, compared with S-Fe3O4/rGO (573.9 mA h g−1, the capacity retention rate of 37.7%). Such a sufficient cyclic stability and rate capability in the H-Fe3O4/rGO composite is mainly attributed to the ultrafine hollow structure for shortening Li+ diffusion path and conductive rGO beneficial to rapid electron transfer.

Clarification of particle size dependence on the rate capabilities of Li[Ni1/2Mn3/2]O4 materials and electrodes by the dilute electrode method

Improving the rate capability of lithium insertion electrodes is essential for high-power lithium-ion batteries used in automobile applications. Several approaches have been studied to improve the rate capability of Li[Ni 1/2 Mn 3/2 ]O 4 (LiNiMO); however, no consensus has been reached owing to conflicting results of particle size dependence on rate capability. Rate capability improvement has been obtained previously using nanosized particles as well as micron-sized particles. To solve this contradiction, we examined the dependence of particle size on rate capabilities of LiNiMO materials and electrodes using the dilurted electrode method. This method allows us to measure two distinct rate capabilities of lithium insertion materials and electrodes. The rate capability of dilute-LiNiMO electrodes is improved by using nanosized particles, while that of conventional electrodes is improved by using micron-sized particles. This indicates two different rate-determining steps: Li-ion transport in LiNiMO particles for the dilute electrodes, and Li-ion diffusion in the electrolyte for the conventional LiNiMO electrodes. Moreover, the rate capability of the dilute-LiNiMO electrodes is superior to that of the conventional electrodes, indicating that Li-ion transport in LiNiMO particles is faster than Li-ion diffusion in an electrolyte with porous electrodes. These findings provide important insights for designing electrodes with excellent rate capabilities. • Two types of rate capabilities of Li[Ni 1/2 Mn 3/2 ]O 4 materials and electrodes were measured. • The rate capability of dilute electrodes is improved by using nanosized particles. • The rate capability of conventional electrodes is improved by using micro-sized particles. • The rate capability of the dilute LiNiMO electrodes is superior to that of the conventional electrodes. • Li-ion transport in LiNiMO particles is faster than Li-ion diffusion in an electrolyte with porous electrodes.

A Study to Explore the Suitability of LiNi0.8Co0.15Al0.05O2/Silicon@Graphite Cells for High-Power Lithium-Ion Batteries.

Silicon–graphite (Si@G) anodes are receiving increasing attention because the incorporation of Si enables lithium-ion batteries to reach higher energy density. However, Si suffers from structure rupture due to huge volume changes (ca. 300%). The main challenge for silicon-based anodes is improving their long-term cyclabilities and enabling their charge at fast rates. In this work, we investigate the performance of Si@G composite anode, containing 30 wt.% Si, coupled with a LiNi0.8Co0.15Al0.05O2 (NCA) cathode in a pouch cell configuration. To the best of our knowledge, this is the first report on an NCA/Si@G pouch cell cycled at the 5C rate that delivers specific capacity values of 87 mAh g−1. Several techniques such as X-ray diffraction (XRD), scanning electron microscopy (SEM), electrochemical impedance spectroscopy (EIS) and gas chromatography–mass spectrometry (GC–MS) are used to elucidate whether the electrodes and electrolyte suffer irreversible damage when a high C-rate cycling regime is applied, revealing that, in this case, electrode and electrolyte degradation is negligible.

Ionic Liquid-Mediated Mass Transport Channels for Ultrahigh Rate Lithium-Ion Batteries.

Excellent mass transport capability is indispensable to building a high-power lithium-ion battery (LIB) system. Nanomaterials with enhanced electrochemical properties have been used for next-generation high-performance LIBs. However, due to the high surface free energy, nanomaterials tend to form agglomeration. The resulting insufficient mass transport channels limit the high rate performance of nanomaterials. Here, we increase the electrolyte accessibility of nanomaterials through a facile ionic liquid (IL) mediation method. The fluidity and affinity enable the IL to infiltrate into the interstitials of nanomaterial aggregations under capillary force, enhancing the electrochemically active contact area and ensuring rapid mass transport. As a proof of concept, IL-mediated LiFePO4 electrodes delivered extraordinary rate performance (112 and 95 mAh g-1 at 200 and 300 C, respectively). In terms of simplicity, the IL mediation can be used as a general strategy to achieve an ultrahigh rate for LIBs.

A Survey of Wireless Battery Management System: Topology, Emerging Trends, and Challenges

An effective battery management system (BMS) is indispensable for any lithium-ion battery (LIB) powered systems such as electric vehicles (EVs) and stationary grid-tied energy storage systems. Massive wire harness, scalability issue, physical failure of wiring, and high implementation cost and weight are some of the major issues in conventional wired-BMS. One of the promising solutions researchers have come up with is the wireless BMS (WBMS) architecture. Despite research and development on WBMS getting momentum more than a decade ago, it is still in a preliminary stage. Significant further upgradation is required towards developing an industry-ready WBMS, especially for high-power LIB packs. Therefore, an in-depth survey exclusively on WBMS architectures is presented in this article. The aim is to provide a summary of the existing developments as well as to present an informative guide to the research community for future developments by highlighting the issues, emerging trends, and challenges. In-depth analysis of the existing WBMS topologies will not only help the researchers to understand the existing challenges and future research scopes clearly but at the same time enthuse them to focus their research inclination in the domain of WBMS.

Enhanced rate capability of Li4Ti5O12 anode material by a photo-assisted sol–gel route for lithium-ion batteries

Photo-assisted sol–gel route was developed to prepare nanostructured Li4Ti5O12 anode materials. Resulting Li4Ti5O12 powders were characterized by phase analysis, microscopic morphology observation, and cyclic charge/discharge tests. As lithium-ion battery electrode, synthesized Li4Ti5O12 material by using the photo-assisted sol–gel route shows specific capacity of 120 mAh g−1 at 10C discharge rate. Furthermore, this material retains 98% of its initial specific capacity at 0.2C after 50 charge/discharge measurements. Outstanding rate capability and cycling stability of this Li4Ti5O12 material are mainly derived from grain refinement induced by ultraviolet photo-irradiation. Therefore, this photo-assisted sol–gel route is attractive alternative method for preparing electrode materials for high-power lithium-ion battery applications.

Aging investigations and consideration for automotive high power lithium-ion batteries in a 48\xa0V mild hybrid operating strategy

This paper focuses on the battery aging of automotive high power lithium-ion batteries intended for 48 V mild hybrid systems. Due to a long vehicle lifetime, battery aging is of high importance, and its consideration within a hybrid system is crucial to ensure a sufficient lifetime for the battery. At the moment, only a few aging investigations and models specifically for automotive high power cells are available. Consequently, all present aging consideration methods are based on the few published aging models focusing on consumer cells. This paper describes the development of an aging model for automotive high power cells and the integration into a mild hybrid operating strategy to actively control the battery aging process during its operation. The underlying aging investigations of high-power battery cells are shown to analyze the main influences of temperature, state of charge, and C-rate. These tests are used to develop the aging model, capable of considering the main influences on the aging process. Based on this model and all gained insights, different methods for considering battery aging in a mild hybrid system are investigated. The goal is to control the aging process during operation and consequently decrease the negative influence. Two active intervention methods are developed and integrated into a 48 V mild hybrid operating strategy to validate their potential. It is possible to control the aging process and at the same time to use the insights for improving the basic hybrid powertrain design regarding reduced aging and battery costs.

FeNb11O29, anode material for high-power lithium-ion batteries: Pseudocapacitance and symmetrisation unravelled with advanced electrochemical and in situ/operando techniques

In the last years of research on new anode materials for advanced Lithium-Ion Batteries (LIBs), niobium-based oxides are raising growing attention due to very high theoretical capacities and high working potential that can prevent the formation of lithium dendrites, ensuring the safety of the batteries. Most of them crystallizes in open Wadsley–Roth shear structures, which show large Li-ion diffusion coefficients and promising applications in energy storage systems. In this paper, the complex structural and electrochemical features of FeNb11O29 are unravelled with in situ Raman Spectroscopy, operando X-Ray Diffraction and electrochemical techniques. The intrinsic pseudocapacitance shown by the iron niobate is correlated to large channels in its structure that cause weak Li+-host interactions and very little charge-transfer resistances. The symmetrisation of the octahedral framework that occurs after the reduction of Nb5+ cations, detected for the first time, seems to be the key of the electrochemistry of FeNb11O29, which shows excellent features for advanced high-power density LIBs.

Facile route to biomass-derived 1D carbon fiber supported high-performance MnO-based nanocomposite anode material

The combination of carbon fiber and MnO is considered as a promising strategy to enhance the electrochemical performance of MnO. Herein, the carbon fiber/MnO/C composite material was fabricated using cotton with pyrrole as the carbon source through two-step pyrolysis with KOH as the activator. The irregular shaped MnO particles with a carbon coating of about 1 nm thickness were embedded on the surface of the carbon fiber. Benefiting from this unique structure, the composite material demonstrated remarkable electrochemical performance. It retained 35.3% of the initial discharge capacity (1239 mA·h·g−1) at 0.1 A g−1. Under the stepwise increase of the current densities from 0.1 to 3 A g−1, a superior rate capability was obtained compared to that for MnO/C (30.2%) and MnO (26.4%). Moreover, the material manifested an outstanding long-term tolerance over 200 cycles. This excellent electrochemical performance is on account of the one-dimensional carbon fiber and the surrounding carbon coating, which not only enhances the electronic connectivity and conductivity but also facilitates the transportation of lithium ions as well as preventing the agglomeration of MnO nanoparticles during the electrochemical cycle. This work introduces a facile scalable approach of utilizing sustainable biomass materials to make electrodes for high-power lithium-ion batteries.

Smart construction of Ti2Nb10O29 /carbon nanofiber core-shell composite arrays as anode materials for lithium-ion batteries

With high capacity and its safety, titanium niobium oxide (Ti2Nb10O29, TNO) has been reported as a potential anode for high power lithium ion batteries. However, insufficient diffusion coefficient of Lithium-ions and poor electronic conductivity for this electrode material significantly constrains its rate capability. To solve this issue, we propose a simple strategy to encapsulate Ti2Nb10O29 in porous carbon nanofiber arrays (porous TNO/CNF) and to prepare heterogeneous Ti2Nb10O29/carbon nanofiber core-shell composite arrays (TNO NPs-CNF) using polyamic acid (PAA) as the polymer precursor. TNO NPs-CNF electrode exhibits significantly superior electrochemical performance (286 mA h g−1 at 10 C and 225 mA h g−1 at 20 C), much better than the TNO and porous TNO/CNF. For the electrode of porous TNO/CNF and TNO NPs-CNF, the omni-directional conducting networks of CNF skeleton greatly enhances electron transport of the whole electrode, reduces energy barrier of lithium ion transport, and maintains the structure’s stability.

A novel streamlined particle‐unscented Kalman filtering method for the available energy prediction of lithium‐ion batteries considering the time‐varying temperature‐current influence

Effective energy prediction is of great importance for the operational status monitoring of high-power lithium-ion battery packs. It should be embedded in the battery system performance evaluation, energy management, and safety protection. A new Streamlined Particle-Unscented Kalman Filtering method is proposed to predict the available energy of lithium-ion batteries, in which an Adaptive-Dual Unscented Transform treatment is conducted to realize the precise mathematical expression of its working conditions. For the accurate mathematical description purpose, an improved Synthetic-Electrical Equivalent Circuit modeling method is introduced into the internal effect equivalent process considering the influence of time-varying temperature and current conditions. As can be known from the experimental results, the proposed prediction method has a maximum estimation error of 2.27% and an average error of 0.80%, for the complex varying-current Beijing Bus Dynamic Stress Test. Under the Urban Dynamometer Driving Schedule working conditions, the available energy prediction has high accuracy with a maximum error of 1.83% and a voltage traction error of 3.28%. It provides vehicle-mounted available energy prediction schemes for effective management and safety protection of high-power lithium-ion batteries. Highlights A new Streamlined Particle-Unscented Kalman Filtering method is proposed to predict the available energy of lithium-ion batteries. Improved Synthetic-Electrical Equivalent Circuit modeling strategies are established to describe the nonlinear battery characteristics. Adopted predictive correction is investigated by considering the time-varying temperature and current influence. For effective convergence, an adaptive windowing function factor is introduced into the correction process with a maximum estimation error of 2.27% and an average error of 0.80% for the complex varying-current Beijing Bus Dynamic Stress Test working conditions. The vehicle battery available energy prediction is realized with a maximum error of 1.83% and a maximum voltage traction error of 3.28% for the Urban Dynamometer Driving Schedule working conditions.

Simple Strategy for Synthesizing LiNi0.8Co0.15Al0.05O2 Using CoAl-LDH Nanosheet-Coated Ni(OH)2 as the Precursor: Dual Effects of the Buffer Layer and Synergistic Diffusion.

Ni-rich layered oxide LiNi0.8Co0.15Al0.05O2 is a promising cathode material for high-power lithium-ion batteries due to its high energy density and low cost. However, obtaining LiNi0.8Co0.15Al0.05O2 with a large and uniform particle size and without undesired Al-related phases using some conventional synthesis methods is quite difficult. These problems seriously affect the electrochemical performance of LiNi0.8Co0.15Al0.05O2, thus impeding its wide application. Here, we propose a simple strategy to synthesize LiNi0.8Co0.15Al0.05O2 using CoAl-layered double hydroxide (CoAl-LDH) nanosheet-coated Ni(OH)2 as the precursor. Compared with LiNi0.8Co0.15Al0.05O2 synthesized from nickel-cobalt-aluminum hydroxide and Al(OH)3-coated nickel-cobalt hydroxide precursors, LiNi0.8Co0.15Al0.05O2 produced using the proposed approach showed good sphericity, a large and uniform particle size, a pure phase, and excellent electrochemical performance. The superior properties are attributed to the dual effects of the buffer layer and synergistic diffusion. Specifically, the CoAl-LDH coating layer reacts with LiOH during the lithiation-calcination process to form a Li1-x(Co0.75Al0.25)1+xO2 mesophase as the buffer layer, which increases the formation temperature of the layered structure and reduces Li+/Ni2+ cation mixing, making a well-ordered crystal structure. Moreover, spectroscopic analysis results and density functional theory calculations indicated a synergistic diffusion effect between Co and Al, and the presence of Co on the surface promotes the diffusion of Al during the lithiation-calcination process, thus avoiding the formation of undesired Al-related phases and allowing for a uniform element distribution.

Li-incorporated porous carbon monoliths derived from carboxymethyl cellulose as anode material for high power lithium-ion batteries

In this study, a simple and effective pre-lithiation strategy is developed to resolve a large initial irreversible capacity loss and low initial Coulombic efficiency of porous carbon as anode material for Li-ion batteries. By using this strategy, Li-incorporated hierarchically porous carbon monoliths (PCMs) are successfully obtained from water-soluble carboxymethyl cellulose lithium salt. The prepared PCMs are found to have amorphous carbon structures and well-developed hierarchical pore architectures. In addition, Li is found to be homogeneously incorporated throughout the prepared PCMs. The electrochemical properties of the prepared Li-incorporated PCMs are studied in a half-cell configuration and exhibit lower initial irreversible capacity along with stable cycle life at high current densities. Generally, the resistance of LIB cell remarkably increases at the end of discharge. Here, the galvanostatic intermittent titration technique analyses reveal a higher diffusion coefficient of the prepared Li-incorporated PCMs, specifically at the end of discharge. Therefore, this inherent property of Li-incorporated PCMs is favorable for high-power LIB anode materials.

Insight into the Redox Reaction Heterogeneity within Secondary Particles of Nickel-Rich Layered Cathode Materials.

Nickel-rich LiNixCoyMn1-x-yO2 (nickel-rich NCM, 0.6 ≤ x < 1) cathode materials suffer from multiscale reaction heterogeneity within the electrode during the electrochemical energy storage process. However, owing to the lack of appropriate diagnostic tools, the systematic understanding and observation on the redox reaction heterogeneity at the individual secondary-particle level is still limited. Raman spectroscopy can not only reflect the depth of the redox reaction through probing the vibrational information on the metal-oxygen coordination structure but also sensitively detect the local structure changes of different regions within the secondary particle with suitable spatial resolution. Therefore, Raman spectroscopy is applied here to conveniently conduct the high-resolution and in-depth analysis of the rate-dependent reaction heterogeneity within nickel-rich NCM secondary particles. It is found that, under high-rate conditions, the oxidation/reduction reaction mainly occurs in the surface region of the particles and the cause of this particle-scale reaction heterogeneity is the limitation of the slow solid-phase Li+ diffusion and the transient charging/discharging processes. In addition, this reaction heterogeneity would aggravate the structural instability of the material continuously during the charging/discharging cycles, thus resulting in a slowdown in the kinetics of Li+ de/intercalation and the apparent capacity decay. This work can not only provide fundamental insight into the rational modification of high-power nickel-rich NCM materials but also guide the setting of electrochemical operating conditions for high-power lithium-ion batteries (LIBs).

LiFePO4/Carbon Nanomaterial Composites for Cathodes of High-Power Lithium Ion Batteries

LiFePO4/Carbon Nanomaterial Composites for Cathodes of High-Power Lithium Ion Batteries

Hierarchically porous MXene decorated carbon coated LiFePO4 as cathode material for high-performance lithium-ion batteries

As a promising cathode material for high-power lithium-ion batteries, LiFePO4 (LFP) suffers from low lithium-ion diffusivity and poor electronic conductivity. In this work, hierarchically porous Ti3C2Tx MXene decorated LFP@C (LFP@C/MXene) composite was successfully synthesized by a facile and efficient electrostatic self-assemble method at room temperature with the help of CTAB to regulate charge state. The introduction of MXene sheets realized the formation of hierarchically porous structure and “dot-to-surface” conductive network, enabling the fast ion and electrons transfer for redox reactions, meanwhile, MXene will be partially oxidized into TiO2 and carbon during cycling, which further improved the diffusion of lithium-ion. Together with the optimized structure of LFP@C nanoplates, the resulting LFP@C/MXene exhibits robust high-rate capability (139 mAh·g−1 at 20 C) as well as a long-life cycling stability (156.6 mAh·g−1 at 1 C with superior capacity retention of 94.8% for 500 cycles). This ingenious design highlights the great potential of 2D MXene for facilely constructing multifunctional electrode materials for application in high-rate Li-ion batteries.

Ultra-Thin Mesoporous LiV3O8 Nanosheet with Exceptionally Large Specific Area for Fast and Reversible Li Storage in Lithium-Ion Battery Cathode

LiV3O8 is considered as a promising cathode material for lithium-ion batteries (LIBs) because of its high specific capacity and good safety. However, the poor cycling stability and low rate capability of LiV3O8 electrode seriously hinder its application in long lifespan and high power LIBs. In the present work, an ultra-thin mesoporous LiV3O8 nanosheet is synthesized on the basis of efficient reduction-oxidation and freeze-dried strategy. The mesoporous nanosheet with a very small thickness of about 1.36 nm has an exceptionally large surface area of 397 m2 g−1 (8–20 times of LiV3O8 reported). The electrochemical test results indicate that the nanosheet cathode can deliver a reversible capacity of 301 mA h g−1 at current density of 100 mA g−1 in the potential range of 2.0–4.0 V and retain 87.1% of the starting discharge capacity after 550 discharge/charge cycles at 2000 mA g−1. The excellent lithium storage performances are ascribed to its unique flexible ultra-thin mesoporous nanosheet structure, which can increase the surface pseudocapacitive lithium storage, reduce the charge transfer resistance, improve the diffusion coefficient of lithium ion, and accommodate the volume changes of LiV3O8 in the discharge-charge processes. These results demonstrate a great potential of ultra-thin mesoporous LiV3O8 nanosheet for large capacity, long life and high power LIBs.

Rapid Model-Free State of Health Estimation for End-of-First-Life Electric Vehicle Batteries Using Impedance Spectroscopy

The continually expanding number of electric vehicles in circulation presents challenges in terms of end-of-life disposal, driving interest in the reuse of batteries for second-life applications. A key aspect of battery reuse is the quantification of the relative battery condition or state of health (SoH), to inform the subsequent battery application and to match batteries of similar capacity. Impedance spectroscopy has demonstrated potential for estimation of state of health, however, there is difficulty in interpreting results to estimate state of health reliably. This study proposes a model-free, convolutional-neural-network-based estimation scheme for the state of health of high-power lithium-ion batteries based on a dataset of impedance spectroscopy measurements from 13 end-of-first-life Nissan Leaf 2011 battery modules. As a baseline, this is compared with our previous approach, where parameters from a Randles equivalent circuit model (ECM) with and without dataset-specific adaptations to the ECM were extracted from the dataset to train a deep neural network refined using Bayesian hyperparameter optimisation. It is demonstrated that for a small dataset of 128 samples, the proposed method achieves good discrimination of high and low state of health batteries and superior prediction accuracy to the model-based approach by RMS error (1.974 SoH%) and peak error (4.935 SoH%) metrics without dataset-specific model adaptations to improve fit quality. This is accomplished while maintaining the competitive performance of the previous model-based approach when compared with previously proposed SoH estimation schemes.