US 12,362,404 B2
Method and apparatus for operating an early warning system for robustly detecting an error of an electrical energy store for a device by means of machine learning methods
Christoph Woll, Gerlingen (DE); Christian Simonis, Leonberg (DE); and Csaba Domokos, Simmozheim (DE)
Assigned to Robert Bosch GmbH, Stuttgart (DE)
Filed by Robert Bosch GmbH, Stuttgart (DE)
Filed on Oct. 3, 2022, as Appl. No. 17/937,489.
Claims priority of application No. 10 2021 211 158.0 (DE), filed on Oct. 4, 2021.
Prior Publication US 2023/0104003 A1, Apr. 6, 2023
Int. Cl. H01M 10/48 (2006.01); G01R 31/367 (2019.01); G01R 31/392 (2019.01); G06N 3/088 (2023.01); G07C 5/08 (2006.01); H01M 10/42 (2006.01); G07C 5/00 (2006.01)
CPC H01M 10/48 (2013.01) [G01R 31/367 (2019.01); G01R 31/392 (2019.01); G06N 3/088 (2013.01); G07C 5/0808 (2013.01); H01M 10/4285 (2013.01); G07C 5/008 (2013.01); H01M 2220/20 (2013.01)] 15 Claims
OG exemplary drawing
 
1. A computer-implemented method for determining an error of a behavior of an electrical energy store in a technical device, comprising:
sensing an operating variable profile of at least one operating variable of the electrical energy store;
determining at least one operating feature from the operating variable profile of the at least one operating variable of the electrical energy store;
evaluating a variational autoencoder with a supplied input variable vector, which comprises an operating feature point from the at least one operating feature, in order to determine a degree of deviation from a distribution in a latent state space or a reconstruction error;
detecting the error of the behavior depending on (i) the degree of deviation or the reconstruction error, and (ii) a predetermined error threshold; and
signaling the error of the behavior.