A semiconductor device capable of monitoring the state of a battery or the like is provided. The states of a plurality of batteries in a battery module is easily acquired. The semiconductor device that can be attached to an electrode of a battery or the like includes a first substrate, an element layer, and first to third conductive layers. The element layer includes a first circuit and a second circuit and is provided on a side of a first surface of the first substrate. The first conductive layer and the second conductive layer are provided on a side of a second surface positioned opposite to the first surface of the first substrate. The first circuit is electrically connected to each of the first conductive layer and the second conductive layer through an opening provided in the first substrate. The third conductive layer is provided to be stacked on a side opposite to the first substrate side of the element layer and electrically connected to the second circuit. The first conductive layer and the second conductive layer each function as a terminal, and the third conductive layer functions as an antenna.

This evaluation device comprises: a mathematical model acquisition unit that acquires a mathematical model expressing the state of a power storage element; an operation data acquisition unit that acquires operation data which includes time-series input data input during operation of a system constructed on the basis of the numerical model, and time-series output data output by the system on the basis of the time-series input data; a processing unit that inputs the time-series input data to the numerical model and executes processing causing time-series model output data to be output from the numerical model; and an evaluation unit that evaluates the design and the operation of the system on the basis of the time-series output data and the time-series model output data.

This evaluation device comprises: a mathematical model acquisition unit that acquires a mathematical model expressing the state of a power storage element; an operation data acquisition unit that acquires operation data which includes time-series input data input during operation of a system constructed on the basis of the numerical model, and time-series output data output by the system on the basis of the time-series input data; a processing unit that inputs the time-series input data to the numerical model and executes processing causing time-series model output data to be output from the numerical model; and an evaluation unit that evaluates the design and the operation of the system on the basis of the time-series output data and the time-series model output data.

Various embodiments include a method for producing a quality test system executing a quality test model with a filter mask and a quality model to determine a quality feature of a battery cell. The system has an electrochemical impedance spectroscopic unit for capturing test data relating to the battery within a frequency range. The method includes: creating the model; and producing the system. Creating the model includes: capturing spectroscopic learning data; creating the filter mask using a first machine learning method with analysis data from part of the frequency range by consulting the filter mask and creating the model using a second machine learning method. The first and the second learning method are coupled based on the learning data. The first machine learning method creates a filter mask determining the analysis data such that the second machine learning method creates a quality model optimized with respect to maximizing the quality.

Various embodiments include a method for producing a quality test system executing a quality test model with a filter mask and a quality model to determine a quality feature of a battery cell. The system has an electrochemical impedance spectroscopic unit for capturing test data relating to the battery within a frequency range. The method includes: creating the model; and producing the system. Creating the model includes: capturing spectroscopic learning data; creating the filter mask using a first machine learning method with analysis data from part of the frequency range by consulting the filter mask and creating the model using a second machine learning method. The first and the second learning method are coupled based on the learning data. The first machine learning method creates a filter mask determining the analysis data such that the second machine learning method creates a quality model optimized with respect to maximizing the quality.

A simulation system and a method for simulating a performance of at least one storage unit of an energy storage. The simulation system includes at least one respective model for a respective storage unit of the energy storage. The encoder-decoder model includes at least one recurrent neural network or at least one neural network having a transformer architecture. The encoder processes an encoder input sequence that describes a measured temporal course of current and voltage or of power and voltage of the storage unit assigned to the model. The encoder generates an initial state of the model. The decoder processes a decoder input sequence describing a temporal course to be simulated of the current or of the power of the storage unit. The decoder generates a decoder output sequence that describes a simulated temporal course of the voltage of the storage unit assigned to the model.

To generate a charging path for a battery, a method includes generating simulation data for charging currents based on a battery model indicating an internal state of a battery, generating an initial look-up table (LUT) for the charging currents and preset battery voltage limits based on the simulation data, the initial LUT representing initial charging limit conditions of the battery for stages corresponding to the charging currents, generating a modified LUT by adjusting at least one of the initial charging limit conditions of the initial LUT, in response to the initial LUT failing to satisfy a threshold, determining a final LUT based on the modified LUT, in response to the modified LUT satisfying the threshold, and generating a charging path for the battery based on the final LUT.

To generate a charging path for a battery, a method includes generating simulation data for charging currents based on a battery model indicating an internal state of a battery, generating an initial look-up table (LUT) for the charging currents and preset battery voltage limits based on the simulation data, the initial LUT representing initial charging limit conditions of the battery for stages corresponding to the charging currents, generating a modified LUT by adjusting at least one of the initial charging limit conditions of the initial LUT, in response to the initial LUT failing to satisfy a threshold, determining a final LUT based on the modified LUT, in response to the modified LUT satisfying the threshold, and generating a charging path for the battery based on the final LUT.

A prediction method of battery health includes: obtaining an environment temperature, and a discharge capacity and an operating parameter of a battery; determining an estimated operating parameter in a preset temperature at the discharge capacity according to the environment temperature, the discharge capacity and the operating parameter; determining an estimated capacity of the battery according to the estimated operating parameter; and determining a health level of the battery according to the estimated capacity of the battery and a reference capacity, at the discharge capacity of the battery.

The present disclosure discloses an uncertain noisy filtering-based fault diagnosis method for a power battery management system and belongs to the field of power battery fault diagnosis. The method comprises: establishing an electro-thermal coupling model of a power battery system; extending an output vector of the system according to a state constraint of a power battery, and expanding a state vector of the system according to a fault of the power battery system to obtain an augmented system of the power battery system; obtaining an estimation interval of a power battery sensor fault by using a zonotope Kalman filtering method; judging whether the power battery management system has a fault according to upper and lower bounds of fault estimation, and if a fault occurs, determining a fault type and a fault time according to a result. Compared with an existing fault diagnosis method for a system without a state constraint, the present application solves the problem of fault diagnosis of a system with a state constraint by extending the state constraint of the system to the system output vector.