US 12,253,408 B2
Battery malfunction detection by correlating vibrations response spectrums
Xiaoyu Li, Shenzhen (CN); Shaohong Lin, Shenzhen (CN); Jindong Tian, Shenzhen (CN); and Yong Tian, Shenzhen (CN)
Assigned to SHENZHEN UNIVERSITY, Shenzhen (CN)
Appl. No. 16/972,572
Filed by SHENZHEN UNIVERSITY, Shenzhen (CN)
PCT Filed Jun. 16, 2020, PCT No. PCT/CN2020/096303
§ 371(c)(1), (2) Date Dec. 2, 2021,
PCT Pub. No. WO2021/164156, PCT Pub. Date Aug. 26, 2021.
Claims priority of application No. 202010099104.8 (CN), filed on Feb. 18, 2020.
Prior Publication US 2023/0138536 A1, May 4, 2023
Int. Cl. G01H 9/00 (2006.01); G01M 7/02 (2006.01); G01N 29/12 (2006.01); G01N 29/34 (2006.01); G01N 29/40 (2006.01); G01N 29/44 (2006.01); G01N 29/46 (2006.01); G01R 31/392 (2019.01)
CPC G01H 9/00 (2013.01) [G01M 7/025 (2013.01); G01N 29/12 (2013.01); G01N 29/348 (2013.01); G01N 29/40 (2013.01); G01N 29/4427 (2013.01); G01N 29/46 (2013.01); G01N 2291/0231 (2013.01); G01N 2291/2698 (2013.01); G01R 31/392 (2019.01)] 15 Claims
OG exemplary drawing
 
1. A battery malfunction detection method, comprising:
applying a preset number of vibration signals with different frequencies to a battery to be detected through a vibration generation device;
collecting a response signal of the battery to be detected through a response collection device, wherein the response signal comprises the preset number of vibration response amplitudes of the battery to be detected, and the preset number of vibration response amplitudes of the battery to be detected are generated by the battery to be detected in response to the vibration signals with different frequencies respectively;
determining whether the battery to be detected is malfunctioning according to the response signal;
wherein the step of determining whether the battery to be detected is malfunctioning according to the response signal comprises:
matching the response signal with response samples of a plurality of battery samples in a preset response sample library one by one, wherein the response sample library comprises a response sample of at least one malfunctioning battery sample and a response sample of at least one battery sample that is not malfunctioning, and each response sample corresponds to a piece of identification information used to indicate whether the battery sample associated with the response sample is malfunctioning, and the response sample comprises the preset number of vibration response amplitudes of the associated battery sample, wherein the preset number of vibration response amplitudes of the battery sample are generated by the associated battery sample under an action of the vibration signals with different frequencies respectively;
determining a best matching response sample which has the largest match degree with the response signal from the response samples of the plurality of the battery samples;
determining whether the battery to be detected is malfunctioning according to the identification information corresponding to the best matching response sample;
wherein the step of matching the response signal with response samples of a plurality of battery samples in a preset response sample library one by one comprises:
generating a response spectrum according to a target response sample, wherein an abscissa of the response spectrum indicates the frequency, an ordinate of the response spectrum indicates the vibration response amplitude, and the target response sample is the response sample of any one battery sample in the response sample library;
acquiring a peak point in the response spectrum and at least two neighboring points nearest to the peak point, and determining the peak point and the neighboring points as detection points;
calculating a standard deviation of the vibration response amplitudes corresponding to the detection points and determining this standard deviation as a first standard deviation;
acquiring the vibration response amplitudes of the battery to be detected from the response signal under the action of the vibration signals with at least two target frequencies so as to acquire at least two target vibration response amplitudes, wherein the at least two target frequencies are frequencies corresponding to the detection points;
calculating a standard deviation of the at least two target vibration response amplitudes and determining this standard deviation as a second standard deviation;
calculating an absolute difference between the first standard deviation and the second standard deviation;
acquiring a match degree between the response signal and the target response sample according to the absolute difference.