Periodic Reporting for period 1 - NEMO (NExt-generation MOdels for advanced battery electronics)
Okres sprawozdawczy: 2023-05-01 do 2024-10-31
The NEMO project aims to advance the hardware and software of the present BMS demonstrating the developments in two types of battery concepts at TRL 4. NEMO has pledged six well-defined objectives to achieve the target development. The first objective is to demonstrate improved sensor signal acquisition and increased computational resources in the BMS. This is to improve the typical voltage and current sensors acquisition rates where EIS is added for a more reliable physical signal, data communication is set to wireless, and an advanced microprocessor is used for increased computational resources. In the 2nd objective, the automatic SoC model is updated and validated. This is supported by the improved EIS signal acquisition which updates the SoC estimator parameters during the lifetime undergoing cycle life tests. The SoH and RUL are aimed to improve in the 3rd objective by developing advanced models. NEMO's 4th objective targets to improve the current state-of-art of passive balancing in a string by SoH-balancing concept via cell switching demonstrating the improvement in lifetime. NEMO will also work on the safety of the battery systems which is stated in its 5th objective. NEMO plans to detect the cell failure via EIS signals developing the mechanical swelling model and core temperature estimation which will be capable of preventing thermal runaway. Apart from the technical developments, NEMO also aims to develop a cloud platform to store the generated data from the project and disseminate it among the research community following the FAIR principle.
Upon achieving the above objectives, NEMO will not only advance the hardware components but also the software tools will significantly improve the BMS performance and the battery lifetime. The EIS-based signal acquisition and accurate SoC estimation throughout the lifetime will contribute to better performance and could potentially be extended to the 2nd life ensuring reliable performance meaning financial and social acceptance of extended battery use. The lifetime extension by 20% significantly improves the battery usage time while the SoH and RUL prediction models can provide inputs for optimal usage of the battery systems. On safety-critical points, NEMO will address the battery failure prediction, and SoH balancing during the lifetime with the basis of onboard EIS measurement. Moreover, the utilization of the latest generation of EIS chips with an advanced microprocessor for improved computational power. Considering the initial investment in battery hardware, the return on investment will be high thanks to the NEMO's advancement of improved performance and increased lifetime.
Upon achieving the above objectives, NEMO will not only advance the hardware components but also the software tools will significantly improve the BMS performance and the battery lifetime. The EIS-based signal acquisition and accurate SoC estimation throughout the lifetime will contribute to better performance and could potentially be extended to the 2nd life ensuring reliable performance meaning financial and social acceptance of extended battery use. The lifetime extension by 20% significantly improves the battery usage time while the SoH and RUL prediction models can provide inputs for optimal usage of the battery systems. On safety-critical points, NEMO will address the battery failure prediction, and SoH balancing during the lifetime with the basis of onboard EIS measurement. Moreover, the utilization of the latest generation of EIS chips with an advanced microprocessor for improved computational power. Considering the initial investment in battery hardware, the return on investment will be high thanks to the NEMO's advancement of improved performance and increased lifetime.
Below are the updates per work package (WP) that have been performed during the first 18 months of the project. For this, only technical WP achievements are considered.
WP2: In this WP, the requirements and specifications are defined and delivered to all the technical WPs. This includes the development of models and algorithms in WP3/WP4, hardware and software advancements in WP5/WP6, and the testing and validation activities in WP7. The resulting documents provide a detailed description of the specifications for the various hardware and software solutions, their capabilities, and the required testing procedures.
WP3: The pseudo-2D (P2D) battery ageing model is achieved by incorporating and determining the model parameters. The specific cell parameters are determined for the NEMO cell. For the SoC estimator, the cells were cycled, EIS was carried out, using the distribution of relaxation time allowing calibration of the model parameters. For the SoT, an EIS-based model was put in place and validated against several temperature points. For the MSM, cells were characterized under various preload force conditions and cycling profiles. The thickness change of cells is simulated based on an electrode swelling algorithm. In another task, the first version of the SoC algorithm is running in the TTTECH cloud, the algorithm was coded in C.
WP4: A detailed literature study is performed to identify the advanced data-driven modeling techniques motivated by large language models. The first trial is made with MOIRAI. The partners have worked extensively to set a common framework for data format, automatic data upload structure, etc. Research is performed to define the SoS parameters and safety-relevant thresholds. Mechanical abuse tests are performed deforming the cell in the centre at quasi-static conditions.
WP5: The single-cell sensing boards are developed for the BMS. The layout of the single-cell sensing board was designed and adjusted for the NEMO cell. 24 PCBs were populated, tested, and supplied by conducting functional testing on each single cell board. Cell-based EIS boards were integrated between the cell tabs. The cell-active balancing boards are also designed and ordered. Two demonstrator HV battery systems were developed. The development of the zBMS+ prototypes is also finished, parts are ordered and the manufacture is planned for 2025.
WP6: The development of the BMS driver and data handling firmware has been completed developing the firmware and software for the master boards. The software stack has been successfully deployed. The cloud-related software development has progressed, setting up the infrastructure and implementing data processing capabilities. There is progress on model and algorithm integration; the SoC algorithm was deployed, Aurix TC4 evaluation boards were provided, and packaging possibilities for physics-based models are studied including model integration in the software architecture.
WP7: Testing activities on the cell level and on the system level serve as the basis for the model validation and demonstration of the defined KPIs. Cycling activities on the prototypes are ongoing and the hardware is commissioned. The testing activities on the cell level have started for the model development.
WP2: In this WP, the requirements and specifications are defined and delivered to all the technical WPs. This includes the development of models and algorithms in WP3/WP4, hardware and software advancements in WP5/WP6, and the testing and validation activities in WP7. The resulting documents provide a detailed description of the specifications for the various hardware and software solutions, their capabilities, and the required testing procedures.
WP3: The pseudo-2D (P2D) battery ageing model is achieved by incorporating and determining the model parameters. The specific cell parameters are determined for the NEMO cell. For the SoC estimator, the cells were cycled, EIS was carried out, using the distribution of relaxation time allowing calibration of the model parameters. For the SoT, an EIS-based model was put in place and validated against several temperature points. For the MSM, cells were characterized under various preload force conditions and cycling profiles. The thickness change of cells is simulated based on an electrode swelling algorithm. In another task, the first version of the SoC algorithm is running in the TTTECH cloud, the algorithm was coded in C.
WP4: A detailed literature study is performed to identify the advanced data-driven modeling techniques motivated by large language models. The first trial is made with MOIRAI. The partners have worked extensively to set a common framework for data format, automatic data upload structure, etc. Research is performed to define the SoS parameters and safety-relevant thresholds. Mechanical abuse tests are performed deforming the cell in the centre at quasi-static conditions.
WP5: The single-cell sensing boards are developed for the BMS. The layout of the single-cell sensing board was designed and adjusted for the NEMO cell. 24 PCBs were populated, tested, and supplied by conducting functional testing on each single cell board. Cell-based EIS boards were integrated between the cell tabs. The cell-active balancing boards are also designed and ordered. Two demonstrator HV battery systems were developed. The development of the zBMS+ prototypes is also finished, parts are ordered and the manufacture is planned for 2025.
WP6: The development of the BMS driver and data handling firmware has been completed developing the firmware and software for the master boards. The software stack has been successfully deployed. The cloud-related software development has progressed, setting up the infrastructure and implementing data processing capabilities. There is progress on model and algorithm integration; the SoC algorithm was deployed, Aurix TC4 evaluation boards were provided, and packaging possibilities for physics-based models are studied including model integration in the software architecture.
WP7: Testing activities on the cell level and on the system level serve as the basis for the model validation and demonstration of the defined KPIs. Cycling activities on the prototypes are ongoing and the hardware is commissioned. The testing activities on the cell level have started for the model development.
The following NEMO results go beyond the state of the art.
- NEMO will implement BMS concepts with the inclusion of a cell-level EIS chip, distributed computational approach, wireless data communication, and increased computational resources.
- The physics-based BESTIMATOR framework will have continuous updates throughout its lifetime.
- EIS-based physics- and data-driven ageing models will accurately predict the SoH and RUL on the edge and cloud computing.
- The active balancing concept based on ageing models extending the lifetime.
- EIS-driven MSM model will predict early failure isolating affected cells by switching.
- NEMO will share 100% of BMS data to the research community.
- NEMO will implement BMS concepts with the inclusion of a cell-level EIS chip, distributed computational approach, wireless data communication, and increased computational resources.
- The physics-based BESTIMATOR framework will have continuous updates throughout its lifetime.
- EIS-based physics- and data-driven ageing models will accurately predict the SoH and RUL on the edge and cloud computing.
- The active balancing concept based on ageing models extending the lifetime.
- EIS-driven MSM model will predict early failure isolating affected cells by switching.
- NEMO will share 100% of BMS data to the research community.