Lithium battery health state assessment method and system based on data analysis

Abstract

The invention discloses a lithium battery health state assessment method and system based on data analysis, and relates to the technical field of new energy battery health state assessment, and the method comprises the following steps: extracting historical data from a lithium battery operation log, constructing a working condition rhythm profile, recognizing an operation rhythm cycle inflection point, and generating a rhythm fingerprint band as a time alignment reference; and performing phase rearrangement on the time axes of the multiple operation characteristics by using the rhythm fingerprint band, so that the operation characteristics are synchronously expanded under the unified rhythm reference, and positioning a short-term synchronous stable section and generating a consistent crack cable. According to the method, by constructing the working condition rhythm profile and the rhythm fingerprint band, multi-feature time alignment and dynamic analysis are achieved, and the false stability phenomenon caused by the working condition rhythm is accurately recognized; and in combination with cooperative regulation and control of the dynamic threshold chain and the shadow beat dome, the false stable state is continuously stripped, the real degradation track of the battery is recovered, and dynamic tracking and high-precision evaluation of the health state are realized.

Description

Technical Field

[0001] This invention relates to the field of health status assessment technology for new energy batteries, specifically to a method and system for assessing the health status of lithium batteries based on data analysis. Background Technology

[0002] Data-driven lithium battery health status assessment refers to an evaluation method that uses naturally generated operational data throughout the entire lifecycle of a lithium battery as its core basis. This involves continuously processing and correlating voltage, current, temperature, capacity, energy efficiency, and time characteristics collected under charging, discharging, and resting conditions. The assessment extracts key features reflecting performance retention and degradation changes from the data and combines this with the evolution of multiple parameters over time to comprehensively judge the battery's current performance status. This assessment does not rely on a single test result or fixed empirical thresholds. Instead, it comprehensively characterizes the overall state of the lithium battery in terms of capacity retention, energy conversion characteristics, internal structural stability, and operational consistency through the analysis of long-term operational trajectories and the relationships between multi-dimensional characteristics. This makes the health status assessment closer to the actual battery operation process and provides continuity and traceability.

[0003] The existing technology has the following shortcomings: In existing technologies, lithium battery health status assessment based on data analysis typically relies on a comprehensive judgment of the synchronous changing trends of multiple operating characteristics within a certain time window. However, in actual operation, a staged self-consistency illusion can easily occur, where multiple key operating characteristics, such as capacity changes, internal resistance response, voltage curve morphology, and temperature characteristics, simultaneously exhibit a statistically stable or even slowly improving trend within a short time window. This apparent stability does not originate from a genuine improvement in the internal electrochemical or structural state of the lithium battery, but rather from changes in external conditions such as charging / discharging cycle switching rhythm, load level adjustment, or intermittent operating characteristics, creating a mutually offsetting effect of degradation characteristics at the statistical level. When determining health status, existing technologies often struggle to effectively distinguish between this statistical stability phenomenon caused by changes in operating cycle rhythm and the actual improvement in battery performance. This can easily lead to misjudging batteries that are actually in a rapid instability critical stage as healthy, thus delaying necessary risk identification, intervention, or maintenance, increasing the risk of sudden performance changes or safety failures in subsequent operation.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for assessing the health status of lithium batteries based on data analysis, so as to solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a lithium battery health status assessment method based on data analysis, comprising the following steps: Historical data is extracted from lithium battery operation logs to construct a working condition rhythm profile, identify the inflection point of the operating rhythm cycle, and generate a rhythm fingerprint as a time alignment benchmark. The time axis of multiple operational features is rearranged using rhythm fingerprint strips, so that the operational features are synchronously unfolded under a unified rhythm benchmark, the short-term synchronous stable section is located and a consistent crack line is generated. Based on the difference in the offset of the consistent crack cable during the continuous time window, feature cancellation information caused by the working condition rhythm is extracted, a list of cancellation traces is formed and pseudo-stable annotation frames are generated. Based on pseudo-stable labeled frames, rhythmic adaptive weight adjustment is applied to the running feature set. Sensitivity is increased and thresholds are adjusted at rhythm switching nodes to form a dynamic threshold chain. The original running sequence is sampled in reverse phase according to the dynamic threshold chain, and information is collected in a denser manner at rhythm abrupt changes and pulse turning points to reconstruct the running trajectory bundle that reflects the true performance evolution. By running the trajectory beam-driven timing and temperature coordinated control module, the charge / discharge rate and resting rhythm are adjusted and the energy slow-release structure is enabled to remove the pseudo-steady state and output the true health assessment results.

[0007] Preferably, the steps for constructing the operating condition rhythm profile and forming the rhythm fingerprint band are as follows: Historical data, including voltage change sequences, charge / discharge current curves, temperature change curves, capacity change records, energy efficiency change curves, and internal impedance change trajectories, were extracted from the lithium battery operation logs and rearranged in chronological order to form continuous time data. Based on continuous time data, a working condition rhythm profile is constructed. By superimposing and analyzing the changes of voltage, current, temperature, capacity, energy efficiency and internal impedance over time, the temporal distribution of the charging stage, discharging stage and resting stage is determined. Based on the operating condition rhythm profile, identify the turning points of changes in voltage curves, temperature curves, capacity curves, and internal impedance curves, and extract the periodic inflection point information of the operating rhythm. The inflection point information of the cycle is continuously woven in chronological order to form a rhythm fingerprint strip, which is used to establish a unified rhythm benchmark for subsequent multi-feature time alignment.

[0008] Preferably, the step of performing phase rearrangement on the time axis of multiple operational features and generating consistent crack lines using rhythmic fingerprint bands includes: Based on the periodic inflection points of the rhythm fingerprint, the time axes of the voltage change sequence, discharge current curve, temperature change record, capacity retention curve, energy conversion efficiency curve, and internal impedance change curve are repositioned to align all features in time. Under a unified rhythmic benchmark, all operational features are synchronously unfolded in cyclical order, so that each feature is arranged in parallel within the same rhythmic cycle. In a synchronously unfolded multi-feature time structure, identify time segments in which multiple operational features simultaneously exhibit stable trends, and record the rate of change and duration of each feature. The identified synchronous stable segments are connected in chronological order to form a consistency crack line that reflects the cooperative behavior of multiple features, which is used to reveal the synchronous stability relationship during battery operation.

[0009] Preferably, the steps for identifying pseudo-stable segments and generating pseudo-stable labeled frames within a continuous time window based on consistent crack lines are as follows: Based on the time nodes of the consistent crack, a continuous time window covering multiple rhythmic cycles is established, so that the voltage change curve, discharge current change trajectory, temperature change record, capacity retention curve, energy efficiency change record and internal impedance change curve are continuously arranged in the time dimension. Within a continuous time window, the changing trends of multiple operational features are tracked, and the direction, rate, and duration of change are recorded point by point to form a feature offset trajectory. Based on the feature offset trajectory, identify feature pairs with opposite directions of change, corresponding amplitudes of change, and consistent duration within the same rhythm segment, and record the feature names, time start and time end points, direction of change, amplitude of change, and rhythm segment number to form an offset trace list; The list of offset traces is mapped to the time series of consistent cracks, and time intervals with stable surfaces but internal feature offsets are selected to generate pseudo-stable labeled frames containing time information, feature names, change directions and rhythm segment numbers.

[0010] Preferably, when generating pseudo-stable annotation frames, overlapping intervals in the offset trace list that are continuously distributed in time and involve multiple operational features are identified as pseudo-stable segments. A time-series annotation structure is established based on the start and end times, participating feature names, change amplitudes, durations, and rhythm segment numbers of the pseudo-stable segments to identify the formation and duration range of pseudo-stable states in the rhythm dimension.

[0011] Preferably, the steps for constructing a dynamic threshold chain around the pseudo-stable labeled frame are as follows: Using the pseudo-stable label frame as a reference, the time distribution and weight parameters of voltage change curve, discharge current curve, temperature change record, capacity change trend curve, energy conversion efficiency change curve and internal impedance change record are initialized so that the operating characteristics correspond synchronously with the pseudo-stable state on the time axis. Based on the periodic nodes of the rhythm fingerprint, a weighted breathing gate that changes with the rhythm is introduced, so that the weights are periodically adjusted during the charging, discharging and resting phases for different operating characteristics. Adjust the rate of change of the weighted ventricular plexus at the rhythm switching node to enhance the sensitivity to characteristic changes and reduce inertial compliance; A dynamic threshold chain is established based on the trend of weight changes and the evolution of rhythm over time, so that the threshold nodes record the rhythm stage, weight value, critical level and response direction, thereby achieving adaptive matching with the running state.

[0012] Preferably, the steps of performing inverse traction sampling on the original running sequence and reconstructing the running trajectory bundle based on the dynamic threshold chain are as follows: Using a dynamic threshold chain as the basis for time control, the original running data is divided into time zones and rhythms are located to establish a sampling framework with time continuity. Reverse-phase traction sampling is performed with the rhythm change position and pulse inflection position as the sampling center, and key data are collected in a dense manner during the stages of voltage, current, temperature, capacity, energy efficiency and internal impedance changes. The encrypted data is spliced ​​and sequentially arranged with the original time series to achieve continuous connection between rhythmic abrupt changes and stable periods in the time dimension. The spliced ​​multidimensional time series is continuously arranged according to the rhythmic cycle to form an operation trajectory bundle containing characteristic data of the charging stage, discharging stage and resting stage, which is used to reflect the real evolution trend of lithium battery performance.

[0013] Preferably, in the encrypted acquisition process of reverse-phase traction sampling, the sampling range is extended bidirectionally between the end of the rhythmic stable region and the starting point of the next rhythm, with the rhythmic change center as the axis of symmetry. The sampling density is increased at voltage inflection points, current drop points, temperature reversal points, capacity decay starting points, energy efficiency fluctuation peak points, and internal impedance change starting points to ensure the continuity and integrity of key state transition information in the time series.

[0014] Preferably, the steps for driving the shadow beat dome by running the trajectory beam and achieving pseudo-steady state stripping are as follows: Based on the time and temperature correspondence information in the running trajectory bundle, a time-series and temperature coordinated response framework is established to enable the electrical parameters and thermal properties to form a dynamic coupling relationship. Based on the rhythmic periodic characteristics of the running trajectory bundle, a shadow beat dome structure is constructed, and the charging active phase, the discharge release phase, the static buffer phase and the temperature recovery phase are mapped as dome control areas. The charging rate, discharging rate and resting rhythm are alternately regulated by the shadow beat dome, so that energy input, release and recovery form a continuous cycle; An energy-releasing ridge structure consisting of a time-releasing layer, a temperature-releasing layer, and a capacity recovery layer is introduced at the beat transition point to smooth the energy conversion process. By utilizing the synergistic effect of the shadow beat dome and the energy-releasing ridge, pseudo-stable states are stripped away, resulting in an assessment of the battery's health evolution.

[0015] A data-driven lithium battery health status assessment system includes a rhythm modeling module, a feature alignment module, a pseudo-stability identification module, a weight adjustment module, a trajectory reconstruction module, and a health assessment module. The rhythm modeling module extracts historical data from lithium battery operation logs, constructs a working condition rhythm profile, identifies the inflection point of the operating rhythm cycle, and generates a rhythm fingerprint as a time alignment benchmark. The feature alignment module uses rhythm fingerprint strips to rearrange the phases of the time axes of multiple running features, so that the running features can be synchronously unfolded under a unified rhythm benchmark, locate short-term synchronous stable sections and generate consistent crack lines. The pseudo-stability identification module tracks the difference in operational feature offsets of the consistent crack cable within a continuous time window, extracts feature cancellation information caused by the working condition rhythm, forms a list of cancellation traces, and generates pseudo-stability annotation frames. The weight adjustment module applies rhythmic adaptive weight adjustment to the running feature set based on pseudo-stable labeled frames, increases sensitivity and adjusts thresholds at rhythm switching nodes, forming a dynamic threshold chain. The trajectory reconstruction module performs reverse-phase traction sampling on the original running sequence according to the dynamic threshold chain, and densifies the information collection at rhythm abrupt changes and pulse turning points to reconstruct the running trajectory bundle that reflects the true performance evolution. The health assessment module, through the operation of the trajectory beam-driven timing and temperature coordinated control module, adjusts the charge / discharge rate and resting rhythm, and enables the energy slow-release structure to eliminate pseudo-steady states and output true health assessment results.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention introduces the construction of operating condition rhythm profiles and rhythm fingerprints, enabling rhythm alignment and dynamic expansion of the time axis of operational data across multiple feature dimensions. This effectively identifies feature cancellation phenomena caused by changes in external operating condition rhythms, thereby avoiding misjudgments of staged spurious stability in health assessments. The method establishes a time-consistent benchmark at the feature level, allowing operational features such as capacity, internal resistance, voltage, and temperature to be correlated under a unified rhythm, achieving continuous capture of the true performance change trajectory and significantly improving the consistency between health status assessment results and the actual battery operating state.

[0017] This invention utilizes the coordinated regulation of a dynamic threshold chain and a shadow beat dome to enable an adaptive feedback mechanism in the time, temperature, and energy dimensions of the health assessment process. This mechanism continuously strips away pseudo-steady states during rhythm switching and energy release, ensuring that subtle degradation characteristics exhibited by the battery during dynamic operation are gradually restored, thereby achieving continuous tracking of health evolution trends. This method transforms battery health assessment from static judgment to dynamic evolution identification, possessing higher sensitivity and long-term predictive reliability. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart of the lithium battery health status assessment method based on data analysis according to the present invention.

[0020] Figure 2 This is a schematic diagram of the modules of the lithium battery health status assessment system based on data analysis of the present invention. Detailed Implementation

[0021] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0022] This invention provides, for example Figure 1 The data analysis-based lithium battery health status assessment method shown includes the following steps: Historical data is extracted from the operation logs of lithium batteries to construct a working condition rhythm profile, identify the periodic inflection points in the operating rhythm, and continuously weave the periodic inflection point information into a rhythm fingerprint band to provide a unified rhythm benchmark for subsequent time alignment. To accurately represent the rhythmic patterns formed by lithium batteries at different operating stages over time, and to ensure that subsequent multi-dimensional operating characteristics can be aligned within a unified time frame, it is necessary to construct a rhythmic fingerprint band that reflects the changing patterns of operating conditions. The specific implementation steps are as follows: Historical data containing operational behavior characteristics is extracted from the operation logs generated throughout the lithium battery's entire service life. The operation logs include various parameter information automatically recorded during battery charging and discharging, specifically: voltage change sequences, charging current curves, and charging duration recorded during the charging phase; discharge current change trajectories, discharge voltage curves, and depth of discharge change records recorded during the discharging phase; and temperature change curves, terminal voltage holding status, and internal impedance change trends recorded during the resting phase. Simultaneously, time-related operational information is extracted from the operation logs, including cycle numbers, ambient temperature records, operating mode switching times, and load power level change records. Through the extraction of this data, the operational trajectory of the lithium battery from initial use to the current state can be completely reconstructed. After extraction, all historical data is rearranged according to timestamp order, removing discontinuous record segments and abnormal jump points to ensure that charging / discharging, resting, and load switching information remain continuous on the timeline. The processed data is labeled according to operational phases, enabling subsequent analysis to accurately identify the operational characteristics and temporal relationships corresponding to each phase, providing a basic foundation for constructing an operational rhythm profile.

[0023] Based on the compiled continuous-time data, a circadian rhythm profile reflecting the changing characteristics of lithium-ion batteries over time is constructed. This profile is obtained by comprehensively analyzing the temporal relationships of multiple operating parameters. Specifically, the curves of voltage, current, temperature, capacity, energy efficiency, and internal impedance changes within the same time period are overlaid and analyzed to identify characteristic combinations exhibiting synchronous changing trends over time. In each operating cycle, the battery experiences a charging rise phase, a constant-voltage charging phase, a resting phase, a discharging decline phase, and a second resting phase. By calculating the temporal arrangement and duration of these phases, the basic form of the circadian rhythm is formed. During the construction process, the voltage rise rate during the charging phase, the current drop amplitude during the discharging phase, the temperature recovery rate during the resting phase, and the capacity recovery ratio are used as core reference indicators to depict the connection and transition patterns between different operating conditions. The final circadian rhythm profile uses time as the horizontal axis and the relative changes of multiple parameters as the vertical axis. By marking the charging / discharging switching nodes and temperature recovery points in the continuous time space, a profile layer reflecting the rhythm of operating condition changes is formed. This profile fully describes the rhythmic transition pattern of the battery between different usage stages, providing a temporal structure basis for the next step of identifying cycle inflection points.

[0024] After obtaining the operating rhythm profile, key locations reflecting rhythm changes are identified one by one to determine the cyclic inflection points of the operating rhythm. Specifically, the nodes where the voltage curve changes from rising to flat during the charging phase, from flat to falling during the discharging phase, from rising to falling during the temperature phase, from increasing to decreasing during the capacity curve, and from increasing to decreasing during the internal impedance curve are identified in the profile curve. Each node corresponds to a critical point of change in operating state, representing the moment when the battery operating condition transitions from one operating rhythm to another. During the identification process, the positions of these nodes on the time axis are recorded, and the time intervals between nodes, the magnitude and direction of parameter changes are stored together to form a complete set of cyclic inflection point information. To ensure the continuity of the rhythm, each inflection point is accompanied by a description of the state of the operating stages before and after it; for example, the previous stage belongs to the high-rate discharge stage, and the next stage belongs to the low-load resting stage. In this way, the cyclic inflection points not only include the time position but also retain the state characteristics of the operating condition transition, making the rhythm information contextually continuous and traceable.

[0025] After obtaining the complete set of cycle inflection points, this inflection point information is continuously woven in chronological order to form a rhythmic fingerprint strip for global time alignment. The rhythmic fingerprint strip connects the cycle inflection points sequentially along the time axis, creating a periodic and directional time chain for different operating stages. Specifically, in the fingerprint strip, each inflection point corresponds to a rhythmic node, the time interval between adjacent inflection points is defined as the rhythmic interval, and the connection direction between inflection points indicates the trend of operating condition changes. Through this continuous connection method, the repetitive structure of the lithium battery operating cycle can be completely displayed on the fingerprint strip. For example, a typical cycle may sequentially include a charging ramp-up node, a constant voltage conversion node, a resting recovery node, a discharge start node, a discharge termination node, and a temperature stabilization node. During the formation of the rhythmic fingerprint strip, the time information, parameter change description, and state transition direction of each node are embedded into the time axis in a structured form, making the rhythmic information extensible and globally consistent in the time dimension. After the rhythmic fingerprint strip is generated, it can serve as a unified rhythmic benchmark, providing a time reference for subsequent multi-feature time axis alignment, phase rearrangement, and synchronous stable segment identification. By establishing rhythm fingerprints, the time structure of lithium battery operating data can be standardized. The rhythm differences between different operating cycles can be mapped and compared on the same benchmark, enabling subsequent data analysis to be carried out under a unified time coordinate system, and realizing cross-cycle feature correlation and continuous performance change identification.

[0026] The rhythm fingerprint strip is used to perform phase rearrangement operation on the time axis of multiple operating features, so that all operating features are synchronously unfolded under a unified rhythm benchmark, the time segment that shows a synchronous and stable trend in the short term is located, and a consistency crack line reflecting the convergent behavior of multiple features is generated. To reveal the true correlations between battery operating characteristics under a unified time benchmark, ensuring rhythmic synchronization of different characteristics and identifying short-term cooperative stable behavior, it is necessary to use rhythmic fingerprinting to rearrange the time axes of each operating characteristic and expand all features under a unified rhythm benchmark to generate a structured result—a consistent crack loop—characterizing the synchronous stable trend of multiple features. The specific implementation steps are as follows: After obtaining the rhythm fingerprint, the time axes of each feature are precisely repositioned and aligned with the reference points according to the time nodes defined by the rhythm fingerprint, based on the multi-dimensional operating features extracted from the lithium battery operation log. Specifically, each cycle inflection point in the rhythm fingerprint is used as a time positioning reference point, and the original time stamps of the voltage change sequence during the charging phase, the current response curve during the discharging phase, the temperature change record during the resting phase, the capacity retention curve, the energy conversion efficiency sequence, and the internal impedance change curve are remapped. This mapping process ensures that the key nodes representing changes in operating conditions in different features are strictly aligned in time. For example, the charging rise node, constant voltage retention node, discharge start node, resting recovery node, and temperature inflection point identified in the rhythm fingerprint are used as common alignment references, so that each feature corresponds to the same operating phase at the same time. To maintain the continuity and integrity of the data, the time interval of each feature is redefined so that it can still maintain the original sampling density and time order after rearrangement. Through this step, the operating features that were originally scattered at different time scales are unified into the same rhythmic time frame, laying a precise time reference for subsequent synchronization.

[0027] After aligning the time of each operating characteristic, all operating characteristics are synchronously unfolded under a unified rhythmic benchmark according to the periodic sequence defined by the rhythmic fingerprint. Synchronous unfolding refers to arranging the changes of each characteristic within the same rhythmic cycle in parallel on a unified time axis, ensuring that characteristics such as voltage, current, temperature, capacity, energy efficiency, and internal impedance have a one-to-one corresponding time position at the same rhythmic node. To achieve this, the inflection point of the rhythmic fingerprint cycle is used as a dividing marker, dividing the entire operation process into multiple rhythmic units, each representing a complete operating cycle. Within each rhythmic unit, the voltage rise segment of the charging phase, the current fall segment of the discharging phase, the temperature stabilization segment of the resting phase, the capacity recovery segment, and the energy efficiency change segment are synchronously unfolded. This parallel unfolding method ensures that all operating characteristics exhibit a consistent starting point, the same rhythmic interval, and a continuous time progression relationship in the time dimension, thereby eliminating time misalignment problems caused by differences in operating conditions. For example, within a unified rhythm framework, the rising position of the voltage curve corresponds to the rising position of the temperature curve, and the steady segment of the capacity curve corresponds to the falling segment of the current curve in time, all within the same rhythmic phase. Through this unified unfolding, the correlation of multiple features in rhythm is clarified, forming a comparable and observable time-synchronized structure.

[0028] After all features have been synchronously unfolded, time segments in which multiple operating features simultaneously exhibit stable trends under a unified rhythm benchmark are identified. The key to this step is to discover time periods in the synchronously unfolded multi-feature time structure where different features exhibit similar directions of change and stabilization characteristics in the short term. Specifically, the periodic units defined by the rhythm fingerprint are used as observation windows, and the change trends of each feature within each window are analyzed segment by segment. Within the same time period, if the voltage change amplitude gradually decreases, the current fluctuation intensity continues to converge, the temperature change rate tends to level off, the capacity change gradient approaches zero, and the energy efficiency fluctuation range significantly narrows, then this time period can be identified as the synchronous stable region of multiple features. Within each synchronous stable region, the time position, direction of change, rate of change, and duration of each feature are further recorded, and the relative time offset between features is calculated to characterize their degree of synchronization. By organizing this information, several multi-feature synchronous stable segments can be identified on the time axis. These segments reflect the stage in which different parameters of the lithium battery simultaneously enter a steady state during operation, i.e., the cooperative stabilization behavior among operating features. The identified synchronous stable segments provide a time basis for subsequently establishing a consistent structure.

[0029] After identifying the synchronous stable segments, the information from these segments is structured and woven into a consistency crack cable reflecting the cooperative behavior of multiple features. The consistency crack cable is a time-series structure used to describe the cooperative stabilization relationship between various operating features under a unified rhythmic benchmark. Specifically, all synchronous stable segments identified in the previous sub-step are arranged sequentially in chronological order, with each segment corresponding to a set of features exhibiting a synchronous stabilization trend. For each set, the name, start time, end time, direction of change, magnitude of change, and synchronization difference value with other features are recorded for each feature within that set. Subsequently, feature sets from adjacent time periods are linked together to form a continuous time chain, ensuring the complete preservation of the connection between consecutive stable segments. Each chain node represents the synchronous stable state of a feature group, and the connecting lines between nodes represent the synchronous evolution path between features in different time periods. When multiple sets of features exhibit a continuous stable trend in adjacent rhythmic cycles, these nodes form continuous chain segments, constituting a complete consistency crack cable. This crack cable can intuitively reflect the cooperative change relationship of different operating features over time, revealing the stable state evolution trajectory of the battery during short-term operation. By establishing a consistent crack, we can not only identify the cooperative stability mode among multiple features, but also provide a key basis for the subsequent identification of pseudo-stability caused by operating condition rhythm, so that the battery operation data has an inherent connection and dynamic consistency in the rhythm dimension.

[0030] Based on the consistent crack line, the subtle offset differences of multiple operating features are tracked within a continuous time window. The mutual cancellation information of features caused by the operating condition rhythm is extracted to form a cancellation trace list. Based on the cancellation trace list, pseudo-stable segments are selected and corresponding pseudo-stable annotation frames are generated. Identifying statistical pseudo-stability caused by changes in battery operating rhythm during operation requires relying on the formed consistent crack lines to track subtle shifts between multiple operating features within a continuous time window. This involves extracting the parameter cancellation relationships caused by rhythm switching and constructing a list of cancellation traces. This list is then used to filter out time segments that appear stable but actually hide degradation trends, generating corresponding pseudo-stability annotation frames. The specific implementation process is as follows: Based on the obtained consistent crack lines, a continuous time window covering multiple rhythmic cycles is established to track the temporal evolution of various operational characteristics. Specifically, the time of each node recorded in the consistent crack line is used as the reference point for window division, and the time interval between nodes is used as the width of a single window, forming a continuous and seamless time-sliding structure. Each time window covers a complete operational rhythm segment, including charging, constant voltage, resting, discharging, and dormancy phases. Then, the voltage change curve, discharge current change trajectory, temperature change record, capacity retention curve, energy efficiency change record, and internal impedance change curve are sequentially unfolded according to the time order of the consistent crack lines, ensuring all features are continuously arranged in the time dimension. To ensure the continuity of observation, the start time of each window partially overlaps with the end time of the previous window, allowing for a smooth transition of feature changes between windows. Within each time window, a time set containing multiple feature curves is formed, providing a complete and continuous data environment for subsequent identification of feature shifts.

[0031] Within a continuous time window, the changing trends of operating characteristics such as voltage, discharge current, temperature, capacity, energy efficiency, and internal impedance are meticulously tracked to identify subtle shifts between them. This tracking process uses each time window as an analysis unit, recording the direction, rate of change, and duration of each characteristic's change on the time axis point by point. Taking the charging phase as an example, as the voltage curve continuously rises, the changing trend of the current curve and the rate of temperature rise are observed simultaneously. When the voltage continues to rise while the current slows down or slightly decreases, and the temperature change tends to stabilize, it indicates that a relative shift has occurred between the characteristics. Similarly, in the discharging phase, if the capacity curve decreases more significantly while the rate of rise of the internal impedance curve is delayed, it also indicates a shift between the two. In this process, the direction of change of each characteristic over time is labeled as rising or falling, the rate of change is described in words as accelerating, slowing down, or remaining stable, and its duration is recorded. Through this point-by-point recording and continuous time arrangement, a complete characteristic shift trajectory can be formed, reflecting the subtle differences between the characteristics during rhythm switching. These shift trajectories are direct evidence revealing the mutual cancellation relationships between characteristics.

[0032] After obtaining the multidimensional feature offset trajectories, the mutual cancellation phenomena caused by changes in the operating rhythm are identified and recorded in chronological order to form a list of cancellation traces. In practice, the feature offset trajectories within each time window are compared pairwise to find feature pairs that exhibit opposite directions, similar amplitudes, and similar durations within the same rhythmic segment. For example, if the voltage curve continuously rises over a certain period while the current curve slowly declines over the same period, and both have the same duration and similar amplitude ratio, it indicates that the increase in voltage is offset by the decrease in current. Similarly, if the temperature curve rises during the discharge phase while the energy efficiency curve declines, and both have the same time position and duration, this also constitutes a cancellation relationship. For each identified cancellation relationship, the feature name, time start point, time end point, direction of change, amplitude of change, and corresponding rhythmic segment number are recorded item by item and arranged in chronological order to form a list of cancellation traces. This list fully describes the offsetting relationships between features, revealing the complementary behavior of internal parameters behind the apparent stable phase. By using the offset trace list, one can intuitively understand which features reach a balance within a specific time period, and whether this balance is caused by the switching of operating rhythms.

[0033] Based on the information recorded in the offset trace list, pseudo-stable segments are selected on the time axis, and corresponding pseudo-stable annotation frames are generated. A pseudo-stable segment refers to a time interval that appears stable on the surface, but where multiple operating features within it exhibit mutual cancellation behavior. Specifically, the time records in the offset trace list are mapped to the time series of the uniform crack strands, and the time distribution is statistically overlapped. When the cancellation relationships of multiple features converge in the same rhythmic phase and the time span is continuous, this continuous interval is marked as a pseudo-stable segment. After identifying the pseudo-stable segment, its corresponding start time, end time, involved feature names, direction of change of each feature, magnitude of change, duration, and rhythmic segment number are extracted, and this information is summarized to form a pseudo-stable annotation frame. The pseudo-stable annotation frame uses time as the main axis, providing a structured description of the start and end points, participating features, cancellation relationships, and rhythmic position of each pseudo-stable segment. Through this annotation result, the mutual cancellation process of features such as voltage, discharge current, temperature, capacity, energy efficiency, and internal impedance during apparent stability can be visually presented in the time dimension. Each pseudo-stable labeled frame corresponds to a potential pseudo-stable stage, reflecting that although the battery operating parameters appear stable during this stage, the internal state has already experienced degradation and offsetting phenomena. The generation of pseudo-stable labeled frames enables the entire battery operating time series to have identifiable pseudo-stable identifiers, providing a precise location basis for subsequent accurate assessment of health status and pseudo-stable separation.

[0034] A rhythm-dependent weighted breathing gag is introduced around the pseudo-stable labeled frame to the running feature set. This increases the sensitivity to feature changes and reduces inertial compliance at rhythm switching nodes, and constructs a dynamic threshold chain that adaptively matches the running state. After identifying pseudo-stable segments, to enable operational features to respond instantly to rhythm changes during analysis and avoid inertial misjudgments or response lags caused by operational condition switching, it is necessary to adjust the weights of the operational feature set according to rhythm changes based on the pseudo-stable annotation frames formed in the previous stage, and actively adjust the sensitivity of features to changes at rhythm switching nodes to reduce inertial compliance, ultimately establishing a dynamic threshold chain that matches the operational state in real time. The specific implementation steps are as follows: Using pseudo-stable annotation frames as a reference, the temporal distribution and weight parameters of the operational feature set are initialized to ensure a one-to-one correspondence between each operational feature and the pseudo-stable annotation information in the time dimension. Specifically, the time start point, time end point, involved operational feature names, rhythm stage numbers, cancellation durations, and change directions contained in the pseudo-stable annotation frames are extracted sequentially and mapped to the time series of operational features. The operational feature set includes voltage change curves, discharge current curves, temperature change records, capacity change trend curves, energy conversion efficiency change curves, and internal impedance change records. For operational features within pseudo-stable segments, their weight values ​​are adjusted to a lower level to reduce the interference of the pseudo-stable stage in the overall state assessment. For feature segments outside the pseudo-stable annotation range, the weights are maintained at the baseline level. Through this distribution initialization, each operational feature has a weight base synchronized with the pseudo-stable state on the time axis, allowing for adjustable weights in subsequent rhythm changes and providing a clear initial boundary for rhythmic regulation.

[0035] After weight initialization, based on the rhythm cycle defined by the rhythm fingerprint, a weighted breathing gate is introduced into the set of operating features, periodically adjusted according to the rhythm state, allowing different features to alternately gain dominance in the rhythm cycle. In specific implementation, the cycle nodes defined in the rhythm fingerprint are used as the control beat points of the breathing gate. Each rhythm cycle is divided into several stages, including the charging rise stage, constant voltage holding stage, discharging decline stage, resting recovery stage, and environmental equilibrium stage. During the charging rise stage, the weight values ​​of voltage, discharging current, and capacity characteristics are gradually increased, as these characteristics most directly reflect the battery's charging response process; simultaneously, the weight values ​​of internal impedance, energy conversion efficiency, and temperature characteristics are gradually decreased to reduce their impact during the charging stage. During the discharging stage, the weights are adjusted in reverse, increasing the weight values ​​of discharging current, energy efficiency, and temperature, and decreasing the influence of voltage and capacity; during the resting recovery stage, the weights of temperature and internal impedance are gradually increased to reflect the battery's internal recovery state during the dormant phase. By adjusting the weights in a breathing-like manner that changes with the rhythm state, the operating characteristics automatically switch priority response objects at different stages, ensuring that the analysis process remains consistent with the actual operating rhythm of the battery in the time dimension. The adjustment range of the weight breathing gate is set according to the duration of the rhythm cycle, the rate of change of the characteristics, and the interval between rhythm nodes, making its change process both smooth and responsive.

[0036] At rhythm switching nodes, the rate of change of the weighted breathing valve is precisely adjusted to enhance feature sensitivity and weaken inertial compliance, enabling the analysis results to respond quickly to abrupt changes in rhythm state. In specific implementation, the rhythm switching nodes recorded in the rhythm fingerprint are used as control trigger points. When a phase transition occurs in the rhythm (e.g., from charging to constant voltage holding, from discharging to rest, or from temperature recovery to the start of the next cycle), the rate of change of feature weights is automatically adjusted. During the preparation phase before node switching, the impending operational condition transition is identified in advance, causing the weights of operating features related to that rhythm phase to decrease slowly, forming a transition band. When the switching node arrives, the rate of increase in the weights of sensitive features is rapidly increased. For example, at the end of the charging phase, the weight of the voltage feature increases briefly, while the weight of the discharge current decreases rapidly, and the weight of the temperature feature subsequently increases to reflect the response during the thermal equilibrium phase. Simultaneously, the weight of the internal impedance feature is temporarily reduced after the rhythm switching to minimize its inertial influence on short-term state changes. At each rhythm switching node, a weight adjustment transition zone is established. The time span of the transition zone is determined according to the rhythm interval to ensure that the weight changes are neither abrupt nor inconsistent, and accurately respond to the state transition. In this way, at the rhythm switching node, the running feature set can quickly redistribute weights in a short period of time, thereby improving the responsiveness to feature changes and reducing judgment errors caused by inertial delays.

[0037] After completing the rhythm synchronization weight adjustment and node response adjustment, a dynamic threshold chain that adaptively matches the operating state is constructed based on the weight change trend of the operating features and the time evolution law of the rhythm state. In specific implementation, the weight value of each operating feature in the time series is associated with the state of the corresponding rhythm stage one by one, and threshold adjustment rules are established using the rhythm nodes as reference points. During rapid rhythm changes (such as discharge initiation or charging completion), the critical value of the dynamic threshold chain decreases, enabling the system to more sensitively capture feature changes; during stable rhythm changes (such as constant voltage maintenance or static recovery), the critical value of the threshold chain gradually increases, weakening the analysis process's response to small fluctuations and thus suppressing the influence of non-real fluctuations. The dynamic threshold chain runs chronologically throughout the entire operating cycle, with each threshold node recording the current rhythm stage, weight value, critical level, and response direction. During operation, the threshold chain is continuously updated with changes in the weighted breathing threshold, ensuring that the judgment criteria at each moment remain consistent with the current operating rhythm. For example, when the cycle is in the discharge phase, the critical values ​​of the dynamic threshold chain are sequentially associated with three characteristics: discharge current, temperature, and energy efficiency. When the cycle enters the resting recovery phase, the threshold chain automatically shifts its focus to the temperature and internal impedance change thresholds. Through this continuous threshold adaptive mechanism, the judgment boundaries of operating characteristics in each cycle phase are always dynamically matched, ensuring that changes in the battery's true state can be correctly identified under different operating conditions.

[0038] Based on the dynamic threshold chain, reverse phase traction sampling is performed on the original running sequence, and key change information is collected in a denser manner at rhythm abrupt change positions and pulse turning points to reconstruct the running trajectory bundle that reflects the true performance evolution trend; To accurately extract the true change characteristics of the battery during the abrupt change phase of its operating rhythm after the establishment of the dynamic threshold chain, and to avoid missing key state information due to uneven sampling intervals or fluctuations in operating conditions, it is necessary to perform reverse-phase traction sampling on the original operating data sequence based on the dynamic threshold chain. In this process, the sampling focus is on rhythm abrupt change locations and pulse transition locations. Enhancing the sampling density to supplement transitional information that cannot be captured by conventional sampling intervals, and forming an operating trajectory bundle that reflects the true performance evolution of the battery through time splicing and multi-dimensional reconstruction, is then achieved. The specific implementation process is as follows: Using a dynamic threshold chain as the basis for time control, the raw operating data is partitioned by time and mapped to corresponding rhythms to establish a basic time framework that can be sampled in reverse phase. Specifically, the rhythm stage number, rhythm node time, threshold change trend, response direction, and weight adjustment information recorded in the dynamic threshold chain are read and mapped one-to-one with the time axis of the raw operating log. The raw operating log includes charging voltage change curves, discharging current change curves, temperature change curves, capacity retention rate change curves over time, energy conversion efficiency change curves, and internal impedance change records over time. Through the critical value distribution of the dynamic threshold chain, the entire time series is divided into several rhythmic period segments, each covering the complete time period from the start to the end of the rhythm. Within each rhythmic period, it is further subdivided into rapid change areas, gradual change areas, rhythm switching areas, and state recovery areas based on threshold change characteristics. The original timestamp and corresponding feature value are retained within each segment to ensure accurate traceability in subsequent sampling processes. In this way, a partitioned structure with temporal continuity and rhythmic consistency is established, enabling each operating stage to have precisely locatable sampling boundaries.

[0039] After time partitioning, the rhythm abrupt change positions and pulse transition positions identified by the dynamic threshold chain are used as sampling centers to perform reverse-phase traction sampling, thereby enhancing the capture of key state change information. Specifically, the rhythm abrupt change position is used as the center point, extending the time window forward to the end of the previous rhythm's stable region and backward to the start of the next rhythm, forming a bidirectional traction sampling range. Within this range, continuous data points of features such as voltage, current, temperature, capacity, energy efficiency, and internal impedance are extracted from the original time series. For voltage change curves, dense sampling is performed at the transition point where the voltage changes from rising to slowly increasing during the charging phase, ensuring a complete record of charging rate changes. For discharge current curves, dense time data is collected within the segment where a rapid drop occurs before the end of discharge, capturing the true duration of the current plunge. For temperature change records, dense sampling is performed near the inflection points where the temperature changes from rising to falling or vice versa, ensuring that information on the response of thermal changes to energy loss is not missed. For capacity retention curves, dense sampling is performed at locations where the rate of capacity change abruptly changes, accurately recording the critical inflection point of battery activity degradation. For energy conversion efficiency curves, high-density sampling is performed during the period of greatest energy utilization fluctuation during rhythm switching, revealing the direct correlation between efficiency changes and operating rhythm. For internal impedance change records, dense sampling is performed during the sudden phase where impedance begins to rise or fall, ensuring a complete preservation of the correspondence between internal structural changes and operating states. Through this reverse-phase traction method, all sampling operations are carried out around the center of rhythm abrupt changes, ensuring complete coverage of key change phases.

[0040] After reverse-phase traction sampling is completed, the encrypted rhythmic abrupt change data and the original time series are spliced ​​and ordered to ensure continuous connection of sampling data from different rhythmic segments in the time dimension. Specifically, the encrypted sampling data of each feature is re-inserted into the original time series according to the rhythmic time sequence, ensuring that the start, transition, and end points of each rhythmic cycle are interconnected. To guarantee temporal continuity, time calibration is performed on the transition between rhythmic abrupt change segments and stable segments, ensuring that the inflection points of the voltage curve, the peak points of the current curve, the peak points of the temperature curve, the decay start points of the capacity curve, the fluctuation start points of the energy efficiency curve, and the initial rising point of the internal impedance curve are consistent in time. After splicing, a complete continuous time trajectory is formed, with each trajectory containing the complete change process from the start to the end of the rhythm. Subsequently, all feature curves are superimposed in chronological order to construct a time-aligned multidimensional feature structure. In this structure, each time point simultaneously contains the corresponding values ​​of voltage, current, temperature, capacity, energy efficiency, and internal impedance, forming a synchronous change record in the time dimension. Through this multi-dimensional splicing method, the originally independent and scattered operational data is reorganized into a whole operational trajectory with temporal continuity and physical consistency.

[0041] Based on the spliced ​​and organized multidimensional time series, an operational trajectory bundle reflecting the true performance evolution trend of lithium batteries is reconstructed. In specific implementation, the multidimensional feature sequence of each rhythmic cycle is treated as a time unit, and all cycles are arranged sequentially to form an operational trajectory set composed of multiple rhythmic units. Each time unit contains characteristic data for the charging phase, constant voltage holding phase, discharging phase, resting phase, and temperature recovery phase. For the trajectory bundle of the charging phase, the voltage rise curve, current change curve, and temperature increase curve are synchronized to form the charging response trajectory; for the trajectory bundle of the discharging phase, the discharge current decrease curve, capacity decay curve, and temperature rise curve are superimposed to form the discharging response trajectory; for the trajectory bundle of the resting phase, the temperature recovery curve, internal impedance stability curve, and energy efficiency recovery curve are combined to form the resting recovery trajectory; for the integrated trajectory of the entire cycle, the trajectories of the above three stages are connected sequentially to form a complete rhythmic evolution curve. In this way, multiple rhythmic cycles are continuously integrated into a unified operational trajectory bundle. Each point in the trajectory bundle corresponds to a real-time node, recording the changes in voltage, discharge current, temperature, capacity, energy efficiency, and internal impedance at that moment. The operational trajectory bundle not only shows the individual changing trends of each feature in each rhythmic stage but also reveals the temporal synergistic relationships between them, thus clearly reflecting the continuous evolution of lithium battery performance over time. The resulting operational trajectory bundle provides continuous, realistic, and traceable operational data for subsequent health status assessments, making health evaluation results closer to the actual degradation process of the battery.

[0042] By driving a shadow beat dome that coordinates timing and temperature with the operation of the trajectory beam, the charging rate, discharging rate and resting rhythm are alternately regulated, and the energy-releasing ridge structure is activated simultaneously, so that the pseudo-steady state is continuously stripped away during the dynamic regulation process, thereby obtaining an assessment result that reflects the true health evolution state. After obtaining the operating trajectory bundle, further stripping away the pseudo-stable state, restoring the battery's true operating trajectory, and identifying its health evolution patterns requires establishing a collaborative control structure driven by time variation and temperature response, with the operating trajectory bundle as the core driver. This involves using a shadow beat dome to alternately regulate the charging rate, discharging rate, and resting rhythm, while simultaneously introducing an energy-releasing ridge structure to maintain a dynamic balance between energy release and absorption. This gradually strips away the pseudo-stable state hidden beneath the surface of stable operation, yielding a health assessment result that reflects the battery's true performance evolution characteristics. The specific implementation process is as follows: Based on the time-temperature correspondence information contained in the operating trajectory bundle, a time-temperature coordinated response framework is established to determine the changing patterns of the battery's operating state in different rhythmic cycles. Specifically, the sequences of charging voltage, discharging current, temperature changes, capacity changes, energy conversion efficiency, and internal impedance changes over time recorded in the operating trajectory bundle are extracted and arranged sequentially along the time axis to form a continuous time stream. Within each rhythmic cycle, the direction and magnitude of the temperature change rate are monitored, identifying the temperature rise phase as the energy input zone, the temperature fall phase as the energy release zone, and the stable temperature zone as the resting recovery zone. In the time dimension, the voltage rise interval, current fall interval, capacity recovery interval, and temperature change interval are mapped one-to-one, establishing a dynamic coupling relationship between the battery's electrical parameters and thermal characteristics. In this way, a coordinated framework with temperature change as the time-driven signal is established, enabling the simultaneous recording and coordinated response of the battery's energy input, output, and recovery characteristics in different rhythmic stages, providing a stable temperature-time coupling foundation for the subsequent construction of the shadow beat dome.

[0043] After the timing and temperature coordination framework is established, a shadow beat dome structure is constructed based on the rhythmic periodic characteristics in the running trajectory bundle to control multi-layered energy distribution and state transitions within the time domain. In specific implementation, each rhythmic cycle is divided into four parts: an active charging phase, a discharge release phase, a resting buffer phase, and a temperature recovery phase. These parts are mapped to four control zones of the shadow beat dome in chronological order. The upper beat controls the frequency of charging rate changes, the middle beat controls the rhythm of discharge rate changes, the lower beat controls the distribution and duration of resting time, and the bottom beat controls the temperature recovery rate. The beats at each layer are interconnected through time bridging regions, forming a complete rhythmic cycle of energy input and release. Driven by the running trajectory bundle, the shadow beat dome adjusts the start time and duration of each beat in real time as time progresses. For example, when the operating trajectory beam shows an increased rate of voltage change and a rapid rise in temperature, the dome automatically extends the duration of the lower layer's temperature recovery cycle; when the temperature remains stable and capacity changes slow down, the dome shortens the resting cycle and increases the charging cycle frequency, making energy distribution more balanced. Through this multi-layered temporal coordination, the shadow beat dome forms a dynamic breathing-like regulation mechanism throughout the entire operating cycle, providing a time basis for the rhythmic balance of battery operation.

[0044] After the shadow beat dome is established, the charging rate, discharging rate, and resting rhythm are alternately regulated to create a continuous cycle of energy input, energy release, and energy recovery in the battery over time. Specifically, the rate of change of the charging rate is directly correlated with the duration of the upper-level beat. At the start of the charging beat, the voltage input intensity is gradually increased, causing a synchronous increase in current response. As the upper-level beat reaches its end, the charging rate is gradually decreased to suppress the upward trend in the battery's internal temperature. Subsequently, the middle-level beat takes over control, entering the discharging phase, gradually increasing the discharging rate to maintain a controllable ratio between the change in discharge current and the temperature rise, ensuring a stable energy release process. When the discharging beat ends, the shadow beat dome automatically switches to the resting beat. The lower-level beat puts the battery in a resting state without current flow, the internal temperature gradually decreases, the electrochemical reaction rate decreases, and the internal impedance tends to stabilize. During the resting beat, the beat duration is adjusted according to the capacity recovery rate and energy recovery curve. When the capacity recovery rate is too slow, the resting time is extended; when the temperature drops too quickly, the resting period is shortened. Through this alternating regulation method, charging, discharging and resting form a coherent energy cycle in the time dimension, so that the energy transfer process in the running trajectory bundle remains continuous and balanced, preventing the accumulation of pseudo-stability caused by excessive continuity of a single process.

[0045] During the alternating regulation process, an energy-releasing ridge structure is introduced to smooth the energy shocks and thermal fluctuations during rhythm transitions, thereby achieving a stable energy transition during continuous operation. Specifically, energy-releasing ridges are set at the junctions of each layer of the shadow beat dome. Through the coordinated work of a time-releasing layer, a temperature-releasing layer, and a capacity recovery layer, energy is gradually released at the conversion nodes. The time-releasing layer controls the beat switching rate; when a charging beat ends and a discharging beat begins, the time-releasing layer extends the transition time between the two, preventing sudden reversals of energy flow that could cause abrupt changes in internal temperature. The temperature-releasing layer reduces the heat release per unit time by controlling the heat transfer rate of the heat dissipation path during beat switching, thus stabilizing temperature changes. The capacity recovery layer promotes the redistribution of active ions by adding a small voltage excitation during the resting phase, thereby reducing electrode polarization after discharge. The three-layered release structure works together to transform the energy conversion process from abrupt to gradual, allowing for the simultaneous release of internal thermal and electrochemical stresses. Under the continuous effect of the energy-releasing ridge, the battery no longer forms a short-term equilibrium illusion during the rhythm switching process, and the performance degradation process hidden in apparent stability can be continuously exposed.

[0046] Under the combined action of the shadow beat dome and the energy release ridge, pseudo-steady states are gradually stripped away during dynamic regulation, generating assessment results that truly reflect the battery's health evolution. Specifically, as the dome cycle continues to run, the pseudo-steady segments identified in the pseudo-steady label frame gradually enter the rhythm regulation range. When a pseudo-steady segment coincides with the energy input or release node of the dome beat, the system enhances beat sensitivity and extends the energy release duration, fully revealing hidden characteristic shifts. During this process, slight upward fluctuations in the voltage curve, delayed decay of the discharge current, slight hysteresis changes in temperature, secondary inflection points in the capacity recovery curve, and subtle increases in internal impedance are all recorded in real time. By continuously recording these real change characteristics and mapping them back to the time series of the operating trajectory bundle, a new trajectory update segment is formed. The updated operating trajectory bundle is continuous in time and reflects real change trends in state. Based on this trajectory, key indicators such as the rate of change of energy conversion efficiency, capacity degradation rate, and temperature response sensitivity of the battery at different rhythm stages can be calculated. Ultimately, through comprehensive analysis of these continuous indicators, an assessment result reflecting the health evolution status of the lithium battery is obtained. This assessment result can reveal the true degradation path and potential instability trends of the battery's internal performance, providing a reliable quantitative basis for health prediction and lifespan management.

[0047] This invention introduces the construction of operating condition rhythm profiles and rhythm fingerprints, enabling rhythm alignment and dynamic expansion of the time axis of operational data across multiple feature dimensions. This effectively identifies feature cancellation phenomena caused by changes in external operating condition rhythms, thereby avoiding misjudgments of staged spurious stability in health assessments. The method establishes a time-consistent benchmark at the feature level, allowing operational features such as capacity, internal resistance, voltage, and temperature to be correlated under a unified rhythm, achieving continuous capture of the true performance change trajectory and significantly improving the consistency between health status assessment results and the actual battery operating state.

[0048] This invention utilizes the coordinated regulation of a dynamic threshold chain and a shadow beat dome to enable an adaptive feedback mechanism in the time, temperature, and energy dimensions of the health assessment process. This mechanism continuously strips away pseudo-steady states during rhythm switching and energy release, ensuring that subtle degradation characteristics exhibited by the battery during dynamic operation are gradually restored, thereby achieving continuous tracking of health evolution trends. This method transforms battery health assessment from static judgment to dynamic evolution identification, possessing higher sensitivity and long-term predictive reliability.

[0049] This invention provides, for example Figure 2 The data analysis-based lithium battery health status assessment system shown includes a rhythm modeling module, a feature alignment module, a pseudo-stability identification module, a weight adjustment module, a trajectory reconstruction module, and a health assessment module. The rhythm modeling module extracts historical data from lithium battery operation logs, constructs a working condition rhythm profile, identifies the inflection point of the operating rhythm cycle, and generates a rhythm fingerprint as a time alignment benchmark. The feature alignment module uses rhythm fingerprint strips to rearrange the phases of the time axes of multiple running features, so that the running features can be synchronously unfolded under a unified rhythm benchmark, locate short-term synchronous stable sections and generate consistent crack lines. The pseudo-stability identification module tracks the difference in operational feature offsets of the consistent crack cable within a continuous time window, extracts feature cancellation information caused by the working condition rhythm, forms a list of cancellation traces, and generates pseudo-stability annotation frames. The weight adjustment module applies rhythmic adaptive weight adjustment to the running feature set based on pseudo-stable labeled frames, increases sensitivity and adjusts thresholds at rhythm switching nodes, forming a dynamic threshold chain. The trajectory reconstruction module performs reverse-phase traction sampling on the original running sequence according to the dynamic threshold chain, and densifies the information collection at rhythm abrupt changes and pulse turning points to reconstruct the running trajectory bundle that reflects the true performance evolution. The health assessment module, through the operation of the trajectory beam-driven timing and temperature coordinated control module, adjusts the charge / discharge rate and resting rhythm, and enables the energy slow-release structure to eliminate pseudo-steady states and output true health assessment results.

[0050] The lithium battery health status assessment method based on data analysis provided in this embodiment of the invention is implemented through the aforementioned lithium battery health status assessment system based on data analysis. For details of the specific methods and processes of the lithium battery health status assessment system based on data analysis, please refer to the embodiments of the lithium battery health status assessment method based on data analysis described above, which will not be repeated here.

[0051] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A data analysis-based method for assessing the health status of lithium batteries, characterized in that, Includes the following steps: Historical data is extracted from lithium battery operation logs to construct a working condition rhythm profile, identify the inflection point of the operating rhythm cycle, and generate a rhythm fingerprint as a time alignment benchmark. The time axis of multiple operational features is rearranged using rhythm fingerprint strips, so that the operational features are synchronously unfolded under a unified rhythm benchmark, the short-term synchronous stable section is located and a consistent crack line is generated. Based on the difference in the offset of the consistent crack cable during the continuous time window, feature cancellation information caused by the working condition rhythm is extracted, a list of cancellation traces is formed and pseudo-stable annotation frames are generated. Based on pseudo-stable labeled frames, rhythmic adaptive weight adjustment is applied to the running feature set. Sensitivity is increased and thresholds are adjusted at rhythm switching nodes to form a dynamic threshold chain. The original running sequence is sampled in reverse phase according to the dynamic threshold chain, and information is collected in a denser manner at rhythm abrupt changes and pulse turning points to reconstruct the running trajectory bundle that reflects the true performance evolution. By running the trajectory beam-driven timing and temperature coordinated control module, the charge / discharge rate and resting rhythm are adjusted and the energy slow-release structure is enabled to remove the pseudo-steady state and output the true health assessment results.

2. The lithium battery health status assessment method based on data analysis according to claim 1, characterized in that, The steps for constructing the operating condition rhythm profile and forming the rhythm fingerprint are as follows: Historical data is extracted from the lithium battery operation logs and rearranged in chronological order to form continuous time data; Based on continuous time data, a working condition rhythm profile is constructed. By superimposing and analyzing the changes of voltage, current, temperature, capacity, energy efficiency and internal impedance over time, the temporal distribution of the charging stage, discharging stage and resting stage is determined. Based on the operating condition rhythm profile, identify the turning points of changes in voltage curves, temperature curves, capacity curves, and internal impedance curves, and extract the periodic inflection point information of the operating rhythm. The inflection point information of the cycle is continuously woven in chronological order to form a rhythm fingerprint strip, which is used to establish a unified rhythm benchmark for subsequent multi-feature time alignment.

3. The lithium battery health status assessment method based on data analysis according to claim 2, characterized in that, The steps for performing phase rearrangement and generating consistent crack strands on the time axis of multiple operational features using rhythmic fingerprinting include: Based on the periodic inflection points of the rhythm fingerprint, the time axes of the voltage change sequence, discharge current curve, temperature change record, capacity retention curve, energy conversion efficiency curve, and internal impedance change curve are repositioned to align all features in time. Under a unified rhythmic benchmark, all operational features are synchronously unfolded in cyclical order, so that each feature is arranged in parallel within the same rhythmic cycle. In a synchronously unfolded multi-feature time structure, identify time segments in which multiple operational features simultaneously exhibit stable trends, and record the rate of change and duration of each feature. The identified synchronous stable segments are connected in chronological order to form a consistent crack line that reflects the cooperative behavior of multiple features.

4. The lithium battery health status assessment method based on data analysis according to claim 3, characterized in that, The steps for identifying pseudo-stable segments and generating pseudo-stable labeled frames within a continuous time window based on consistent crack lines are as follows: Based on the time nodes of the consistent crack, a continuous time window covering multiple rhythmic cycles is established, so that the voltage change curve, discharge current change trajectory, temperature change record, capacity retention curve, energy efficiency change record and internal impedance change curve are continuously arranged in the time dimension. Within a continuous time window, the changing trends of multiple operational features are tracked, and the direction, rate, and duration of change are recorded point by point to form a feature offset trajectory. Based on the feature offset trajectory, identify feature pairs with opposite directions of change, corresponding amplitudes of change, and consistent duration within the same rhythm segment, and record the feature names, time start and time end points, direction of change, amplitude of change, and rhythm segment number to form an offset trace list; The list of offset traces is mapped to the time series of consistent cracks, and time intervals with stable surfaces but internal feature offsets are selected to generate pseudo-stable labeled frames.

5. The lithium battery health status assessment method based on data analysis according to claim 4, characterized in that, When generating pseudo-stable annotation frames, overlapping intervals in the offset trace list that have continuous time distribution and involve multiple operational features are identified as pseudo-stable segments. A time-series annotation structure is established based on the start and end times, participating feature names, change amplitudes, durations, and rhythm segment numbers of the pseudo-stable segments.

6. The lithium battery health status assessment method based on data analysis according to claim 4, characterized in that, The steps for constructing a dynamic threshold chain around pseudo-stable labeled frames are as follows: Using the pseudo-stable label frame as a reference, the time distribution and weight parameters of voltage change curve, discharge current curve, temperature change record, capacity change trend curve, energy conversion efficiency change curve and internal impedance change record are initialized so that the operating characteristics correspond synchronously with the pseudo-stable state on the time axis. Based on the periodic nodes of the rhythm fingerprint, a weighted breathing gate that changes with the rhythm is introduced, so that the weights are periodically adjusted during the charging, discharging and resting phases for different operating characteristics. Adjust the rate of change of the weighted ventricular plexus at the rhythm switching node to enhance the sensitivity to characteristic changes and reduce inertial compliance; A dynamic threshold chain is established based on the trend of weight changes and the evolution of rhythm over time, so that the threshold nodes record the rhythm stage, weight value, critical level and response direction.

7. The lithium battery health status assessment method based on data analysis according to claim 6, characterized in that, The steps for performing inverse traction sampling and reconstructing the trajectory bundle based on the dynamic threshold chain for the original running sequence are as follows: Using a dynamic threshold chain as the basis for time control, the original running data is divided into time zones and rhythms are located to establish a sampling framework with time continuity. Reverse-phase traction sampling is performed with the rhythm change position and pulse inflection position as the sampling center, and key data are collected in a dense manner during the stages of voltage, current, temperature, capacity, energy efficiency and internal impedance changes. The encrypted data is spliced ​​and sequentially arranged with the original time series to achieve continuous connection between rhythmic abrupt changes and stable periods in the time dimension. The spliced ​​multidimensional time series are arranged continuously according to the rhythmic cycle to form a running trajectory bundle.

8. The lithium battery health status assessment method based on data analysis according to claim 7, characterized in that, In the process of encrypted acquisition of reverse-phase traction sampling, the sampling range is extended bidirectionally between the end of the rhythm stable region and the starting point of the next rhythm, with the rhythm change center as the axis of symmetry. The sampling density is increased at voltage inflection points, current drop points, temperature reversal points, capacity decay starting points, energy efficiency fluctuation peak points, and internal impedance change starting points.

9. The lithium battery health status assessment method based on data analysis according to claim 7, characterized in that, The steps for driving the shadow beat dome by running the trajectory beam and achieving pseudo-steady state stripping are as follows: Based on the time and temperature correspondence information in the running trajectory bundle, a time-series and temperature coordinated response framework is established to enable the electrical parameters and thermal properties to form a dynamic coupling relationship. Based on the rhythmic periodic characteristics of the running trajectory bundle, a shadow beat dome structure is constructed, and the charging active phase, the discharge release phase, the static buffer phase and the temperature recovery phase are mapped as dome control areas. The charging rate, discharging rate and resting rhythm are alternately regulated by the shadow beat dome, so that energy input, release and recovery form a continuous cycle; An energy-releasing ridge structure consisting of a time-releasing layer, a temperature-releasing layer, and a capacity recovery layer is introduced at the beat transition point to smooth the energy conversion process. By utilizing the synergistic effect of the shadow beat dome and the energy-releasing ridge, pseudo-stable states are stripped away, resulting in an assessment of the battery's health evolution.

10. A data analysis-based lithium battery health status assessment system, used to implement the data analysis-based lithium battery health status assessment method according to any one of claims 1-9, characterized in that, It includes a rhythm modeling module, a feature alignment module, a pseudo-stability identification module, a weight adjustment module, a trajectory reconstruction module, and a health assessment module. The rhythm modeling module extracts historical data from lithium battery operation logs, constructs a working condition rhythm profile, identifies the inflection point of the operating rhythm cycle, and generates a rhythm fingerprint as a time alignment benchmark. The feature alignment module uses rhythm fingerprint strips to rearrange the phases of the time axes of multiple running features, so that the running features can be synchronously unfolded under a unified rhythm benchmark, locate short-term synchronous stable sections and generate consistent crack lines. The pseudo-stability identification module tracks the difference in operational feature offsets of the consistent crack cable within a continuous time window, extracts feature cancellation information caused by the working condition rhythm, forms a list of cancellation traces, and generates pseudo-stability annotation frames. The weight adjustment module applies rhythmic adaptive weight adjustment to the running feature set based on pseudo-stable labeled frames, increases sensitivity and adjusts thresholds at rhythm switching nodes, forming a dynamic threshold chain. The trajectory reconstruction module performs reverse-phase traction sampling on the original running sequence according to the dynamic threshold chain, and densifies the information collection at rhythm abrupt changes and pulse turning points to reconstruct the running trajectory bundle that reflects the true performance evolution. The health assessment module, through the operation of the trajectory beam-driven timing and temperature coordinated control module, adjusts the charge / discharge rate and resting rhythm, and enables the energy slow-release structure to eliminate pseudo-steady states and output true health assessment results.

Images