Abstract: Systems and methods of optimizing a charging of a vehicle battery are disclosed. Using one or more electronic battery sensors, observable battery state data is determined regarding the charging of the battery. A neural network feature extractor extracts features from preceding vehicle battery state information. A reinforcement learning model, such as an actor-critic model, includes an actor model configured to produce an output associated with a charge command to charge the battery, and a critic model configured to output a predicted reward. The reinforcement learning model is trained based on the vehicle battery state information and the extracted features. This includes updating weights of the actor model to maximize the predicted reward output by the critic model, and updating weights of the feature extractor and weights of the critic model to minimize a difference between the predicted reward and health-based rewards received from charging the battery.

Abstract: Methods and systems for smoothening the transition of reward systems or datasets for actor-critic reinforcement learning models. A reinforcement model such as an actor-critic model is trained on a first dataset and a first reward system. The weights of the actor model and the critic model are frozen. While these weights are frozen, an affine transformation layer is attached to a final layer of the critic model, and the affine transformation layer is trained with a second dataset and a second reward system in order to adjust a weight of the final layer of the critic model. Then, the weights of the critic model are unfrozen which allows the adjusted weight of the final layer of the critic model to be implemented. The reinforcement learning model is retrained on the second dataset and second reward system, first with just the critic weights unfrozen, and then with both actor and critic weights unfrozen.

Abstract: Methods and systems of optimizing battery charging are disclosed. Battery state sensors are used to determine anode overpotential of a battery multiple times during multiple charge cycles. In a first phase, a reinforcement learning model (e.g., actor-critic model) is trained with rewards given throughout each charge cycle of the battery to optimize training. The reinforcement learning model can determine state-of-health characteristics of the battery over the charge cycles, and in a second phase, the reinforcement learning model is augmented accordingly. During this augmentation, the reinforcement learning model is trained with rewards given on a charge cycle-by-cycle basis, wherein rewards are given after looking at the charging optimization after the conclusion of each charge cycle.

Abstract: A method, device, and computer-readable medium for operating an electrical energy store is provided. The electrical energy store includes a storage cell for storing electrical energy and a control unit. State variables of the electrical energy store are detected. The state variables are transmitted to a computation unit outside of the electrical energy store. The electrical energy store is operated based on operating parameters provided by the computation unit.

Abstract: The invention relates to a computer-implemented method for predicting a modeled state of health of an electrical energy store having at least one electrochemical unit, in particular a battery cell, having the following steps: providing a data-based state of health model trained to assign a modeled state of health to the electrochemical energy store on the basis of characteristics of operating variables of the energy store; providing time characteristics of the operating variables that characterize operation of the electrical energy store; and determining a present or predicted modeled state of health on the basis of the generated characteristics of the operating variables using the state of health model, wherein data gaps in the time characteristics of the operating variables owing to a phase of inactivity are completed based on a characteristic of a temperature of the energy store that is derived from at least one provided ambient condition.

Abstract: The disclosure relates to a computer-implemented method for operating a system comprising a central processing unit, which is in communication with a multiplicity of battery-operated machines that each have an equipment battery, and for providing an action for extending the service life of the equipment battery of the machine in a machine-specific manner.

Abstract: A method for detecting a fault of a device battery or battery cell using a battery model, wherein the battery model is based on a differential equation system and designed to indicate a progression of an operational variable dependent on an internal battery state determined by a model parameter of the battery model. The method includes providing a temporal operational variable progression of multiple operational variables for a period of time. The method includes adjusting the model parameter of the electrochemical battery model based on the operational variable progression over the period of time. The method includes detecting a fault type dependent on a predetermined rule, wherein the rule indicates a fault condition dependent on a deviation of the model parameter from a corresponding predetermined model parameter, and, when the fault condition of the rule is satisfied, signaling the fault type.

Abstract: A method for adjusting a parameter of a model of a battery, including providing a temporal operating variable profile of a plurality of operating variables for a time period, and modelling a profile of one of the operating variables using the model based on an operating variable of the provided operating variable profile within the time period. The method includes determining an operating state of the battery when the modeled operating variable deviates from the corresponding operating variable by more than a threshold value during the time period. The method includes performing a charging process, wherein, in the presence of the operating state, a current pulse or a voltage pulse of a specified height and a specified duration is superimposed on a charging current or a charging voltage while a further temporal operating variable profile is captured and adjusting the parameter based on the further temporal operating variable profile.

Abstract: A method for determining charging profiles for device batteries of battery-operated devices. In one instance, the method includes selecting device batteries having the same usage-related load and the same aging state; dividing the selected device batteries into groups; assigning different charging profiles to the groups of device batteries, wherein the charging profiles indicate for a charging operation a maximum permissible charging current depending on a charge level range; operating the device batteries of all groups with the respectively assigned charging profiles for a predetermined period of time, so that charging operations are executed depending on the respectively assigned charging profile; detecting a change in the average aging state for each group of device batteries between the beginning of the predetermined time period and the end of the predetermined time period; and adjusting the charging profile depending on the change in the average aging state.

Abstract: The invention relates to a method for adjusting an anode overvoltage of a lithium-ion battery (310). The invention furthermore relates to a method for improving a capacity state of health of a lithium-ion battery (310). The invention also relates to a vehicle having at least one lithium-ion battery (310) whose anode overvoltage is adjusted using the method for adjusting the anode overvoltage of the lithium-ion battery (310) and/or whose capacity state of health is improved using the method for improving the capacity state of health of the lithium-ion battery (310). The invention also relates to a fleet management system that is designed to perform the method for adjusting the anode overvoltage of the lithium-ion battery (310) and/or the method for improving the capacity state of health of the lithium-ion battery (310).

Abstract: The disclosure relates to a method for monitoring battery cells and continuously providing cell aging states of battery cells of a device battery. The method includes providing temporal cell-operating value profiles of operating values for each of the battery cells and determining cell aging states of each battery cell based on one or more aging state models, respectively, depending on cell-operating value profiles provided with a high temporal resolution, after a respective evaluation period. The method includes selecting a portion of the battery cells for continuously capturing the cell-operating value profiles with the high temporal resolution in the respective next evaluation period, and continuously providing the cell- operating value curves of the selected battery cells with the high temporal resolution in the next evaluation period, while for remaining battery cells of the device battery no operating values or cell-operating value curves with low temporal resolution are provided.

Abstract: A method for determining an internal-resistance based state-of-health of a battery in a device includes determining the internal-resistance based state-of-health of the battery using a hybrid state-of-health model including a physical ageing model and a correction model by (i) determining, based on an electrochemical battery model, a physical state-of-health according to one or more operating parameters of the battery using the physical ageing model, (ii) mapping operating features derived from the one or more operating parameters onto a correction parameter using the correction model that includes a data-based configuration, and (iii) applying the correction parameter to the physical state of health to determine the internal-resistance based state-of-health. The method further includes determining a modeled battery voltage, and determining a characteristic of the modeled battery voltage using an adaptation model.

Abstract: A method for parameterizing an electrochemical battery model of a specific battery of a device, includes providing a current operating feature point of the specific battery and operating feature points of further batteries of a plurality of further devices for various evaluation periods. The operating feature points of the further batteries characterize an operation of a respective further battery within one or more of the evaluation periods based on several operating features. The several operating features are produced depending on characteristics of operating variables of the respective further battery. The method further includes selecting operating feature points of the further batteries that are similar to the current operating feature point of the specific battery, and providing model parameters of the electrochemical battery model that are associated with the current operating feature point and the selected operating feature points.

Abstract: A method for controlling a charging current limiting value for a battery management system. In one example, the method includes determining, for a measured temperature and a prescribed state of charge, reference currents for various time intervals; calculating a corresponding reference time constant for each reference current by using a model for the calculation of a mean value of a charging current based on a continuous current; constituting a diagram for the relationship between the reference time constant and the reference current; determining a predictive time constant by the comparison of a measured value of a charging current with the reference currents; calculating a predictive limiting mean value of the charging current; and calculating a first predictive limiting value ipredS for a short predictive time tpredS, a second predictive limiting value ipredL for a long predictive time tpredL, and a third predictive limiting value ipredP for a continuous predictive time.

Abstract: A method for carrying out a process for charging an appliance battery of an appliance, with the following steps: providing a charging profile that specifies a maximally permissible charging current for charging the appliance battery in a manner depending on a state of charge of the appliance battery; ascertaining a current state of health of the appliance battery; providing a reference state of health that predetermines a customary state of health for a calendrical age of the appliance battery; charging the appliance battery in a manner depending on a corrected maximally permissible charging current, the corrected maximally permissible charging current being ascertained by applying a correction value to the maximally permissible charging current, the correction value being determined in a manner depending on a correction function, depending on a difference between the reference state of health and the current state of health.

Abstract: A method for controlling a cell current limiting value for a battery management system. In some examples, the method includes determining quadratic reference currents of a battery cell; calculating a corresponding reference time constant for each reference current using a model for the calculation of a RMS value of a cell current by reference to a continuous current; constituting a diagram for the relationship between the reference time constant and the quadratic reference current; determining a predictive time constant by the comparison of a quadratic measured value of a cell current with the quadratic reference currents; calculating a predictive RMS limiting value of the cell current; calculating a first predictive limiting value for a short predictive time, a second predictive limiting value for a long predictive time, and a third predictive limiting value for a continuous predictive time; and calculating additional RMS limiting value for the cell current.

Abstract: The invention relates to a method for estimating the state of an energy store comprising at least one electrochemical battery cell (12, 14, 16, 18, 20, 22, 24, 26, 28) using a battery management system (BMS) which comprises an impedance spectroscopy chip, having at least the following steps: a) determining the frequency-dependent impedance of the at least one electrochemical battery cell (12, 14, 16, 18, 20, 22, 24, 26, 28) using a data set recording taken in real-time, b) training an artificial neural network (60) with temperature-based training spectra as the input and a specification for temperature values belonging to each training spectrum as the output, c) taking into consideration a battery cell-to-battery cell variance (30) between the electrochemical battery cells (12, 14, 16, 18, 20, 22, 24, 26, 28) when testing the artificial neural network (60) using weighting functions ascertained during step b) and test spectra and estimating the temperature values belonging to the test spectra according to the

Abstract: Method for balancing states of charge of an electrical energy store with a plurality of battery cells.

Abstract: A computer-implemented method provides an electrochemical battery model and a state of health model for a device battery. The electrochemical battery model is based on a system of differential equations, models an equilibrium state, and reports a dependence between operating parameters of the device battery and a state of charge of the device battery. The state of health model includes at least one physical aging model based on a further system of differential equations, and models the state of health depending on progressions of the operating parameters of the device battery.

Abstract: The disclosure relates to a computer-implemented method for operating a system comprising a central processing unit, which is in communication with a multiplicity of battery-operated machines that each have an equipment battery, and for providing an action for extending the service life of the equipment battery of the machine in a machine-specific manner.