We are happy to announce the new release of DATTES. During this summer, enjoy with this refreshing new release by downloading it from the official repository on https://gitlab.com/dattes/dattes/-/releases.
You can also keep up to date with the development version using git
: https://gitlab.com/dattes/dattes/.
Main new features
- Added EIS impedance identification
dattes_analyse(result,'e')
- More information in results.profiles:
- added amp-hour counters from cycler
- added step counters from cycler
- these counters will be used in future releases to improve DATTES functions (
calcul_soc
,which_mode
)
- Updating metadata after automatic generation:
- users can now export metadata with
dattes_export(result,'M')
- this will create a *.meta file beside *.xml file
- users can then edit this file to put accurate u_max, u_min, capacity values
- finally, by running
result = dattes_structure(result,'MSvs')
metadata is updated in the DATTES structure
- users can now export metadata with
- filter_phase_nr in
dattes_configuration
:- now users can use filter_phase_nr is capacity, pseudo_ocv, ocv_points, ica, resistance and impedance
- this allow to manually remove phases from analysis that had been auto-detected by
configurator
Main improvements
- Improvement in OCV by points identification: added
spike_filter
- Improvement in resistance identification:
- windowing region around delta time to filter noise around with linear polynomial
- added U_sim to visualise the quality of identification (
dattes_plot
) - added err_U to visualise and measure quality of identification
- Improvement in impedance identification:
- new method for RRC identification (
ident_rrc2
) - added U_sim to visualise the quality of identification (
dattes_plot
) - added rsquare and mverr to measure quality of identification
- new method for RRC identification (
Known issues:
-
Impedance identification under Octave:
Impedance identification (ident_cpe and ident_rrc) is not working properly under Octave. In most cases, no minimum is found with fminsearch and equivalent circuit parameters are all set to NaN.
-
Mode identification (which_mode):
Depending of data quality (time/voltage/current sampling), which_mode fails to properly identify CC/CV phases. Some long CC phases (e.g. pseudo OCV measurements) are cut and considered as CV and CC phases while they should be considered as one long CC phases.
-
Neware and Digatron files import has been tested on very few data. Univ. Eiffel do not have such type of cyclers, data from other institutions is welcome to perform supplementary tests.
-
Neware/Bitrode/Arbin files under octave: file import seem to be quite slow, specially when reading mixed (numeric/text) csv files.
-
Arbin res import not tested under Windows (needs to install mdbtools for Windows)
-
Arbin xls import not working under Windows/Octave