Method and apparatus for identifying abnormal battery cell, electronic device, and storage medium
A method for identifying abnormal battery cell includes determining, based on target feature data of a battery cell, whether the battery cell is an abnormal battery cell. The target feature data includes feature data used for differentiation between an abnormal battery cell and a normal battery cell generated during a formation process.
1 . A method for identifying abnormal battery cell, comprising:
determining, based on target feature data of a battery cell, whether the battery cell is an abnormal battery cell; and
in response to the determining that the battery cell is abnormal, marking the battery cell as abnormal for removal from subsequent production,
wherein:
the target feature data comprises feature data used for differentiation between an abnormal battery cell and a normal battery cell generated during a formation process, the target feature data comprises a first target parameter value corresponding to a first arrival of a variation of a target parameter at a first process variation and a second target parameter value corresponding to a first arrival of the variation of the target parameter at a second process variation, the first process variation and the second process variation are different, the target parameter is selected from film-forming peak (dQ/dV), dynamic internal resistance (V/I), or dV/dQ, the first arrival of the variation of the target parameter at the first or second process variation corresponds to an earliest occurrence during a first stage of the formation process at which a difference between two consecutive sampling points of the target parameter equals the first or second process variation, respectively, the first or second target parameter value is obtained from the sampling points associated with the first arrival of the variation of the target parameter at the first or second process variation, respectively; and
determining, based on the target feature data of the battery cell, whether the battery cell is an abnormal battery cell comprises:
inputting the first and second target parameter values into a two-dimensional Gaussian distribution model constructed from sample battery cells to obtain an identification result of whether the battery cell is an abnormal battery cell, the obtaining the identification result comprising calculating a probability density of the battery cell; and
in response at least to the probability density being less than a preset probability density threshold, determining that the battery cell is an abnormal battery cell.
2 . The method for identifying abnormal battery cell according to claim 1 , wherein the target feature data comprises feature data influenced by electrolyte in the battery cell generated during the formation process.
3 . The method for identifying abnormal battery cell according to claim 1 , wherein the target feature data comprises feature data influenced by water content in the battery cell generated during the formation process.
4 . The method for identifying abnormal battery cell according to claim 1 , wherein determining, based on the target feature data of the battery cell, whether the battery cell is an abnormal battery cell comprises:
determining, during the formation process of the battery cell based on the target feature data of the battery cell, whether the battery cell is an abnormal battery cell.
5 . An electronic device, comprising a processor and a memory; wherein
the processor is configured to execute one or more instructions stored in the memory to implement the method for identifying abnormal battery cell according to claim 1 .
6 . A non-transitory computer-readable storage medium storing one or more instructions, the instructions being executable by a processor to implement the method for identifying abnormal battery cell according to claim 1 .
7 . An apparatus for identifying battery cell with electrolyte abnormalities, comprising:
a processor; and
a memory coupled to the processor, the memory storing instructions which, when executed the processor, cause the processor to:
determine, based on target feature data of a battery cell, whether the battery cell is an abnormal battery cell; and
in response to the determining that the battery cell is abnormal, mark the battery cell as abnormal for removal from subsequent production,
wherein:
the target feature data comprises feature data used for differentiation between an abnormal battery cell and a normal battery cell generated during a formation process, the target feature data comprises a first target parameter value corresponding to a first arrival of a variation of a target parameter at a first process variation and a second target parameter value corresponding to a first arrival of the variation of the target parameter at a second process variation, the first process variation and the second process variation are different, the target parameter is selected from film-forming peak (dQ/dV), dynamic internal resistance (V/I), or dV/dQ, the first arrival of the variation of the target parameter at the first or second process variation corresponds to an earliest occurrence during a first stage of the formation process at which a difference between two consecutive sampling points of the target parameter equals the first or second process variation, respectively, the first or second target parameter value is obtained from the sampling points associated with the first arrival of the variation of the target parameter at the first or second process variation, respectively; and
determining, based on the target feature data of the battery cell, whether the battery cell is an abnormal battery cell comprises:
inputting the first and second target parameter values into a two-dimensional Gaussian distribution model constructed from sample battery cells to obtain an identification result of whether the battery cell is an abnormal battery cell, the obtaining the identification result comprising calculating a probability density of the battery cell; and
in response to the probability density being less than a preset probability density threshold and the target feature data being within a preset feature range, determining that the battery cell is an abnormal battery cell.