Systems and methods for predicting remaining useful life in batteries and assets
In one aspect, a computer-implemented method for a cloud-based computing system may include receiving data pertaining to a battery pack of a vehicle, wherein the data is measured by one or more sensors associated with the vehicle. The method may include determining, based on the data, whether a threshold is satisfied, wherein the threshold relates to a physics-based model estimated terminal voltage. Responsive to determining the threshold is satisfied, the method may include determining a trigger event has occurred and calibrate one or more parameters of the physics-based model, a machine learning model, or both.
1 . A computer-implemented method, wherein the computer-implemented method comprises:
receiving a plurality of data pertaining to a battery pack of a vehicle, wherein the plurality of data comprises first data including one or more measurements by one or more sensors associated with the vehicle, and second data pertaining to simulation of the battery pack when sensor data is missing;
determining, based on the plurality of data, whether a threshold is satisfied, wherein the threshold relates to a physics-based model estimated terminal voltage;
responsive to determining the threshold is satisfied, determining a trigger event has occurred and calibrate one or more parameters of a hybrid model comprising the physics-based model and a machine learning model that uses scaling to modify the plurality of data to predict at least a remaining useful life of each cell of the battery pack; and
using output comprising at least the remaining useful life of each cell of the battery pack from the hybrid model to cause an operating parameter of the vehicle to change.
2 . The computer-implemented method of claim 1 , wherein the physics-based model outputs one or more properties pertaining to the battery pack of the vehicle, and the machine learning model uses the one or more properties to predict the remaining useful life of each cell of the battery pack.
3 . The computer-implemented method of claim 2 , wherein the output is fed as input into a physics edge model with a temperature of the battery pack, a current of the battery pack, and an initial state of charge of the battery pack.
4 . The computer-implemented method of claim 2 , further comprising using, in the machine learning model, the first data from the one or more sensors to predict the remaining useful life of each cell of the battery pack.
5 . The computer-implemented method of claim 2 , further comprising extracting one or more features from the plurality of data, lab experiment data, user battery usage profile data, or some combination thereof to predict the remaining useful life of each cell of the battery.
6 . The computer-implemented method of claim 1 , wherein the threshold further relates to a sum of the physics-based model estimated terminal voltage minus an actual voltage of the battery pack.
7 . The computer-implemented method of claim 1 , further comprising training the machine learning model in the cloud-based computing system using at least information from a cell manufacturer of the battery pack and a user battery pack usage profile.
8 . The computer-implemented method of claim 1 , further comprising using a thermal single particle model with electrolyte algorithm for the physics-based model.
9 . The computer-implemented method of claim 1 , wherein the physics-based model is configured to output at least one of a time, a cell voltage, a current, a charge throughput, and an average solid electrolyte interphase thickness.
10 . The non-transitory computer-readable medium of claim 1 , wherein the physics-based model is configured to output at least one of a time, a cell voltage, a current, a charge throughput, and an average solid electrolyte interphase thickness.
11 . A tangible, non-transitory computer-readable medium storing instructions that, when executed, cause a processing device to:
receive a plurality of data pertaining to a battery pack of a vehicle, wherein the plurality of data comprises first data including one or more measurements by one or more sensors associated with the vehicle, and second data pertaining to simulation of the battery pack when sensor data is missing;
determine, based on the plurality of data, whether a threshold is satisfied, wherein the threshold relates to a physics-based model estimated terminal voltage;
responsive to determining the threshold is satisfied, determine a trigger event has occurred and calibrate one or more parameters of a hybrid model comprising the physics-based model and the machine learning model that uses scaling to modify the plurality of data to predict at least a remaining useful life of each cell of the battery pack; and
using output comprising at least the remaining useful life of each cell of the battery pack from the hybrid model to cause an operating parameter of the vehicle to change.
12 . The non-transitory computer-readable medium of claim 11 , wherein the physics-based model outputs one or more properties pertaining to the battery pack of the vehicle, and the machine learning model uses the one or more properties to predict the remaining useful life of each cell of the battery pack.
13 . The non-transitory computer-readable medium of claim 12 , wherein the output is fed as input into a physics edge model with a temperature of the battery pack, a current of the battery pack, and an initial state of charge of the battery pack.
14 . The non-transitory computer-readable medium of claim 12 , further comprising using, in the machine learning model, the first data from the one or more sensors to predict the remaining useful life of each cell of the battery pack.
15 . The non-transitory computer-readable medium of claim 11 , wherein the threshold further relates to a sum of the physics-based model estimated terminal voltage minus an actual voltage of the battery pack.
16 . The non-transitory computer-readable medium of claim 11 , further comprising training the machine learning model in the cloud-based computing system using at least information from a cell manufacturer of the battery pack and a user battery pack usage profile.
17 . The non-transitory computer-readable medium of claim 11 , further comprising using a thermal single particle model with electrolyte algorithm for the physics-based model.
18 . The non-transitory computer-readable medium of claim 11 , further comprising extracting one or more features from the plurality of data, lab experiment data, user battery usage profile data, or some combination thereof to predict the remaining useful life of each cell of the battery.
19 . A system comprising:
a memory device storing instructions; and
a processing device communicatively coupled to the memory device, wherein the processing device executes the instructions to:
receive a plurality of data pertaining to a battery pack of a vehicle, wherein the plurality of data comprises first data including one or more measurements by one or more sensors associated with the vehicle, and second data pertaining to simulation of the battery pack when sensor data is missing;
determine, based on the plurality of data, whether a threshold is satisfied, wherein the threshold relates to a physics-based model estimated terminal voltage; and
responsive to determining the threshold is satisfied, determine a trigger event has occurred and calibrate one or more parameters of the a hybrid model comprising the physics-based model and the machine learning model that uses scaling to modify the plurality of data to predict at least a remaining useful life of each cell of the battery pack; and
using output comprising at least the remaining useful life of each cell of the battery pack from the hybrid model to cause an operating parameter of the vehicle to change.
20 . The system of claim 19 , wherein the physics-based model outputs one or more properties pertaining to the battery pack of the vehicle, and the machine learning model uses the one or more properties to predict the remaining useful life of each cell of the battery pack.