GIC//NMC Solar Battery Synthetic Data 2 - 45,000 x 18 degradation for clear-sky irradiance and cloudy days
Description
This dataset consists of part 2 of the data associated with publication "Data-driven Direct Diagnosis of PV Connected Batteries " The synthetic cycles were generated using the mechanistic modeling approach. See “Big data training data for artificial intelligence-based Li-ion diagnosis and prognosis“ (Journal of Power Sources, Volume 479, 15 December 2020, 228806) and "Analysis of Synthetic Voltage vs. Capacity Datasets for Big Data Li-ion Diagnosis and Prognosis" (Energies 2021, 14, 2371 ) for more details. Two sets of data are available, one for training and one for validation Training dataset: MEDB_PI folder, clear-sky irradiance, 0.025 triplet resolution up to 50% degradation with 2% increment. Validation dataset: MEDB_Cloud, 18 different cloudy days, 0.05 triplet resolution up to 50% degradation with 2% increment. All datasets were generated with slightly different cell parameters to account for cell-to-cell variations. Details are available in publication. For each duty cycle, 3 set of files are provided, the *_V files contains V vs. Q data, the *_t files contains V vs. time data and the *_R files contains rate vs. Q data. For each file, column in the volt, voltT, ot rate variable corresponds to 1 degradation path, the 1001 lines corresponds to the resolution in variable Q (for the capacity based data) or timenorm (for the time-based data). Details of each duty cycle is provided in the pathinfo variable with headers in pathinfo_index ( 1 - % LLI, 2 - % LAMPE, 3 - % LAMNE, 4 - Capacity, 5 - DOD). All simulations were performed with the 2022 version of the alawa toolbox. Voltage and kinetics of electrodes from different manufacturers, with different composition, or with different architecture might differ significantly. MEDB_irradiancedata.mat contains data gathered for 2 years at MEDB site (see publication for details). Data provided courtesy of HNEI’s Severine Busquet, Jonathan Kobayashi, and Richard Rocheleau This matlab structure contains the following variables: dn: time from Matlab reference time dh: hour of the day doy: day of the year dm: month of the year dy: year ghi: global irradiance (W/m2) class: clear sky yes/no perc_clear: clear sky percentage (%) meta: panel metadata tot_POA: clear sky irradiance at POA (W/m2) inc: Solar angle of incidence relative to the POA (degree) tot_horz: clear sky horizontal irradiance (W/m2)