A data processing method for battery residual capacity based on electrical sensing

Abstract

The application discloses a data processing method for battery residual capacity based on electrical sensing, and relates to the technical field of electrical data processing.The method comprises the following steps: a micro planar electromagnetic coil array is embedded in a battery module, a multi-frequency excitation detection sensor is used to obtain an electromagnetic wave signal; a battery management platform is developed to process the capacity, a data interaction is performed between the capacity processor and the planar electromagnetic coil array and an external electrical sensing end, the capacity processor is triggered to perform residual capacity analysis under forward propagation and backward propagation mutual verification, battery residual capacity data is determined and is displayed in a pop-up window, so that the technical problem that the core internal state affecting the capacity cannot be accurately captured in the prior art, the capacity estimation accuracy is insufficient, especially under complex working conditions, the error is large, and the current high-precision and online evaluation requirements cannot be met can be solved, and the automation of detection and analysis can be realized at the same time, and the accuracy of capacity analysis under complex working conditions can be improved.

Description

Technical Field

[0001] This invention relates to the field of electrical data processing technology, and more specifically to a method for processing battery remaining power data based on electrical sensing. Background Technology

[0002] In existing battery management systems, the estimation of remaining battery capacity is typically based on coulometric measurement, open-circuit voltage method, or equivalent circuit model method. Coulometric measurement relies on current integration and is susceptible to sensor zero drift and accumulated errors, resulting in a significant decrease in capacity estimation accuracy over long-term operation. The open-circuit voltage method requires static conditions and is difficult to apply to dynamic operating conditions such as vehicles and energy storage. Although equivalent circuit models and some data-driven methods introduce model correction or algorithm fitting, their parameters are highly dependent on operating condition calibration and cannot accurately reflect the complex electrochemical state changes inside the battery.

[0003] In existing technologies, key information inside the battery cannot be directly perceived, and the battery capacity estimation still relies on external macroscopic signals such as voltage and current. This leads to significant discrepancies between the remaining capacity and the health status assessment results in cases of aging, electrolyte phase change, or local failure.

[0004] Therefore, existing technologies struggle to accurately capture the core internal states that affect power consumption, resulting in insufficient accuracy in power consumption estimation, especially under complex operating conditions where the error is significant, making it difficult to meet current high-precision, online assessment requirements. Summary of the Invention

[0005] This application provides a data processing method for remaining battery power based on electrical sensing, which addresses the technical problem in the prior art where it is difficult to accurately capture the core internal states that affect the power level, resulting in insufficient accuracy in power level estimation, especially with large errors under complex operating conditions, making it difficult to meet the current requirements for high precision and online evaluation.

[0006] In view of the above problems, this application provides a data processing method for battery remaining power based on electrical sensing.

[0007] This application provides a data processing method for battery remaining power based on electrical sensing. The method includes: obtaining electromagnetic wave signals by performing multi-frequency excitation detection sensing through a miniaturized planar electromagnetic coil array embedded in the battery module; developing a power processor on a battery management platform, wherein the power processor is jointly constructed by battery electrochemical-electromagnetic coupling mechanism, battery electrochemical simulation, and electrical data analysis; through data interaction between the power processor and the planar electromagnetic coil array and an external electrical sensing terminal, triggering the power processor to perform remaining power analysis under forward and backward propagation mutual verification based on the obtained electromagnetic wave signals and battery electrical signals, and determining the battery remaining power data; and displaying the battery remaining power in a pop-up window on the visualization interface of the battery management platform.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] This application provides a data processing method for battery remaining power based on electrical sensing. It involves embedding a miniaturized planar electromagnetic coil array within the battery module to perform multi-frequency excitation detection sensing to obtain electromagnetic wave signals. A power processor is developed on a battery management platform, which is jointly constructed based on the battery electrochemical-electromagnetic coupling mechanism, battery electrochemical simulation, and electrical data analysis. Through data interaction between the power processor, the planar electromagnetic coil array, and an external electrical sensing terminal, the power processor is triggered to perform forward and backward propagation cross-validation analysis of remaining power based on the obtained electromagnetic wave signals and battery electrical signals to determine the remaining battery power data. The remaining battery power is displayed in a pop-up window on the visualization interface of the battery management platform. This method addresses the technical problem in existing technologies where it is difficult to accurately capture the core internal states affecting power, leading to insufficient accuracy in power estimation, especially under complex operating conditions, which makes it difficult to meet current high-precision, online evaluation requirements. It can simultaneously automate detection and analysis, improving the accuracy of power analysis under complex operating conditions. Attached Figure Description

[0010] Figure 1 This application provides a schematic flowchart of a data processing method for remaining battery power based on electrical sensing.

[0011] Figure 2 This application provides a schematic diagram illustrating the construction process of the power processor in a data processing method for remaining battery power based on electrical sensing. Detailed Implementation

[0012] This application provides a data processing method for remaining battery power based on electrical sensing to solve the technical problem in the prior art that it is difficult to accurately capture the core internal states that affect the power, resulting in insufficient accuracy in power estimation, especially with large errors under complex operating conditions, which makes it difficult to meet the current requirements for high precision and online evaluation.

[0013] like Figure 1 As shown, this application provides a data processing method for battery remaining power based on electrical sensing, the method comprising:

[0014] S1: Electromagnetic wave signals are obtained by embedding a miniaturized planar electromagnetic coil array within the battery module to perform multi-frequency excitation detection sensing.

[0015] Furthermore, multi-frequency excitation detection sensing is performed to obtain electromagnetic wave signals. Step S1 of this application includes:

[0016] Based on the physical processes of the battery module, an array transmission frequency band is set, wherein the physical processes are divided into ion migration, interface polarization and solid-liquid phase transition, and the array transmission frequency band includes a first frequency band, a second frequency band and a third frequency band corresponding to the physical processes; based on the array transmission frequency band, the planar electromagnetic coil array is driven to perform multiple rounds of parallel detection on the battery module to obtain the electromagnetic wave signal.

[0017] In this application, electromagnetic wave signals are obtained by performing multi-frequency excitation detection sensing through a miniaturized planar electromagnetic coil array embedded in the battery module.

[0018] Optionally, the planar electromagnetic coil array can be arranged in a way that involves attaching, embedding, or stacking within the structural components, heat insulation layer, or cell gap area of ​​the battery module, so as to form a stable electromagnetic coupling path with the cells. The geometry, thickness, and trace width of the coils are adapted to the spatial constraints of the module to facilitate array arrangement at the module scale.

[0019] Among them, the planar electromagnetic coil array refers to the coil being constructed in the form of a planar spiral or a planar multi-turn conductor, and multiple coil units being arranged in an array according to a predetermined spacing and topology, so as to achieve parallel or zoned detection of different areas of the module in space.

[0020] The multi-frequency excitation detection sensor refers to applying excitation signals with multiple frequency components to a planar electromagnetic coil array, making it exhibit differentiated sensitivity to the internal dielectric properties, conductivity properties, and polarization response of the battery at different frequencies, thereby outputting an electromagnetic wave signal containing information such as amplitude and phase.

[0021] In one specific embodiment provided in this application, the array transmission frequency band is set according to the physical processes of the battery module. The physical processes used to define the frequency band selection correspond to the key dynamic mechanisms within the battery. By dividing the physical processes into ion migration, interface polarization, and solid-liquid phase transition, the main contributors to the electrochemical and electromagnetic responses of the battery module are decomposed.

[0022] Specifically, ion migration refers to the diffusion and migration behavior of lithium ions in the electrolyte and porous electrode. This process has a relatively slow time constant and is usually more likely to exhibit conductivity and concentration-related characteristics when subjected to lower frequency electromagnetic disturbances. Interfacial polarization refers to the polarization response caused by the double layer, reaction kinetics, and charge transfer at the electrode / electrolyte interface. Its time constant is between that of ion migration and the faster dielectric response, and it is usually more sensitive to phase lag and polarization loss in the mid-frequency range. Solid-liquid phase transformation is a phase change in the battery's internal materials or electrolyte system under local temperature, concentration, or aging conditions. This type of process often causes changes in dielectric constant, loss factor, and frequency dispersion characteristics, and therefore can exhibit significant amplitude attenuation and dispersion differences at higher frequency ranges.

[0023] Based on the above division, the array transmission frequency band set in this application includes a first frequency band, a second frequency band, and a third frequency band corresponding to the physical process. The first frequency band, the second frequency band, and the third frequency band are exemplary frequency band grouping concepts, used to carry sensitive excitations to ion migration, interface polarization, and solid-liquid phase transition, respectively.

[0024] For example, in one implementation, the first frequency band can be a relatively low frequency sweep range to enhance the response to changes in electrolyte conductivity, the second frequency band can be a mid-frequency range to highlight the difference between interfacial charge transfer and polarization resistance, and the third frequency band can be a higher frequency range to capture changes in dielectric parameters and frequency dispersion characteristics.

[0025] In practice, this application is not limited to specific frequency values. The actual frequency band boundary can be calibrated and set according to the cell system, module structure, electromagnetic coupling distance and noise environment.

[0026] Subsequently, after setting the array's transmission frequency band, the planar electromagnetic coil array is driven to perform multiple rounds of parallel detection on the battery module to acquire the electromagnetic wave signal. That is, the control terminal applies excitation to the planar electromagnetic array coil units sequentially or alternately according to the frequency sets of the first, second, and third frequency bands. The excitation form can be sinusoidal frequency sweep, discrete multi-tone, pulse modulation followed by frequency domain calculation, etc., to ensure that distinguishable response data can be formed in different frequency bands.

[0027] The multi-round parallel detection involves repeatedly performing the excitation-reception-demodulation process several times within one sampling period on the same or multiple frequency bands, so as to reduce random noise and improve feature stability through averaging, filtering or consistency checks.

[0028] For example, in an exemplary process, the planar electromagnetic array coil can be divided into several detection channels. In each round of detection, all channels are excited in parallel in the first frequency band and the amplitude and phase data of the receiver are collected synchronously. Then, the acquisition is repeated in the second frequency band, and finally, the acquisition is repeated in the third frequency band, forming a data cube structure indexed by frequency band-channel-round. Subsequently, the characteristics of amplitude attenuation, phase shift or frequency dispersion of each channel in each frequency band are extracted, and finally, the electromagnetic wave signal characterizing the electromagnetic response state of the battery module is obtained.

[0029] In summary, multidimensional electromagnetic information related to internal ion migration, interfacial polarization, and solid-liquid phase transition can be obtained without relying on battery static conditions, providing a more sufficient data foundation for subsequent energy processors to perform mechanism mapping and energy analysis.

[0030] S2: Develop a power processor in the battery management platform, wherein the power processor is jointly constructed by the battery electrochemical-electromagnetic coupling mechanism, battery electrochemical simulation and electrical data analysis.

[0031] In this application, the battery management platform refers to a software and hardware collaborative system with the ability to collect data, process calculations, and display results. It is used to centrally process multi-source data from battery modules and external sensors. By developing a power processor, a functional unit with multi-layer computing logic, model deduction, and data fusion capabilities is constructed in the battery management platform to complete the analysis and output of battery remaining power related data.

[0032] Specifically, the power processor is jointly constructed by the battery electrochemical-electromagnetic coupling mechanism, battery electrochemical simulation, and electrical data analysis. The battery electrochemical-electromagnetic coupling mechanism is used to describe the correspondence between changes in the internal electrochemical state of the battery and the externally perceptible electromagnetic response. Its core lies in establishing a mapping relationship between internal state variables such as lithium-ion concentration distribution, interface polarization state, electrolyte conductivity, and dielectric properties, and electromagnetic characteristics such as amplitude changes, phase shifts, or frequency dispersion generated under multi-frequency electromagnetic excitation, thereby providing a mechanistic basis for the analysis of the physical meaning of electromagnetic signals.

[0033] Battery electrochemical simulation is used to numerically model and dynamically extrapolate the internal electrochemical behavior of batteries. It establishes a battery electrochemical model that includes material property parameters, geometric structure parameters, and internal state variables to simulate and calculate the voltage, current, and internal state changes of batteries under different states of charge and operating conditions.

[0034] By processing multi-source data in a unified manner, and under the dual influence of mechanistic constraints and data-driven approaches, battery data analysis can calculate the remaining battery capacity more accurately and stably, thus providing a reliable data foundation for the battery management platform's subsequent status display and early warning decisions.

[0035] Furthermore, such as Figure 2 As shown, in developing a power processor on a battery management platform, step S2 of this application includes:

[0036] A battery electrochemical model is established, based on an algebraic equation determined by internal state variables and battery material property parameters. Virtual sensors are embedded at key locations in the battery electrochemical model. A first mapping layer is constructed using the battery electrochemical-electromagnetic coupling mechanism, a second simulation layer is constructed using the battery electrochemical model, and a third computational layer is constructed using electrical data analysis. Through fully connected integration of these layers, an electrical energy processor is formed. The third computational layer establishes electrical data interaction with an external electrical sensing terminal.

[0037] In this application, the electrochemical behavior of the battery module during charging and discharging is abstracted and described through mathematical modeling. The internal state variables characterize key state quantities that change within the battery with time and operating conditions, such as lithium ion concentration in the electrodes, electrolyte concentration distribution, interfacial overpotential, or reaction rate. The battery material property parameters are used to define the inherent characteristics of different battery systems, such as the diffusion coefficient of the electrode material, electrolyte conductivity, specific capacity of the active material, and geometric parameters.

[0038] By introducing the aforementioned internal state variables and material property parameters into algebraic equations, mathematical relationships describing the internal electrochemical processes of the battery are formed, thereby enabling the electrochemical response of the battery under different states of charge and operating conditions to be calculable and predictable. This set of algebraic equations serves as the basis for constructing the battery electrochemical model.

[0039] Subsequently, after establishing the basic battery electrochemical model, virtual sensors are implanted at key locations within the model. These key locations refer to model nodes or physical regions that significantly influence the overall battery performance and output characteristics, such as the surface of electrode particles, the solid-liquid interface, or the vicinity of the current collector.

[0040] Specifically, the virtual sensor in this application is not a real hardware device, but rather a device that uses specific observation points or computing nodes set in the model to output the values ​​of internal state variables at the corresponding locations in real time.

[0041] By implanting virtual sensors, it is possible to obtain state information inside the battery that is difficult to measure directly without adding actual hardware sensors. This information can then be used as an important data source for subsequent data analysis and verification, thereby enhancing the ability to characterize the behavior of real batteries.

[0042] Based on this, this application constructs a first mapping layer using the battery electrochemical-electromagnetic coupling mechanism to realize the correspondence between the internal electrochemical state of the battery and the external electromagnetic response characteristics. By introducing the electrochemical-electromagnetic coupling mechanism, it maps the changes in the internal state variables of the battery into observable electromagnetic disturbance characteristics, thereby providing theoretical support for the physical interpretation and reverse inference of multi-frequency electromagnetic signals.

[0043] Subsequently, a second simulation layer is constructed based on the battery electrochemical model to numerically simulate the evolution of the battery's internal state under different time scales and operating conditions. The simulation results, such as voltage, current, and virtual sensor readings, are output through continuous or discrete calculations to form a dynamic approximation of the actual battery behavior.

[0044] Furthermore, a third computing layer is constructed using electrical data analysis to process, fuse, and verify the actual electrical data collected from external electrical sensing terminals and the data output from the aforementioned mapping and simulation layers. It can perform unified analysis of multi-source data through feature extraction, error evaluation, or sequence operations.

[0045] In this application, the input of the first mapping layer is the electromagnetic wave signal sensed in real time by the planar electromagnetic coil array, and the output is electrochemical state information; the input of the second simulation layer is the electrochemical state information output by the upper layer, and the data initialization and simulation analysis are performed on the battery electrochemical model embedded in the layer, and the output is virtual sensing data of key locations; the input of the third analysis layer is the electrical data output by the upper layer and from the external electrical sensing terminal, and the output is the remaining battery power data.

[0046] By integrating the first mapping layer, the second simulation layer, and the third computation layer through full layer connectivity, that is, by maintaining correlation and interaction between the layers in terms of data flow and computational logic, the electromagnetic mapping results, electrochemical simulation results, and electrical data analysis results can mutually constrain and complement each other, thereby forming an overall collaborative processing structure.

[0047] Preferably, based on the above architecture, this application further adopts a sample-driven training method. By retrieving historical power consumption records, the recorded data is integrated and input / output is identified according to the above three-layer architecture, and used as training samples. The power consumption processor is obtained through supervised training until convergence.

[0048] The third computing layer establishes electrical data interaction with the external electrical sensing terminal to receive real-time collected external electrical signals such as voltage and current, and feeds back the analysis results to the power processor.

[0049] In summary, through the above-mentioned multi-layered joint construction method, the power processor can achieve comprehensive calculation and stable output of the remaining battery power through the combined effect of mechanism modeling, numerical simulation and data analysis.

[0050] Furthermore, virtual sensors are implanted at key locations in the battery electrochemical model. Step S2 of this application includes:

[0051] A first virtual sensor is set on the surface of the electrode particles in the battery electrochemical model. The sensing reading of the first virtual sensor is the lithium ion surface concentration, which directly affects the local overpotential and indirectly affects the total voltage. A second virtual sensor is set on the current collector interface of the battery electrochemical model. The sensing reading of the second virtual sensor is the local current density, and the total current is quantified by integrating the local current density.

[0052] In this application, a first virtual sensor is disposed on the surface of the electrode particles in the battery electrochemical model. The electrode particle surface refers to the interface region where the solid particles constituting the electrode active layer contact the electrolyte. This region is the primary location where lithium-ion insertion and deintercalation reactions occur, and it directly affects the electrochemical performance of the battery.

[0053] Specifically, by introducing an observation unit at the computational node or boundary condition corresponding to the electrode particle surface in the battery electrochemical model, serving as the first virtual sensor to output specific internal state variables, the sensor reading represents the lithium-ion surface concentration. This sensor characterizes the instantaneous concentration of lithium ions per unit volume or unit surface area on the electrode particle surface, reflecting the electrode reaction activity and local concentration gradient changes. In terms of model mechanism, the lithium-ion surface concentration directly affects the magnitude of the local overpotential. Because the concentration gradient alters the kinetics of the electrode reaction, it causes a shift in the interfacial reaction potential. The change in local overpotential further indirectly affects the battery's total output voltage through the voltage superposition relationship in the electrode region.

[0054] Therefore, the lithium-ion surface concentration reading obtained through the first virtual sensor can serve as an important intermediate variable for characterizing the internal reaction state and voltage change trend of the battery.

[0055] Meanwhile, a second virtual sensor is set at the current collector interface of the battery electrochemical model. The current collector interface refers to the connection area between the electrode active material layer and the metal current collector. This area is responsible for collecting the charge generated inside the electrode and conducting it to the external circuit. It is a key location for current formation and output.

[0056] Similarly, by embedding an observation unit for outputting current-related state quantities at the calculation position of the corresponding current collector interface in the battery electrochemical model, as a second virtual sensor, its sensing reading is the local current density, which is used to characterize the current intensity distribution through the current collector interface per unit area. This local current density can reflect the differences in electrode reaction rate and conductivity state in different regions.

[0057] By spatially integrating the local current density at each location on the current collector interface, the total current output of the battery can be quantified, thereby establishing a correspondence between local electrochemical behavior and overall current characteristics at the model level.

[0058] In summary, by setting a first virtual sensor and a second virtual sensor on the surface of the electrode particles and the interface of the current collector, respectively, this application can simultaneously acquire key internal information reflecting the voltage formation mechanism and current output characteristics without adding physical sensors, providing reliable model support for subsequent power processor to perform multi-source data fusion and remaining power analysis.

[0059] Furthermore, regarding the battery electrochemical-electromagnetic coupling mechanism, step S2 of this application includes:

[0060] Multidimensional electromagnetic disturbance features are extracted, wherein the multidimensional electromagnetic disturbance features include at least amplitude attenuation, phase shift, frequency dispersion and polarization rotation; the mapping between key internal states and multidimensional electromagnetic disturbance features is determined as the battery electrochemical-electromagnetic coupling mechanism, wherein the key internal states are determined at least by lithium ion concentration distribution, electrolyte conductivity and dielectric constant, and solid-liquid interface thickness.

[0061] In this application, the multidimensional electromagnetic disturbance features refer to the feature quantities that characterize the changes in the internal state of the battery extracted from different physical and frequency domain dimensions of the electromagnetic signal, which at least include elements such as amplitude attenuation, phase shift, frequency dispersion and polarization rotation.

[0062] Among them, amplitude attenuation is used to describe the degree of energy attenuation of electromagnetic excitation signal during propagation or coupling due to factors such as internal dielectric loss and conductivity changes in the battery. This feature is sensitive to changes in parameters such as electrolyte conductivity and polarization loss.

[0063] Phase offset is used to characterize the lag or advance of the received signal relative to the transmitted signal in phase, reflecting changes in the battery's internal equivalent impedance and polarization behavior.

[0064] Frequency dispersion is used to describe the nonlinear or dispersed characteristics of electromagnetic response as the excitation frequency changes. This characteristic can reflect the combined effects of changes in dielectric constant, interface structure, etc. with frequency.

[0065] Polarization rotation is used to characterize how the polarization state of electromagnetic waves rotates or changes due to material anisotropy, interface structure, or multiphase distribution during propagation or coupling within a battery. By jointly extracting these multiple characteristics, the comprehensive influence of the battery's internal electrochemical state on its electromagnetic response can be reflected from different perspectives.

[0066] In this application, a mapping between key internal states and multidimensional electromagnetic disturbance characteristics is established as the electrochemical-electromagnetic coupling mechanism of the battery. The key internal states refer to internal physical and chemical state quantities that have a decisive influence on changes in battery charge and health, and are determined at least by lithium-ion concentration distribution, electrolyte conductivity and dielectric constant, and solid-liquid interface thickness.

[0067] Among them, the lithium ion concentration distribution is used to describe the spatial distribution of lithium ions in the electrode and electrolyte. Its changes directly affect the electrode reaction rate and local polarization behavior, and are manifested in the electromagnetic level as changes in amplitude attenuation and phase shift.

[0068] Electrolyte conductivity and dielectric constant are used to characterize the ability of an electrolyte to conduct current and respond to electromagnetic fields. Changes in these properties can significantly affect the propagation loss and frequency dispersion characteristics of electromagnetic signals.

[0069] Solid-liquid interface thickness is used to describe the effective thickness variation of the interface layer between the electrode material and the electrolyte, such as the growth or degradation of the solid electrolyte interface film. This thickness variation will change the equivalent dielectric structure and polarization characteristics of the interface, which will be reflected in the electromagnetic characteristics as polarization rotation or frequency domain response differences.

[0070] By establishing the mapping relationship between the aforementioned key internal states and multidimensional electromagnetic disturbance characteristics, a coupling mechanism between the battery's electrochemical state and electromagnetic response can be formed. This allows externally acquired electromagnetic characteristics to be used to infer changes in the battery's internal state, providing physical mechanism support for the power processor to perform remaining power analysis and health status assessment.

[0071] S3: Through data interaction between the power processor, the planar electromagnetic coil array, and the external electrical sensing terminal, based on the obtained electromagnetic wave signal and battery electrical signal, the power processor is triggered to perform a remaining power analysis under forward and backward propagation mutual verification to determine the remaining battery power data.

[0072] In this application, the power processor receives electromagnetic wave signal data from a planar electromagnetic coil array and electrical signal data such as voltage and current from an external electrical sensing terminal in real time or periodically via a communication interface or internal bus, and transmits and updates the processing results or intermediate parameters between internal modules. Optionally, this data interaction process can be controlled by a preset sampling period, triggering conditions, or state change threshold to ensure that power analysis is initiated at an appropriate time.

[0073] Subsequently, the power processor is triggered to perform a remaining power analysis under forward and backward propagation cross-verification.

[0074] The forward propagation process involves using electromagnetic wave signals acquired from a planar electromagnetic coil array as input. Within the energy processor, the electromagnetic disturbance characteristics are mapped and deduced layer by layer along a computational path comprised of the battery's electrochemical-electromagnetic coupling mechanism mapping, simulation, and analysis. This yields a first data sequence reflecting the battery's internal electrochemical state and corresponding energy changes. This forward propagation process is primarily used to infer the battery's internal state and energy parameters from electromagnetic sensing information.

[0075] Backpropagation refers to using the battery electrical signal collected by an external electrical sensor as input, and performing a reverse deduction of the battery's state of charge along a calculation path that is opposite to or complementary to forward propagation, thereby obtaining a second data sequence. This process is used to perform a verification estimate of the state of charge from the perspective of traditional electrical signals.

[0076] Subsequently, through mutual verification, the data sequences obtained from forward propagation and backward propagation are compared, corrected, or fused to determine their consistency in terms of time series and numerical range.

[0077] When two conditions meet the preset consistency conditions or error thresholds, the power processor can fuse or weight the corresponding results to generate stable battery remaining power data; when there is a deviation that exceeds the threshold, data interpolation correction is performed.

[0078] In summary, through the aforementioned remaining power analysis mechanism based on forward and backward propagation verification, this application can improve the reliability and robustness of battery remaining power estimation under multi-source data constraints, thereby obtaining more accurate battery remaining power data.

[0079] Furthermore, triggering the power processor to perform remaining power analysis under forward and backward propagation cross-verification, step S3 of this application includes:

[0080] The battery electrical signal is acquired through an external sensor; the electromagnetic wave signal is input into the power processor to perform forward propagation and determine a first data sequence; the battery electrical signal is input into the power processor to perform backward propagation and determine a second data sequence; the first data sequence and the second data sequence are cross-validated to generate the remaining battery power data.

[0081] The data sequence includes electrochemical simulation data, virtual sensing data, and external battery electrical data; the sensor readings of the virtual sensors are used to synthesize the readings of the virtual external electrical data.

[0082] In this application, battery electrical signals are first acquired through an external sensing terminal. The external sensing terminal refers to a physical sensing unit arranged in the battery module or battery management system, which is used to acquire electrical parameters such as voltage, current, and optional temperature of the battery module in real time or periodically during actual operation.

[0083] The battery electrical signals reflect the macroscopic operating state of the battery and are a commonly available data source in existing battery management systems. Collecting these electrical signals provides benchmark and constraining external observation information for subsequent power analysis.

[0084] After the battery electrical signal acquisition is completed, the electromagnetic wave signal is input into the power processor to perform forward propagation and determine the first data sequence.

[0085] Specifically, the electromagnetic wave signal originates from the response data obtained by the planar electromagnetic coil array during multi-frequency excitation detection. This data is used as the input variable of the power processor and introduced into the calculation path consisting of battery electrochemical-electromagnetic coupling mechanism mapping, simulation, and analysis. The signal is passed and calculated layer by layer according to the layer connection order in the power processor to infer the internal state of the battery and its corresponding power changes based on external electromagnetic disturbance characteristics.

[0086] This forward propagation process can generate a first data sequence to characterize the evolution trend of the battery's internal state.

[0087] Subsequently, the battery electrical signal is input into the power processor to perform back propagation and determine the second data sequence.

[0088] Specifically, using electrical signals such as voltage and current collected by external electrical sensors as input, the battery electrochemical model and data analysis rules are combined within the energy processor to reverse-engineer the battery's state of charge and energy changes. This process, along a computational path complementary to the forward propagation, estimates the remaining battery energy from another dimension, thus forming a second data sequence constrained by external electrical signals.

[0089] In summary, the data sequences generated by forward and backward propagation can characterize the same battery state from different information sources and different mechanistic paths.

[0090] Subsequently, after obtaining the first data sequence and the second data sequence, the first data sequence and the second data sequence are cross-validated, that is, their consistency in the time dimension, numerical range and trend of change is checked and verified. When the two meet the preset consistency conditions or error thresholds, the corresponding data are fused or weighted to output stable and reliable battery remaining power data. If there is a deviation, the result can be adjusted by correction or data compensation to improve the accuracy of power estimation.

[0091] In the first and second data sequences, the data dimensions of the data sequences include electrochemical simulation data, virtual sensing data, and external battery electrical data.

[0092] The electrochemical simulation data refers to the internal state evolution results calculated by the battery electrochemical model; the virtual sensing data refers to the internal state readings output by the virtual sensors implanted in the battery electrochemical model; and the external electrical data of the battery consists of electrical signal data.

[0093] Furthermore, the sensor readings of the virtual sensor are used to synthesize virtual external electrical data readings. That is, by converting the internal state information output by the virtual sensor into an equivalent external electrical data form according to a preset mapping relationship, a synthesized external electrical signal based on electromagnetic sensing and analysis is obtained.

[0094] In summary, the power processor can achieve comprehensive analysis and accurate generation of the remaining battery power through the combined effect of two dimensions.

[0095] Furthermore, to cross-verify the first data sequence and the second data sequence, step S3 of this application includes:

[0096] The first data sequence and the second data sequence are mapped and checked to determine whether they meet the preset error amount. If the preset error amount is met, the mean of the mapped data points is calculated to determine the target data sequence. If there are data points that do not meet the preset error amount, the mean is calculated under the data point re-interpolation supplementation to determine the target data sequence.

[0097] The remaining battery power data includes the battery's intrinsic power and a multidimensional health score. The battery's intrinsic power is determined based on the ratio of the spatial integral of the active lithium ion content in each electrode region to the theoretical maximum lithium ion capacity of the electrode region. The multidimensional health score includes lithium loss, structural degradation, and electrolyte phase change.

[0098] In this application, after obtaining the first data sequence and the second data sequence, data point mapping and verification are first performed on the first and second data sequences. That is, time axis alignment, sampling point correspondence, or feature point matching are performed on the data sequences from different computation paths, so that each data point in the first data sequence can correspond one-to-one with a data point in the second data sequence that has the same physical meaning or time position. Subsequently, numerical comparisons are performed on the mapped corresponding data points to evaluate the consistency between the two in terms of amplitude, trend of change, or statistical characteristics.

[0099] The preset error is a tolerance threshold set according to system accuracy requirements, model confidence interval, or historical operating experience. It is used to determine whether two types of data can be regarded as an effective representation of the same battery state within a reasonable range.

[0100] When it is determined that the corresponding data points in the first data sequence and the second data sequence meet the preset error amount, it is considered that the analysis results obtained by forward propagation and backward propagation have a high degree of consistency in the current state. By performing the mean or weighted mean calculation of the corresponding data points, a more stable target data sequence can be obtained while reducing random errors and single path deviations. This target data sequence is used to characterize the comprehensive state of the battery in the corresponding time or interval.

[0101] When there are data points that do not meet the preset error amount, it indicates that there is a deviation between the two types of data sequences in some intervals or local states.

[0102] To address this issue, this application employs mean calculation through data point re-interpolation. Specifically, for data points that do not meet the error criteria, supplementary data points are generated using interpolation, smoothing, or prediction based on the data change trends of their adjacent time points or spatial locations. This corrects the impact of abnormal or missing data on the overall result. After the interpolation is completed, the mean is calculated for the corrected corresponding data points, thereby generating the target data sequence while ensuring data continuity and consistency.

[0103] In summary, the battery state assessment results can be kept stable even under complex operating conditions or in the presence of local measurement deviations.

[0104] Furthermore, based on the target data sequence, the generated remaining battery power data includes the battery's intrinsic power and a multidimensional health score.

[0105] Among them, the intrinsic charge of the battery is used to reflect the actual amount of electricity that the battery can release in the current state. It is determined based on the ratio of the spatial integral of the active lithium ion storage in each electrode region to the theoretical maximum lithium ion capacity of the electrode region. That is, by calculating the spatial integral of the number of active lithium ions that can participate in the reaction in different electrode regions inside the battery, and then performing ratio processing with the maximum lithium ion storage capacity of the corresponding electrode under ideal conditions, the intrinsic charge index reflecting the true state of charge of the battery is obtained.

[0106] The multidimensional health score characterizes the performance degradation of the battery during long-term use, encompassing multiple dimensions such as lithium loss, structural deterioration, and electrolyte phase change. Lithium loss reflects capacity decay caused by irreversible lithium consumption; structural deterioration reflects damage to electrode material structure or degradation of the conductive network; and electrolyte phase change reflects the impact of changes in electrolyte properties on battery performance. Specifically, the multidimensional health score is directly evaluated based on the target data sequence.

[0107] By incorporating the battery's intrinsic capacity and multidimensional health score into the remaining battery capacity data, this application enables a comprehensive characterization of the battery's available capacity and health status, providing more comprehensive data support for subsequent management, early warning, and decision-making.

[0108] S4: A pop-up window displays the remaining battery power in the visual interface of the battery management platform.

[0109] Furthermore, step S4 of this application includes:

[0110] Based on the remaining battery power data, a battery warning message is generated, which includes a power level warning message and a health status warning message. The remaining battery power data and the battery warning message are integrated and displayed in a pop-up window on the visualization interface of the battery management platform.

[0111] In this application, the battery management platform's visual interface—that is, the human-computer interaction interface provided to users or maintenance personnel—can graphically present the battery's operating status, power information, and related analysis results. After the power processor completes the calculation of the remaining battery power data, the corresponding information is instantly presented in the visual interface in the form of a pop-up window, prompt, or overlay. This allows users to obtain key power status without switching interfaces or actively querying, thereby improving the timeliness and perceptibility of information transmission.

[0112] The content displayed in the pop-up window may include the current remaining battery level, trend indicators, or a brief status description, but the specific presentation format is not limited.

[0113] Furthermore, this application also includes generating battery warning information based on the remaining battery power data, as another pop-up display content.

[0114] Specifically, in the battery management platform, the remaining battery power data output by the power processor is compared and analyzed with pre-set safety thresholds, operating strategies, or health assessment rules to automatically generate information to alert users to anomalies or potential risks. This battery warning information includes power level warnings and health level warnings. Power level warnings indicate when the remaining battery power is below a preset threshold, the rate of decline is abnormal, or the battery is about to fail to meet load demands, prompting users to charge, switch power sources, or adjust their power usage strategies. Health level warnings reflect abnormal battery states in a multi-dimensional health score, such as accelerated lithium loss, excessive structural degradation, or increased electrolyte phase transition risk, indicating potential performance degradation or safety hazards.

[0115] Subsequently, the remaining battery power data and the battery warning information are combined according to a predetermined display rule. For example, the remaining power index and its corresponding warning status are displayed simultaneously in the same pop-up window, or different levels of warning information are presented by means of icons, colors, text prompts, etc.

[0116] In summary, users can not only intuitively obtain the current remaining battery power, but also simultaneously learn whether there are any risks to the battery in terms of power level and health status, thus achieving comprehensive perception and timely response to battery status. This application can effectively improve the information prompting efficiency and user experience of the battery management platform, providing intuitive support for battery operation safety and management decisions.

[0117] This application provides a data processing method for remaining battery power based on electrical sensing, which has the following technical advantages:

[0118] By employing an embedded planar electromagnetic coil array to perform multi-frequency excitation detection and combining the electrochemical-electromagnetic coupling mechanism, a mapping between the internal microscopic state and electromagnetic disturbance characteristics is established. This enables precise capture of internal battery changes at the microscopic level, significantly improving the accuracy of remaining capacity estimation. A capacity processor integrating a coupling mechanism mapping layer, an electrochemical simulation layer, and an electrical data analysis layer is constructed. Virtual electrical data is synthesized using virtual sensors to supplement internal state information, further optimizing the accuracy of capacity analysis. Simultaneously, the detection and analysis are automated, enhancing the intelligence level of battery management. A cross-validation mechanism based on electromagnetic signal forward propagation and external electrical signal backward propagation is introduced to effectively reduce data interference and improve the reliability of detection results. Simultaneously, multi-dimensional health scores, including intrinsic battery capacity, lithium inventory loss, and structural degradation, are output. Combined with early warning information generation, this provides comprehensive data support for battery lifecycle management, helping to proactively mitigate safety risks.

[0119] Through the foregoing detailed description of a data processing method for battery remaining power based on electrical sensing, those skilled in the art can clearly understand the data processing method for battery remaining power based on electrical sensing in this embodiment. As for the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the description in the method section.

[0120] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for processing battery remaining power data based on electrical sensing, characterized in that, The method includes: Electromagnetic wave signals are obtained by embedding a miniaturized planar electromagnetic coil array within the battery module to perform multi-frequency excitation detection sensing. A power processor is developed in the battery management platform, wherein the power processor is jointly constructed by battery electrochemical-electromagnetic coupling mechanism, battery electrochemical simulation and electrical data analysis; Through data interaction between the power processor, the planar electromagnetic coil array, and the external electrical sensing terminal, the power processor is triggered to perform a remaining power analysis under forward and backward propagation mutual verification based on the obtained electromagnetic wave signal and battery electrical signal, thereby determining the remaining battery power data. The remaining battery power is displayed in a pop-up window on the visual interface of the battery management platform.

2. The data processing method for remaining battery power based on electrical sensing as described in claim 1, characterized in that, Perform multi-frequency excitation detection sensing to obtain electromagnetic wave signals, including: Based on the physical processes of the battery module, an array transmission frequency band is set, wherein the physical processes are divided into ion migration, interface polarization and solid-liquid phase transition, and the array transmission frequency band includes a first frequency band, a second frequency band and a third frequency band corresponding to the physical processes. Based on the array's transmission frequency band, the planar electromagnetic coil array is driven to perform multiple rounds of parallel detection on the battery module to acquire the electromagnetic wave signal.

3. The data processing method for remaining battery power based on electrical sensing as described in claim 1, characterized in that, Developing a power processor for a battery management platform includes: An electrochemical model of the battery is established, in which the algebraic equations determined by the internal state variables and battery material property parameters serve as the basis for its construction. Virtual sensors are implanted at key locations in the battery electrochemical model; A first mapping layer is constructed based on the battery electrochemical-electromagnetic coupling mechanism, a second simulation layer is constructed based on the battery electrochemical model, and a third computing layer is constructed based on electrical data analysis. Through full-connection integration of the layers, an electrical processor is formed. The third computing layer establishes electrical data interaction with an external electrical sensing terminal.

4. The data processing method for remaining battery power based on electrical sensing as described in claim 3, characterized in that, Virtual sensors are implanted at key locations in the battery electrochemical model, including: A first virtual sensor is set on the surface of the electrode particles in the battery electrochemical model. The sensing reading of the first virtual sensor is the lithium ion surface concentration. The lithium ion surface concentration directly affects the local overpotential and indirectly affects the total voltage. A second virtual sensor is set at the current collector interface of the battery electrochemical model, wherein the sensing reading of the second virtual sensor is the local current density, and the total current is quantized by integrating the local current density.

5. The data processing method for remaining battery power based on electrical sensing as described in claim 4, characterized in that, The battery electrochemical-electromagnetic coupling mechanism includes: Extract multidimensional electromagnetic disturbance features, wherein the multidimensional electromagnetic disturbance features include at least amplitude attenuation, phase shift, frequency dispersion and polarization rotation; The mapping between key internal states and multidimensional electromagnetic disturbance characteristics is determined as the electrochemical-electromagnetic coupling mechanism of the battery. The key internal states are determined at least by lithium ion concentration distribution, electrolyte conductivity and dielectric constant, and solid-liquid interface thickness.

6. The data processing method for remaining battery power based on electrical sensing as described in claim 1, characterized in that, Triggering the power processor to perform remaining power analysis under forward and backward propagation cross-verification includes: The battery electrical signal is collected through an external sensor. The electromagnetic wave signal is input into the power processor to perform forward propagation and determine the first data sequence; The battery electrical signal is input into the power processor, and back propagation is performed to determine the second data sequence; The first data sequence and the second data sequence are cross-validated to generate the remaining battery power data.

7. The data processing method for remaining battery power based on electrical sensing as described in claim 6, characterized in that, The data sequence includes electrochemical simulation data, virtual sensing data, and external battery electrical data; Among them, the sensor readings of the virtual sensor are used to synthesize the readings of virtual external electrical data.

8. The data processing method for remaining battery power based on electrical sensing as described in claim 7, characterized in that, Mutual verification of the first data sequence and the second data sequence includes: Perform data point mapping and verification on the first data sequence and the second data sequence, and determine whether the preset error amount is met; If the preset error amount is met, the mean of the mapped data points is calculated to determine the target data sequence; If there are data points that do not meet the preset error amount, the mean is calculated by re-interpolating the data points to determine the target data sequence.

9. The data processing method for remaining battery power based on electrical sensing as described in claim 8, characterized in that, The remaining battery power data includes the battery's intrinsic power and a multi-dimensional health score; The intrinsic capacity of the battery is determined based on the ratio of the spatial integral of the active lithium ion content in each electrode region to the theoretical maximum lithium ion capacity of the electrode region. The multidimensional health score includes lithium content loss, structural degradation, and electrolyte phase change.

10. The data processing method for remaining battery power based on electrical sensing as described in claim 1, characterized in that, Based on the remaining battery power data, a battery warning message is generated, which includes a power warning message and a health warning message. The remaining battery power data and the battery warning information are integrated and displayed in a pop-up window on the visual interface of the battery management platform.

Images