Skip to main content
NCBI home page
Scientific Data logoLink to Scientific Data
. 2025 Dec 10;12:1942. doi: 10.1038/s41597-025-06229-5

Lithium-Ion Battery Pack Cycling Dataset with CC-CV Charging and WLTP/Constant Discharge Profiles

Joaquín de la Vega Hernández 1, Juan A Ortega-Redondo 1, Jordi-Roger Riba 2,✉
PMCID: PMC12706042  PMID: 41372221

Abstract

This work presents a database of a lithium-ion battery pack cycling tests generated from a custom test bench that simulates dynamic driving conditions based on the WLTP cycle. Current profiles were derived from speed-time data using MATLAB/Simulink and a Tesla Model 3 vehicle model. The dataset includes time series data on cell voltages, currents, surface temperatures, and pack-level resistance from up to 36 cells arranged in three parallel branches. The data is recorded under controlled thermal conditions and stored in an efficient PARQUET format. The system uses a controller area network (CAN) bus architecture and commercial automotive battery management system (BMS) units to replicate in-vehicle communication constraints, enhancing the dataset’s relevance for real-world battery management system development and validation.

Subject terms: Electrical and electronic engineering, Energy efficiency

Background & Summary

Lithium-ion batteries are a key enabling technology for electric vehicles (EVs)1 due to their high energy density, efficiency, and rechargeability. However, their performance and safety are not static; they evolve over time under varying thermal, electrical, and mechanical conditions.

As batteries undergo cycling and are exposed to different operating environments, they inevitably degrade, which reduces their capacity, increases their internal resistance, and limits their power output2. These processes are driven by electrochemical, thermal, and mechanical mechanisms, including solid electrolyte interphase (SEI) layer growth, active material loss, lithium plating, electrolyte decomposition, and electrode structural fatigue3. The rates of these degradation processes are strongly influenced by operational factors including temperature, state of charge (SOC) working range, depth of discharge, and current profiles4–6. Over time, these changes reduce the available energy and alter the thermal characteristics and safety margins, potentially compromising the reliability and lifespan of electric vehicle battery packs7,8.

To develop reliable models, control strategies, and lifetime predictions, researchers need data that accurately reflects the complex conditions encountered during actual EV operation9. Real driving cycles are subjected to driver, environment, traffic and other conditions. These circumstances result in nonlinear current demands, frequent regenerative braking events, and dynamic thermal profiles that cannot be replicated using simple laboratory tests. Standardized driving cycles, such as the worldwide harmonized light vehicles test procedure (WLTP), are useful tools for producing comparable data under consistent conditions without compromising the dynamics of a driving cycle. These profiles introduce time-varying demands that more accurately capture the stress conditions batteries experience during real-world driving.

Although several battery cycling datasets are publicly available10,11, most focus on constant current (CC) profiles or specialized diagnostic studies such as failure analysis. Existing datasets often lack temperature measurements from multiple locations within each cell, which limits their usefulness for analyzing thermal behavior in detail. Furthermore, very few datasets provide measurements under standardized dynamic cycles with realistic system constraints. Additionally, few studies incorporate realistic vehicular communication architectures into their data acquisition systems. The dataset presented here helps address these gaps by offering comprehensive time series data collected under WLTP-based operation. This allows researchers to analyze battery behavior under profiles that are directly linked to standardized driving patterns. Consequently, it facilitates model validation, thermal management research, and life cycle assessment studies that are directly relevant to electric vehicle (EV) conditions.

The battery’s state of health (SOH) depends on a combination of physical, electrical, and thermal variables that change over time as the battery cycles12. Capacity fade, resistance increase, and thermal behavior are all influenced by measurable parameters such as voltage, current, temperature, and internal resistance. Accurate tracking of these variables is therefore crucial for applications such as thermal and energy management13, aging diagnostics, and predictive modeling. Large thermal gradients can decrease local cell impedance14, potentially causing uneven current distribution within a branch. If these imbalances are not managed properly, they can accelerate the degradation of specific cells, which ultimately reduces the usable capacity and reliability of the entire battery pack. Imbalanced aging and rapid capacity fade, for example, may result from non-uniform temperature distributions across cells or modules. These distributions are often caused by resistive heat generation during dynamic cycling. The presented dataset allows for the study of these thermal-electrical interactions and their influence on state of health (SOH) evolution by providing temperature measurements at multiple points on each cell under controlled environmental conditions. This detailed data supports the calibration of accurate thermal models, the development of advanced thermal management and balancing strategies, and the assessment of degradation mechanisms under realistic and dynamic operating profiles15,16.

In real EV systems, battery information is not acquired directly from laboratory instruments, but rather transmitted through a battery management system (BMS) via a controller area network (CAN) bus17. The BMS continuously monitors cell voltages, currents, and temperatures. It provides crucial protection functions and information about the state of the battery pack to ensure safe operation under various load conditions18. This architecture imposes practical limitations on data resolution, frequency, and accessibility. These limitations are often overlooked in experimental datasets. To address these limitations and the previously described challenges, a custom battery test bench was developed that integrates real automotive BMS units and CAN-based communication. This allows working with data that reflects realistic sampling constraints and signal update frequencies while maintaining high measurement accuracy.

This test bench can perform constant and dynamic charge/discharge cycles, as well as simulate standardized driving scenarios such as the WLTP, all under thermally controlled conditions. It integrates temperature chamber control, multi-point thermal sensing, and automated safety protections to ensure reliable, repeatable testing. The focus of this work is to publish the database generated by the test bench, which captures synchronized voltage, current, temperature, and resistance data from up to 36 lithium-ion cells arranged in a parallel-series configuration. The dataset is intended to support research in battery modeling and degradation analysis. It is a valuable resource for advancing the estimation of battery health, the optimization of life cycles, and the simulation of electric vehicle powertrains under standardized, realistic conditions.

Methods

Cell description

The dataset includes measurements taken from 36 commercially available lithium-ion batteries, which are arranged in three sets of 12 cells each. Table 1 summarizes the specifications of the selected cells19.

Table 1.

Information regarding the cells utilized in the database.

Manufacturer Panasonic
Model NCR18650B
Cathode type NCA (nickel-cobalt-aluminum)
Capacity (rated) 3200 mAh (20 °C)
Voltage (nominal) 3.6 V
Charge temperature 0 °C to + 45 °C
Charge voltage 4.2 V (CC-CV)
Charge C-rate 25 °C: 0.5 C (max); <10 °C: 0.25 C
Charge cut-off 65 mA, 4 h
Discharge temperature −20 °C to +60 °C
Discharge cut-off 2.5 V
Open in a new tab

As shown in Table 1 and according to the manufacturer’s recommendations, the required charging profile for the cells is a constant current–constant voltage (CC-CV) mode. This means the charging voltage per cell should be 4.2 V, which should be maintained until the current flowing into the cell drops to 0.065 A. In contrast, the voltage should not decrease below 2.5 V during discharge to avoid damaging the cells.

Within each branch, the cells are numbered from 1 (most negative) to 12 (most positive). The branches are labeled sequentially from 1 to 3, and the cells within each branch are labeled from 1 to 12. The naming convention uses P# (for parallel) to indicate the branch and S# (for series) to identify the specific cell within that branch. For example, the eighth cell in the second branch is labeled P2S8. This numbering system is illustrated in Fig. 1.

Fig. 1.

Fig. 1

Open in a new tab

Identification of each cell and branch in the battery pack.

Driving cycle simulation

To accurately simulate real-world battery demands, the worldwide harmonized light vehicles test procedure (WLTP) driving cycle was chosen. Since the WLTP is defined in terms of vehicle speed versus time20, it was first converted into a current versus time profile suitable for battery testing.

The vehicle power profile was obtained using classical analytical equations that describe all forces acting on the vehicle. These equations can be found in21 and were integrated into the Powertrain Blockset model22 from MATLAB/Simulink version 2024b. This model directly provides the instantaneous power required by the vehicle. Within the simulation environment, a model of a Tesla Model 3 was constructed, incorporating relevant dynamic parameters, such as vehicle mass, aerodynamic drag coefficient, rolling resistance, braking force distribution, and drivetrain efficiency23,24. The data from the Tesla Model 3 were used to generate the profiles while following WLTP driving cycles. Next, the total instantaneous current required by the vehicle is determined from the instantaneous power and the available voltage provided by the battery pack, and is scaled within the 0–1 A interval. This current profile was linearly scaled so that the current per battery pack branch does not exceed 1.6 A (0.5 C-rate) during charging and 6.4 A (2 C-rate) during discharging, in accordance with the manufacturer’s recommendations. For security reasons, the test bench dynamically adjusts the total current according to the number of active branches so that the scaled current of one branch is multiplied by the number of active branches.

This conversion was performed using MATLAB and Simulink version 2024b22, leveraging the Powertrain Blockset.

The resulting output current profile represents the vehicle’s power demands throughout the WLTP cycle. It was exported and used to drive the real battery system under test. This ensured that the electrical load applied to the battery pack closely mimicked real-world driving conditions.

System overview

A Python 3 script was developed to coordinate the various components of the test bench and ensure reliable operation during extended testing periods. Leveraging the host computer’s multiple processing cores, the script runs concurrent tasks that manage power source control, data acquisition, file storage, and overall system supervision. This design enables real-time monitoring, responsive control actions, and uninterrupted data logging, even during long cycling sessions. Built-in checks and exception-handling routines ensure system stability under variable conditions. Figure 2 presents an overview of the test bench architecture.

Fig. 2.

Fig. 2

Open in a new tab

System overview.

Power control system

We use a programmable, bidirectional ITECH IT3632 DC power source (version 1.1/03. 2021, voltage accuracy <0.06 V, current accuracy ±1% + 0.3 A, ITECH, Taipei City, Taiwan) to apply current profiles derived from the WLTP cycle. As shown in Fig. 3, the power source is controlled via Standard Commands for Programmable Instruments (SCPI), which are sent from the main computer. This enables the precise, automated execution of dynamic load steps during testing.

Fig. 3.

Fig. 3

Open in a new tab

Power source, main computer and CAN interface module.

Sensing and communication infrastructure

Dedicated instrumentation was used to measure the voltage, current, temperature, resistance, and humidity at multiple points within the system. Three battery BMSs (R24, ±1.5 mV, Wattius, Granollers, Spain) collected voltage and current data. Seven Arduino UNO R3 microcontrollers (Arduino, Monza, Italy) collected temperature and humidity data. The Arduinos and BMSs then streamed the data through a CAN bus configured for a 500 kbit/s speed, similar to that used in real vehicle applications25. The main computer gathered all the measurements from the CAN bus.

Figure 4 shows the connection of the microcontrollers to the bus. Voltage and resistance at the pack level were collected using an ITECH IT5101 battery tester (version 2.2. 2021, voltage accuracy ±0.01%, resistance accuracy ±0.4%, ITECH Electronic Co., Taipei City, Taiwan), which communicated directly with the main computer via a serial interface.

Fig. 4.

Fig. 4

Open in a new tab

BMSs and Arduino microcontrollers.

Temperature measurement and control

Temperature data were collected using 10 kΩ NTC thermistors. Two thermistors10 kΩ NTC (3964 K, ±1%, TDK, Tokyo, Japan) were attached to each battery cell, one near the positive terminal and one near the negative terminal. This resulted in a total of 72 temperature sensing points within the pack.

Two additional 10 kΩ NTC thermistors (3950 K, ±1%, Handson Technology, Atlanta, GA, USA) were used to monitor the air temperature inside and outside the thermal chamber housing the battery pack. All thermistors were individually calibrated using a CEM BX-150 dry-well calibrator (CEM, 33 °C to 300 °C, ±0.1 °C, Shenzhen, China).

As shown in Fig. 5, all components are placed inside a programmable PREBATEM 150 L (5 °C to 60 °C, ±0.2 °C) temperature chamber from JP-SELECTA (Abrera, Spain). This ensures consistent and controllable ambient conditions during testing and prevents thermal gradients from affecting test accuracy.

Fig. 5.

Fig. 5

Open in a new tab

Controlled temperature chamber (left) and temperature calibrator (right).

Current measurement

Current was measured using Hall-effect transducers (LEM HASS-50S, ≤ ± 1%, LEM, Geneva, Switzerland). As shown in Fig. 6, each branch of the battery pack was equipped with its own sensor to enable independent current monitoring. It also shows the thermistors allocations.

Fig. 6.

Fig. 6

Open in a new tab

Voltage, current and temperature sensors.

Humidity measurement

The humidity inside the thermal chamber was monitored using an digital humidity sensor (AM2302, ±2% RH, Adafruit, NY, USA). This data, along with the temperature readings, was used to characterize the environmental conditions around the battery pack. As shown in Fig. 7, the humidity sensor was placed near the battery pack.

Fig. 7.

Fig. 7

Open in a new tab

Humidity and in-chamber temperature sensor.

Voltage and resistance measurement

The voltage across each individual cell was measured by three dedicated battery management system (BMS) units, one per branch of the battery pack. An ITECH IT5101 internal resistance tester battery tester was connected to the main busbar terminals to measure the overall voltage and resistance of the pack. The instrument communicated directly with the main computer via a serial interface. Figure 8 shows the ITECH IT5101 internal resistance tester and the position of the out-chamber temperature sensor.

Fig. 8.

Fig. 8

Open in a new tab

IT5101 battery tester and out-chamber temperature sensor.

Data Record

The main computer continuously reads data from the CAN bus and connected measurement devices. It stores the results in a structured database as Apache Parquet files. Due to non-uniform sampling and asynchronous data collection, a zero-order hold (ZOH) method is applied between records. Each record includes its own timestamp, but the system is designed to operate at a minimum sampling rate of 2 Hz.

The dataset includes a unique folder containing all the files. Each file contains the data from one charge/discharge cycle. There are two types of cycles: those that start with several WLTP driving cycles (discharge semicycles) followed by a CC-CV charge semicycle, and those that check the capacity of the cells by applying a CC discharge semicycle followed by a CC-CV charge semicycle. Discharge semicycles end when the voltage cutoff of any cell is reached, and charge semicycles end when one of the cutoff thresholds (voltage, current, or time) is reached.

Each file contains the complete data from one charge and discharge semicycles and consists of 132 elements, including:

  • Cycle number: A sequential number that increases at the start of every discharge semi-cycle. Each cycle contains one charge stage and one discharge stage.

  • Timestamp: In ISO 8601 format.

  • Semicycle: This indicates the operational stage of the power supply. It can be one of the following predefined stages:
    • WLTP: The dynamic driving cycle current profile is limited by the specifications of the cells in the battery pack. The cut-off voltage is 38.4 V. The sequence consists of 1800 steps at regular one-second intervals with charge (positive) and discharge (negative) current values. Once the sequence is complete, it restarts and continues until the cutoff voltage is reached.
    • Constant Charge: CC-CV profile. 1 A per active branch in CC. The cut-off current is 0.07 A per active branch. The cut-off voltage is 50.5 V, with a cut-off time of 43200 seconds.
    • Constant Discharge: 1.6 A per active branch. The cut-off voltage is 38.4 V.
    • Capacity Check Charge: CC-CV profile. 0.5 A per active branch in CC. The cut-off current is 0.07 A per active branch. The cut-off voltage is 50.5 V, with a cut-off time of 46080 seconds.
    • Capacity Check Discharge: 0.64 A per active branch. The cut-off voltage is 38.4 V.
    • Fast Charge: CC-CV profile. 1.62 A per active branch in CC. The cut-off current is 0.07 A per active branch. The cut-off voltage is 50.5 V, and the cut-off timestamp is 43200 seconds.
    • Fast Discharge: 4.16 A per active branch. The cut-off Voltage is 38.4 V.
    • Stand-by: There is no current flow from the power source. It is waiting for the main computer to request the next semicycle stage.
    • Balancing: The BMS has detected an imbalance within a branch or between branches. The system attempts to balance itself.
  • Voltage [V]: There are 46 voltage-related measurements, comprising:
    • 36 individual cell voltages
    • 3 branch-level voltages
    • 2 pack-level voltages
    • 5 statistical values. Minimum, maximum and average voltage value across all cells and the minim and maximum voltage value of the whole battery pack.
  • Current [A]: There are 5 current measurements, including:
    • 1 per branch
    • 1 total current through the battery terminals
    • 2 peak values: maximum charge and maximum discharge current
  • Temperature [°C]: There are 74 temperature measurements in °C, including:
    • 72 cell temperatures (top and bottom of each cell)
    • 2 ambient temperatures (inside and outside the thermal chamber)

  • Relative humidity [%RH] within the test chamber.

  • Resistance [Ohm]: The IT5101 battery tester was used to measure resistance at the bus bars where the battery pack is connected.

  • State of Charge (SOC) [%]: The BMS provides the estimation through Coulomb counting, which is complemented by OCV-based and end-of-charge/discharge calibrations.

Note that the Capacity Check Charge and Capacity Check Discharge semicycles are dedicated procedures used to evaluate battery capacity. These procedures are performed after each maintenance break or after a maximum of ten WLTP driving cycles. Each evaluation consists of two full Capacity Check cycles. The balancing semicycle is an automated stage performed by the test bench when the voltage imbalance between cells or branches exceeds 100 mV. This ensures system stabilization before continuing with subsequent cycling.

The README file accompanying the database contains a complete list of the column names and units included in the data files.

Figure 9 shows three measured cycles in which the WLTP (in blue) and some CC-CV charging profiles (in red) are applied to the entire battery pack.

Fig. 9.

Fig. 9

Open in a new tab

Example cycles of the database showing different the WLTP driving cycle and the CC-CV charge profile.

The dataset is publicly available via the CORA.RDR repository26 through the persistent link https://dataverse.csuc.cat/dataset.xhtml?persistentId=doi:10.34810/data2395, a federated platform for research data from Catalan universities and affiliated research institutions in Spain. The dataset is organized into multiple PARQUET files, with each file corresponding to a complete battery cycling period, which include at least a charge and a discharge stage. A detailed README text file accompanies the dataset and describes the data structure, measurement parameters, units, and file contents. Additionally, a CSV file containing the WLTP time series vectors is provided: the original speed profile (in km/h), the derived current profile from the electromechanical Tesla Model 3 simulation (in amperes), and the adapted current profile applied to the tested battery configuration.

The cycling files are ordered by occurrence and follow a consistent naming convention: “Testcode_Cycle_#_CycleType.parquet,” where Testcode relates to a unique test identifier (“Qtzl” for this dataset), # denotes the cycle number, and CycleType describes the type of cycle (WLTP or Capacity Check) This ensures scalability and uniqueness. The initial release includes data from the first 410 cycles. Though file lengths vary, each file contains an average of 66,445 rows and is approximately 1.96 MB in size.

Figure 10 shows how battery capacity changes with the number of cycles. It also shows the end-of-life threshold that signals the end of the tests.

Fig. 10.

Fig. 10

Open in a new tab

Measured battery capacity in blue as a function of cycle number, showing the capacity fade of the battery pack. The expected End-of-Life threshold that determines the end of the tests is shown in red.

The CORA.RDR repository provides metadata about the general information of the dataset, such as the DOI, publication date, title, authors, and more. In addition, information regarding the terms of use and version of the dataset can be found in the same website.

Technical Validation

To ensure the reliability of the dataset, a comprehensive validation process was conducted during system development and data acquisition. All measurements were collected using calibrated sensors and commercial-grade equipment integrated into a test bench that simulated realistic electric vehicle operating conditions. The surface temperature of each cell was monitored using two 10 kΩ NTC thermistors (one on the top and one on the bottom), which were calibrated with a BX-150 dry-well temperature calibrator to ensure high thermal accuracy. This dual-sensor approach enabled precise monitoring of thermal gradients during cycling, which is critical for analyzing safety margins and degradation patterns under dynamic loads.

Hall-effect sensors were used to monitor the current flowing through each of the three parallel branches, and automotive-grade battery BMSs were used to measure the cell voltages, with one BMS for each branch. The BMS units communicated with the main computer via a CAN bus to mimic in-vehicle communication behavior with regard to signal availability, latency, and bandwidth. Including CAN-based communication helps reflect the realistic data constraints present in onboard vehicle systems. During the initial test cycles, the system experienced CAN bus buffering issues, which resulted in noticeable delays between the BMS data and the IT5101 battery tester readings. This issue was identified and resolved during a maintenance break and did not occur again in later cycles.

Current profiles based on the WLTP were generated via an electromechanical simulation of a Tesla Model 3 using the Powertrain Blockset in MATLAB/Simulink. Then, the profiles were adapted to the electrical characteristics of the tested batteries. Synchronization between the current profile and system execution was a key validation priority. A real-time correction algorithm continuously checked the time steps of the driving cycle and automatically calibrated the time-step commands of the power source to minimize time lag. This ensured a delay of less than two seconds between the theoretical driving profile and the applied current, thereby maintaining fidelity during dynamic cycling.

To further ensure data quality and battery safety, multiple control mechanisms were implemented. The system actively monitors overcurrent, overvoltage, undervoltage, and overtemperature conditions. If any of these thresholds were exceeded, or if the voltage imbalance between branches exceeded 0.1 V, the test would automatically pause and notify the user. The test would only resume once acceptable conditions were reestablished. This robust control architecture minimizes the impact of transient faults and environmental disturbances while ensuring consistent test conditions across cycles. A parallelized Python system continuously logged data, handling real-time acquisition, error monitoring, and efficient storage in PARQUET format. This allowed for reliable long-duration testing with high temporal resolution.

Figure 11 shows example data extracted from multiple cells. The top plot displays the surface temperatures of the cells during cycle 150. The second plot (middle) shows the current flow through the three branches of the battery pack during cycle 160. Finally, the third plot (bottom) illustrates the individual voltages of each cell during cycle 170.

Fig. 11.

Fig. 11

Open in a new tab

Example data considering different cycles and parameters such as voltage, current and temperature.

Usage Notes

This dataset supports a wide range of battery-related research and development tasks. It supports training and validating data-driven models for state of charge (SOC) estimation and thermal behavior forecasting. The inclusion of WLTP-based current profiles makes the dataset especially relevant for studying dynamic degradation patterns under realistic usage conditions. Researchers can use the data to evaluate balancing strategies, detect cell or branch imbalances, and test diagnostic algorithms under real-world communication constraints. Additionally, the dataset can aid in designing and validating BMS algorithms and energy management systems.

Note that there are data gaps ranging from 1 second to 1 hour in cycles 1 to 124.

Note that some temporal gaps between cycles may be present in the dataset. These gaps are due to routine system maintenance, software debugging and updates, or improvements made to the test bench conditions.

Acknowledgements

This project received funding from grant TED2021-130007B-I00, by MICIU/AEI/10.13039/501100011033/ and by ERDF “A way of making Europe,” by the European Union and from the Agència de Gestió d’Ajuts Universitaris i de Recerca-AGAUR (2021 SGR 00392).

Author contributions

The authors contributed to this research project as a team. Joaquín de la Vega designed and conducted the experiments, conceptualized the study, prepared the database and the overall publication design, and wrote the manuscript. J.A. Ortega conceptualized the study and the overall publication design, acquired funding, and uploaded the database. J.-R. Riba conceptualized the study and the overall publication design. He also acquired funding, supervised the experiments, and revised the manuscript. All authors read and approved the manuscript.

Data availability

The dataset is available at CORA.RDR repository26 10.34810/data2395.

Code availability

This study did not use any custom code for data curation, analysis, or validation. The dataset uses the Apache Parquet file format (.parquet), which is a free, open-source data storage format. A comprehensive and detailed description of the database structure and experimental conditions is provided in an accompanying README file in TXT format. All procedures related to data organization and storage were performed using standard commercial software tools. Therefore, no code repository is associated with this publication.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Mahmoudzadeh Andwari, A., Pesiridis, A., Rajoo, S., Martinez-Botas, R. & Esfahanian, V. A review of Battery Electric Vehicle technology and readiness levels. Renew. Sustain. Energy Rev.78, 414–430 (2017). [Google Scholar]
  • 2.Hausbrand, R. et al. Fundamental degradation mechanisms of layered oxide Li-ion battery cathode materials: Methodology, insights and novel approaches. Mater. Sci. Eng. B192(no. C), 3–25 (2015). [Google Scholar]
  • 3.Middlemiss, L. A., Rennie, A. J. R., Sayers, R. & West, A. R. Characterisation of batteries by electrochemical impedance spectroscopy. Energy Reports6, 232–241 (2020). [Google Scholar]
  • 4.Vetter, J. et al. Ageing mechanisms in lithium-ion batteries. J. Power Sources147(no. 1–2), 269–281 (2005). [Google Scholar]
  • 5.Keil, P. & Jossen, A. Charging protocols for lithium-ion batteries and their impact on cycle life—An experimental study with different 18650 high-power cells. J. Energy Storage6, 125–141 (2016). [Google Scholar]
  • 6.Birkl, C. R., Roberts, M. R., McTurk, E., Bruce, P. G. & Howey, D. A. Degradation diagnostics for lithium ion cells. J. Power Sources341, 373–38 (2017). [Google Scholar]
  • 7.Spotnitz, R. Simulation of capacity fade in lithium-ion batteries. J. Power Sources113(no. 1), 72–80 (2003). [Google Scholar]
  • 8.Zhang, J., Zhang, L., Sun, F. & Wang, Z. An Overview on Thermal Safety Issues of Lithium-ion Batteries for Electric Vehicle Application. IEEE Access6, 23848–23863 (2018). [Google Scholar]
  • 9.Offer, G. J., Yufit, V., Howey, D. A., Wu, B. & Brandon, N. P. Module design and fault diagnosis in electric vehicle batteries. J. Power Sources206, 383–392 (2012). [Google Scholar]
  • 10.dos Reis, G., Strange, C., Yadav, M. & Li, S. Lithium-ion battery data and where to find it. Energy AI5, 100081 (2021). [Google Scholar]
  • 11.BatteryArchive.org. [Online]. Available: https://batteryarchive.org/. [Accessed: 27-Nov-2023].
  • 12.Ranjith Kumar, R. et al. Advances in Batteries, Battery Modeling, Battery Management System, Battery Thermal Management, SOC, SOH, and Charge/Discharge Characteristics in EV Applications. IEEE Access11, 105761–105809 (2023). [Google Scholar]
  • 13.Ma, S. et al. Temperature effect and thermal impact in lithium-ion batteries: A review. Prog. Nat. Sci. Mater. Int.28(no. 6), 653–666 (2018). [Google Scholar]
  • 14.Troxler, Y. et al. The effect of thermal gradients on the performance of lithium-ion batteries. J. Power Sources247, 1018–1025 (2014). [Google Scholar]
  • 15.Kremer, P., Cigarini, F., Gohlich, D. & Park, S. Active Cell Balancing for Life Cycle Extension of Lithium-Ion Batteries under Thermal Gradient, Proc. Int. Symp. Low Power Electron. Des., vol. 2021-July (2021).
  • 16.An, Z., Zhang, J., Zhou, Y. & Gao, Z. Bio-inspired hybrid battery thermal management system with gradient flow channels. Appl. Therm. Eng.278, 127211 (2025). [Google Scholar]
  • 17.J. H. Bae, Z. Bai, B. J. Koo, H. R. Kim & D. G. Jang, The CAN communication application on the BMS, Int. Conf. ICT Converg., pp. 1140–1145, (2013).
  • 18.Cheng, K. W. E. W. E., Divakar, B. P. P., Wu, H., Ding, K. & Ho, H. F. F. Battery-management system (BMS) and SOC development for electrical vehicles. IEEE Trans. Veh. Technol.60(no. 1), 76–88 (2011). [Google Scholar]
  • 19.Panasonic, Lithium Ion NCR18650B. [Online]. Available: https://www.batteryspace.com/prod-specs/ncr18650b.pdf?srsltid=AfmBOorNRHxayN0R52KX_Yrn3XspHCpgDFeRgFRugetSTpeIHjSdN7iv. [Accessed: 23-Jun-(2025].
  • 20.U. Nations, UN Global Technical Regulation on Worldwide harmonized Light vehicles Test Procedures (WLTP), pp. 1–499, (1998).
  • 21.Riba, J.-R., Garcia, A. & Romeral, L. A computer experiment to simulate the dynamic behaviour of electric vehicles driven by switched reluctance motors. Int. J. Electr. Eng. Educ.51(no. 4), 368–382 (2014). [Google Scholar]
  • 22.Mathworks, Powertrain Blockset. Mathworks, (2024).
  • 23.Tesla, Tesla Model 3 Ownersmanual, [Online]. Available: https://www.tesla.com/ownersmanual/model3/en_us/Owners_Manual.pdf. [Accessed: 28-Mar-2025] (2024).
  • 24.Arat, M. & Bolarinwa, E. O. Rolling Resistance Effect of Tire Road Contact in Electric Vehicle Systems, SAE Tech. Pap., vol. 2015-April, no. April (2015).
  • 25.Di Natale, M., Zeng, H., Giusto, P. & Ghosal, A. Understanding and using the controller area network communication protocol: Theory and practice. Springer New York (2012).
  • 26.de la Vega Hernández, J., Ortega Redondo, J. A. & Riba Ruiz, J.-R. Lithium-ion battery pack cycling dataset with CC-CV charging and WLTP/constant discharge profiles. CORA.Repositori de Dades de Recerca10.34810/data2395 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The dataset is available at CORA.RDR repository26 10.34810/data2395.

This study did not use any custom code for data curation, analysis, or validation. The dataset uses the Apache Parquet file format (.parquet), which is a free, open-source data storage format. A comprehensive and detailed description of the database structure and experimental conditions is provided in an accompanying README file in TXT format. All procedures related to data organization and storage were performed using standard commercial software tools. Therefore, no code repository is associated with this publication.

Articles from Scientific Data are provided here courtesy of Nature Publishing Group

RESOURCES