In this research thesis, a new data-centric approach is used to determine the individual cell capacities in Li-ion batteries used in electric vehicles (EVs). This approach could also be extended to batteries for static energy storage applications. The method is based on real-world data collected from pouched-shaped cells having Nickel Manganese Cobalt (NMC) based chemistry of different capacities, used in electric buses. An algorithm has been developed which takes as input the charging and discharging current data of the battery, and voltages of the individual cell either from sensors installed on the terminals of a cell or from a battery simulator software. As an output, the algorithm tries for finding the best two resting states on the time axis of the current profile and net charging or discharging of the battery in between those times. A reasonable amount of net charge is required for capacity calculation, for this reason, the recommended SOC difference between those two resting states should be at least ±10% or more. Several experiments were also performed to firm up the results. For the experiments, 100 pouch-shaped NMC-based cells of 40 Ah capacity each were used. These 100 cells were connected in series as 1P100s and it is one module of the xP100s larger battery pack of an electric trolley bus. The algorithm has three levels: at the first level, it uses only simulator data of current and cell voltages to determine the total capacities of individual cells from partial charging or discharging. In the second level, it takes real current data and simulator voltages from the individual cells to determine the same total capacities of individual cells. The third level uses real current charging and discharging data and only minimum, average, and maximum cell voltages to find an indicator of the skewed capacities of cells. Further, the second and third level is compared with the first. For more accuracy and exact calculations of the individual cell capacity, special tests and data collection procedures are proposed as well. Irrespective of the type of data available, a non-destructive diagnostic of the battery is carried out. Abnormal cells are detected with cell number and its location inside the pack in the case where individual cell data are available. In the case where only maximum and minimum cell voltages are available, the cell will be determined which limits the capacity of all the cells connected in series with it. For better diagnostics of the battery new data collection techniques are proposed, given that the owner of the vehicle allows the transfer of data from BMS.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:uu-491180 |
| Date | January 2022 |
| Creators | Bilal, Muhammad |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | MATVET Energiteknik |
Page generated in 0.0024 seconds