Battery state evaluation method and system based on multi-dimensional data fusion and mixed features

Abstract

The invention relates to a battery state evaluation method and system based on multi-dimensional data fusion and mixed features, and the method comprises the following steps: data access and preprocessing: accessing multi-dimensional time sequence data through a battery management system, and preprocessing the data; feature extraction and analysis: extracting composite impedance, voltage difference and balance features based on the data of the battery, extracting composite impedance features through a corresponding algorithm, and calculating the effective capacity of the battery; and health assessment and judgment: comprehensively analyzing a plurality of assessment indexes such as composite impedance, effective capacity and balance according to a set health level threshold model, judging the health state of the battery in real time, and outputting a health level report. According to the method, the accuracy and the real-time performance of battery health assessment are improved, the remaining service life of the battery can be effectively predicted, and potential fault risks can be found in time. And through comprehensive analysis of a plurality of indexes, more comprehensive decision support is provided for the battery management system.

Description

Technical Field

[0001] The present invention relates to the technical field of battery management and status monitoring, and in particular to a battery status assessment method and system based on multi-dimensional data fusion and hybrid features. Background Art

[0002] Rechargeable batteries, particularly lithium-ion batteries, have become core energy components in electric vehicles, energy storage systems, and consumer electronics. Their safety, lifespan, and performance directly impact the reliability and economic viability of the entire system. One of the core responsibilities of a battery management system (BMS) is to accurately assess battery status, including state of charge (SOC), state of health (SOH), and power capability (SOP), and to provide early warning of potential risks.

[0003] Existing battery status assessment methods have different focuses: such as the voltage-based OCV method (offline, insensitive to voltage platforms), the current integration-based ampere-hour method (online, with cumulative errors), the equivalent circuit model-based filtering method (online, dependent on model accuracy and parameter identification), the data-driven machine learning method (great potential, data-dependent, weak interpretability), and the impedance-based method (EIS offline, DCIR online but greatly affected by working conditions).

[0004] However, existing technologies still have shortcomings in comprehensive evaluation:

[0005] 1) Limited evaluation dimensions: often focusing on a single indicator, failing to fully reflect the various aspects of the battery status (such as power and consistency).

[0006] 2) Insufficient data utilization: The multi-dimensional time series data collected by BMS contains rich information, but is often simplified.

[0007] 3) Insufficient capture of dynamic characteristics: It is difficult to accurately reflect the dynamic response of the battery under complex actual working conditions.

[0008] 4) Difficulty in handling the impact of operating conditions: The battery state is affected by the coupling of temperature, SOC, current, etc., and it is difficult to extract intrinsic state indicators.

[0009] 5) Single feature fusion strategy: How to effectively integrate multi-source and heterogeneous features to obtain reliable comprehensive judgments still requires in-depth research.

[0010] 6) Lack of sensor calibration: The evaluation results depend on the data quality, but the sensor validity is often ignored. Summary of the Invention

[0011] The purpose of the present invention is to provide a battery status assessment method and system based on multi-dimensional data fusion and hybrid features to solve the technical problems existing in existing battery management systems when assessing battery health status.

[0012] To achieve the above object, the present invention adopts the following technical solutions:

[0013] A battery state assessment method based on multi-dimensional data fusion and hybrid features includes the following steps: S1: data access and preprocessing: accessing multi-dimensional time series data, including voltage, current, and temperature, through a battery management system and preprocessing the data;

[0014] S2: Feature extraction and analysis: Based on the above battery data, extract the composite impedance, voltage difference, and balance characteristics. Use the corresponding algorithm to extract the composite impedance characteristics and calculate the effective capacity of the battery.

[0015] S3: Health assessment and judgment: Based on the set health level threshold model, comprehensive analysis of multiple evaluation indicators such as composite impedance, effective capacity, and balance is performed to determine the health status of the battery in real time and output a health level report.

[0016] Further preferably, after step S3, report generation and early warning are also included: a structured battery status report is generated based on the health assessment results, and a battery remaining service life prediction and maintenance recommendations are provided. When the battery is abnormal, the system will issue an early warning.

[0017] Further preferably, in step S2, the composite impedance is calculated using three methods: integration method, peak method, and area method, and the impedance values obtained by the three methods are integrated to form a final composite impedance characteristic.

[0018] Further preferably, in step S2, the operating state of the battery is divided into several temperature intervals and voltage intervals, and the voltage difference value is calculated in each interval.

[0019] Further preferably, in step S2, the capacity of the battery is estimated according to the voltage change of the battery in the constant current section.

[0020] Further preferably, in step S2, the voltage dynamic performance factor is also extracted, and the dynamic performance factor calculation formula is:

[0021]

[0022] Where n is the total number of sampling points of the data, ΔV t is the rate of change of voltage calculated at each sampling point.

[0023] More preferably, before step S3, the method further includes performing validity verification on data of the battery's temperature sensor and single cell voltage sensor.

[0024] Further preferably, the validity check includes data consistency check, temperature drift detection, reasonable range check and single cell voltage consistency check.

[0025] Further preferably, in step S3, a threshold model for each health level is established based on a large amount of battery historical data and experimental data through regression analysis or machine learning methods. The threshold model will be dynamically adjusted according to different battery usage conditions and working environments to ensure that the evaluation results have high accuracy and adaptability.

[0026] A battery status assessment system based on multi-dimensional data fusion and hybrid features, including a data interface and processing module, accesses multi-dimensional time series data, including voltage, current, and temperature, and pre-processes the data;

[0027] Feature calculation engine module, used to calculate composite impedance, pressure difference, and balance characteristics;

[0028] Configuration management module, used to store thresholds, parameters, and range definitions;

[0029] The result synthesis and output module is used to output the battery health level report.

[0030] Beneficial effects of the present invention:

[0031] This paper proposes a comprehensive assessment method based on multidimensional time series data fusion, hybrid feature engineering, and multi-stage threshold determination. By comprehensively considering the battery's actual operating time series data, such as current, voltage, and temperature, and employing fused feature extraction and multi-level model determination, this method performs a comprehensive assessment across multiple dimensions of battery operation. This method provides a more accurate and comprehensive prediction of battery health status and effectively avoids misjudgments caused by local feature failure. It also effectively predicts the remaining useful life of the battery and promptly identifies potential failure risks. This method can significantly extend the battery's service life and improve system safety and reliability.

[0032] The battery status assessment method of the present invention can be widely used in electric vehicles, energy storage battery systems and various types of battery management systems, and has strong applicability and scalability. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is a flow chart of battery health status assessment of the present invention;

[0034] Figure 2 It is a flow chart of battery data processing;

[0035] Figure 3 This is the current and balance feature extraction flow chart. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present invention.

[0037] Example:

[0038] like Figure 1-Figure 3 As shown in the figure, the battery status assessment method based on multi-dimensional data fusion and hybrid features includes the following steps:

[0039] Step S1: multi-dimensional data access and preprocessing;

[0040] The voltage, current, temperature and other data from the battery system are uniformly accessed, and pre-processing operations such as data cleaning, denoising, and interpolation are performed to ensure the stability and accuracy of subsequent analysis.

[0041] During battery health assessment, accurate data access and preprocessing are key to ensuring the accuracy of subsequent analysis. This step aims to centrally access multidimensional data from different battery management systems (BMSs), ensuring data integrity and consistency and providing reliable input for subsequent feature extraction.

[0042] The system receives multi-dimensional data from the battery pack, including battery system voltage, current, temperature, and single cell voltage. This data typically has different sampling frequencies and timestamps. To ensure synchronization of subsequent processing, all data must be aligned according to timestamps. If some data points are missing, interpolation is used to fill in the gaps. For example:

[0043]

[0044] This method ensures that all data (such as voltage, current, temperature, etc.) at each moment can be synchronized with other data.

[0045] Step S2: Complex impedance feature extraction

[0046] Composite feature extraction of battery impedance based on current variation segments reflects changes in the battery's internal electrochemical state from different perspectives. These features can help assess the battery's load response capability and its health degradation trend.

[0047] By analyzing the current changes during the battery's charge and discharge processes, we can extract a composite impedance signature that reflects the battery's internal electrochemical state. This impedance signature reveals the battery's responsiveness under different load conditions and helps assess its health and degradation trends.

[0048] During the battery's charge and discharge process, current fluctuations significantly affect the battery's internal impedance. To understand this, we first need to segment the current data, dividing the battery's charge and discharge process into multiple consecutive segments based on significant current fluctuations. Within each current segment, we calculate the relationship between voltage and current to determine the corresponding impedance value.

[0049] For each current change, the system calculates its composite impedance R t, which reflects the load response capability of the battery during this time period. The following three methods are used to calculate the composite impedance:

[0050] Integration method: Calculate the average impedance of the battery by the ratio of the accumulated voltage to the current change:

[0051]

[0052] Peak method: Calculate the instantaneous impedance by the peak difference between voltage and current within the current segment:

[0053]

[0054] Area method: Estimate impedance by the area ratio of the voltage and current traces:

[0055]

[0056] The impedance values R obtained by the above three calculation methods are (A) ,R (B) , and R (C) Fusion is performed to form the final composite impedance characteristic R CDR .

[0057] The value of the composite impedance is closely related to the health of the battery:

[0058] Healthy state: A lower composite impedance value indicates that the internal electrochemical reaction of the battery is good, the battery has a strong load response capability, and the battery is in a healthy state;

[0059] Mild degradation: As the battery is used for a longer time, the internal impedance gradually increases. At this time, the composite impedance value will rise slightly, indicating that the battery performance has slightly declined;

[0060] Severe degradation: If the composite impedance value of the battery increases significantly, it indicates that serious damage has occurred inside the battery or the battery pack is poorly balanced. The battery may be in a state of severe degradation or failure.

[0061] Step S3: Pressure difference and temperature zone mapping model

[0062] By analyzing the pressure difference data in different temperature ranges and voltage ranges, a temperature zone mapping model is constructed to reveal the internal consistency and failure trends of the battery under different operating conditions.

[0063] The voltage differential reflects the voltage difference between battery cells. Generally speaking, the better the battery balance, the smaller the voltage differential. Therefore, by monitoring the changes in the voltage differential in real time, it is possible to effectively assess the consistency of the battery and whether there is a potential failure risk.

[0064] In practical battery applications, changes in temperature and voltage have a significant impact on battery performance. To this end, this step divides the battery's operating state into several temperature and voltage ranges and calculates the voltage differential within each range. Specifically, the system segments the battery's voltage data based on the temperature variations of the battery's operating environment, categorizing the battery's state at each point in time into temperature T and voltage V ranges.

[0065] For each pair of temperature intervals T i and voltage range V j , the system will calculate the voltage difference within this range Defined as the difference between the maximum and minimum voltages within the interval:

[0066]

[0067] in, Indicates that at temperature T i and voltage V j Voltage dataset within the interval.

[0068] Step S4: Consistency analysis and calculation of composite equilibrium index

[0069] Calculate the fluctuations and inconsistencies between battery cell voltages, evaluate the balance of the battery pack through consistency indicators, and provide optimization directions for the battery management system.

[0070] The balance of a battery pack directly impacts its performance and service life. During use, individual cells may experience variations in internal resistance, capacity, and charge / discharge characteristics. These variations can lead to uneven energy distribution in the pack, impacting overall battery performance. Therefore, assessing battery consistency and balance is a critical step in this invention.

[0071] Consistency analysis is mainly evaluated by calculating the voltage fluctuation difference of each single cell in the battery pack. Assume that there are n single cells in the battery pack, and the voltage of the single cell at any time t is v_t,i, where i represents the number of the single cell. The system calculates the average voltage of the battery pack at each time The consistency is reflected by the voltage fluctuation range ΔV_t.

[0072] Voltage mean: For n cells in a battery pack, calculate the mean voltage:

[0073]

[0074] Voltage fluctuation range: Calculate the voltage range of the battery pack at that moment (the difference between the maximum voltage and the minimum voltage):

[0075] ΔV t =max({vt,1 ,v t,2 ,\dots,v t,n})-min({v t,1 ,v t,2 ,\dots,v t,n})

[0076] The larger the voltage fluctuation range ΔV_t, the greater the difference between the individual cells in the battery pack, and the poorer the balance. Conversely, if the voltage fluctuation is smaller, it means that the battery pack is more balanced.

[0077] To quantify the battery pack's balance, this step further calculates a composite balancing index. This index comprehensively considers the voltage fluctuation, mean deviation, and consistency of each cell in the battery pack to assess whether the battery pack requires balancing.

[0078] The calculation of the composite equilibrium index takes into account the following factors:

[0079] (1) Voltage deviation: The deviation between each single cell and the average voltage of the battery pack:

[0080]

[0081] If the voltage deviation of a single battery is large, it means that the battery may have a large capacity difference or internal resistance change, and may need to be replaced or balanced.

[0082] (2) Weighted balancing score: The balancing score of the battery pack is obtained by calculating the weighted average of the voltage deviations of all single cells:

[0083]

[0084] The lower the score, the better the battery pack is balanced. If the composite balance score exceeds the set threshold, it indicates that the battery pack is unbalanced and may need to be balanced.

[0085] Step S5: Approximate estimation of effective capacity

[0086] Based on the voltage change during the constant current charge and discharge phase, the effective capacity of the battery is approximately calculated to evaluate the actual energy storage capacity of the battery.

[0087] Effective capacity is a core metric in battery performance evaluation, typically used to indicate a battery's energy storage capacity in real-world applications. Because the battery's charge and discharge processes are affected by factors such as temperature and current fluctuations, its actual capacity may differ from its nominal capacity. Therefore, the goal of this step is to approximate the battery's constant current phase and calculate its effective capacity, using this metric to reflect the battery's health.

[0088] During the battery charging and discharging process, the current change may not be constant, and the current fluctuation will affect the battery charging and discharging efficiency. Therefore, in order to accurately evaluate the effective capacity of the battery, it is first necessary to use the constant current segment identification algorithm to extract the constant current segment with small current change and relatively stable from the battery charging and discharging curve. Specifically, the system calculates the current change rate ΔI at each moment t =|I t -I t-1 |, when the current change rate is less than the set threshold δ I When , this period of time is considered to be the constant current period.

[0089] For each constant current segment, the system estimates the capacity of the battery based on the voltage change within the segment. The approximate effective capacity of the battery can be calculated using the following formula:

[0090]

[0091] Here, ΔV is the voltage change within the constant current range, and I is the constant current value. This formula shows that the effective capacity of the battery is proportional to the voltage change within the current range. The more stable the current, the more accurate the effective capacity estimation result of the battery.

[0092] Step S6: Extracting voltage dynamic performance factors

[0093] The voltage response performance factor of the battery during the charge and discharge process is extracted to reflect the dynamic response capability of the battery and is used to evaluate the performance of the battery under different loads.

[0094] During the charge and discharge process, a battery's voltage response speed and stability are important factors in evaluating battery performance. Especially under rapid charge and discharge or high-load conditions, the battery's voltage change rate (or dynamic performance) has a direct impact on its safety, efficiency, and service life. The goal of this step is to evaluate the battery's response capabilities under different loads by extracting the voltage dynamic performance factor, providing supplementary data for a comprehensive assessment of battery health.

[0095] The calculation of the voltage dynamic performance factor is based on the voltage change during the battery charging and discharging process. By calculating the rate of change of voltage at each time point, the system can measure the response speed of the battery under load changes. Voltage change rate ΔV t It can be calculated by the following formula:

[0096] ΔV t =|V t -V t-1 |

[0097] Among them, V t is the voltage at the current moment, V t-1 is the voltage at the previous moment. Voltage change rate ΔP tIt reflects the battery's response speed to changes in external load. A higher rate of change usually means that the battery is more sensitive under high load conditions.

[0098] Dynamic performance factor R of the battery dyn It is a comprehensive indicator to measure the battery voltage response speed, usually calculated based on the voltage change rate. Specifically, the system reflects the battery's response performance by counting the average and maximum values of the battery's voltage change rate during the charge and discharge process. The dynamic performance factor calculation formula is:

[0099]

[0100] Where n is the total number of sampling points of the data, ΔV t is the rate of change of voltage calculated at each sampling point. dyn A higher value means the battery responds faster to load changes, while a lower value means the battery responds slower.

[0101] Voltage dynamic performance factor P dyn This is a key indicator of battery performance, and its value changes directly reflect the battery's internal state and health. As the battery ages and uses, its internal resistance increases, slowing its response to load changes. Therefore, the dynamic performance factor tends to decline as the battery ages.

[0102] Healthy battery: The battery can respond quickly to load changes and has a high dynamic performance factor;

[0103] Degraded battery: The battery's response speed slows down and the dynamic performance factor is low.

[0104] By monitoring the dynamic performance factors of the battery in real time, the system can promptly detect the performance degradation of the battery under high load conditions, help predict the remaining battery life and take measures in advance.

[0105] Dynamic performance factor P dyn As a key indicator of battery health, it provides real-time health assessments for the battery management system. By combining this with other health indicators (such as effective capacity and composite impedance), the system can comprehensively assess the battery's operating status and issue early warnings for maintenance or replacement when necessary.

[0106] Step S7: Sensor validity check

[0107] Perform validity verification on sensor data such as battery temperature and cell voltage to ensure the authenticity and reliability of the collected data and eliminate possible sensor failures or data errors.

[0108] Temperature sensors and cell voltage sensors in the battery management system can distort data due to failure, drift, or external environmental interference. Therefore, the goal of this step is to ensure the authenticity and validity of collected sensor data through a validity verification mechanism, thereby providing reliable data support for subsequent health assessments.

[0109] Temperature has a significant impact on battery performance. Batteries in high or low temperature environments can behave significantly differently than they do in normal temperature conditions. To ensure the validity of temperature data, this step first checks the rationality of sensor data using the following strategies:

[0110] Data consistency check: First, the system checks the consistency of the readings from each temperature sensor. If there are multiple temperature sensors, the system compares the readings from the different sensors. If the reading from a particular sensor deviates significantly from the readings from the other sensors (for example, outside the set tolerance range), the sensor data is considered to be problematic.

[0111] Temperature drift detection: Temperature sensors may drift, causing data distortion. To verify sensor validity, the system analyzes temperature fluctuation patterns based on historical data. If it detects rapid temperature changes or large fluctuations within a short period of time, the sensor is marked as invalid.

[0112] The battery cell voltage is an important data source for evaluating the battery health status. To ensure the accuracy of the voltage data, this step uses the following method to perform validity verification:

[0113] Reasonable range check: The battery cell voltage value should be within a reasonable range (for example, the battery voltage is usually between 2.5V and 4.3V). If a battery cell voltage is outside this range, the sensor is considered to be faulty or abnormal.

[0114] Cell voltage consistency check: All cells in a battery pack should have similar voltage values. If the voltage of a cell differs significantly from the others and does not conform to the normal operation of the battery pack, the system will detect it through an algorithm and mark the sensor as invalid.

[0115] Step S8: Health level determination and threshold model

[0116] Based on the multiple characteristic indicators extracted above, the battery status is graded through the health level model, and whether the battery needs to be repaired or replaced is determined based on the threshold.

[0117] Specifically, it includes data collection: obtaining the current health data of the battery from the battery management system, including composite impedance, effective capacity, balance and other indicators; indicator calculation: calculating various health assessment indicators based on the collected data; level judgment: comparing various indicators with the threshold model to determine the health level of the battery; output results: generating a battery health status report and providing maintenance or replacement recommendations.

[0118] By comprehensively considering characteristic data from multiple dimensions, including the battery's composite impedance, effective capacity, balance indicators, temperature, and voltage fluctuations, a health grade determination model is constructed to dynamically assess the battery's health status. This model, based on threshold judgments, can grade the battery's health status and provide data support for subsequent maintenance decisions.

[0119] Different health levels are defined to reflect the working status and service life of the battery. Each health level is determined by multiple key indicators, including:

[0120] Composite impedance R CDR : Reflects the battery's response capability under load;

[0121] Effective capacity C eff : Actual energy storage capacity of the battery;

[0122] Composite Balance Index C bal : Consistency of battery cells;

[0123] Voltage dynamic performance factor P dyn : The battery's response speed to load changes;

[0124] Temperature fluctuation and voltage fluctuation: reflect the working stability of the battery under extreme conditions.

[0125] These indicators will be judged according to the set health thresholds and ultimately output the battery health level. The health level is generally divided into the following levels:

[0126] Health status: All indicators are within normal range and battery performance is good;

[0127] Mild degradation: Some indicators slightly deviate from the normal range, and battery performance slightly decreases;

[0128] Moderate degradation: Multiple indicators deviate significantly from the normal range, and battery performance decreases significantly;

[0129] Severe degradation: All key indicators are seriously deviated from the normal range, the battery performance is seriously degraded, and replacement may be required.

[0130] To determine the battery health level, this step uses regression analysis or machine learning to establish threshold models for each health level based on a large amount of historical and experimental battery data. These threshold models are dynamically adjusted based on the battery's varying usage conditions and operating environments, ensuring high accuracy and adaptability of the assessment results.

[0131] Each health level corresponds to one or more thresholds, and the specific thresholds are set as follows:

[0132] Healthy status: All indicators such as composite impedance, effective capacity and voltage dynamic performance are within the healthy range. Composite impedance R CDR Low, the battery is well balanced, and the effective capacity is close to the nominal capacity;

[0133] Mild recession: Some indicators such as composite equilibrium C bal Or dynamic performance factor P dyn etc. deviate from the normal range, but are still within the acceptable range;

[0134] Moderate decay: composite impedance R CDR Higher, the effective capacity drops significantly, and the battery performance shows a significant decline;

[0135] Severe degradation: Multiple key indicators deviate seriously from the normal range, the battery's energy storage capacity and load response capabilities are greatly reduced, and there is a risk of failure.

[0136] Determining the battery health level is crucial for battery management systems. By monitoring the battery health level in real time, the system can issue early warnings when battery degradation is occurring, enabling users or maintenance personnel to take timely action to extend the battery life and ensure battery system safety. The health level model not only predicts the remaining battery life but also optimizes battery charge and discharge strategies, reducing overuse and unnecessary battery wear.

[0137] Through the health level determination in this step, the battery management system can achieve a transition from simple status detection to comprehensive health assessment, and provide decision support for efficient battery utilization and safe management.

[0138] Step S9: Comprehensive output report generation

[0139] Finally, a structured battery status report is generated based on the battery health level and performance evaluation results, providing a decision-making basis for battery management.

[0140] The battery status assessment method of the present invention has the following characteristics:

[0141] Multidimensional data fusion: By integrating multidimensional time-series data such as battery voltage, current, and temperature, we can comprehensively consider the dynamic changes of batteries under different operating conditions, avoiding evaluation errors caused by a single indicator. After preprocessing, the time-series data from each input data source is aligned on a unified time axis to ensure the accuracy of subsequent analysis.

[0142] Hybrid feature engineering: This mechanism integrates multiple data processing and feature extraction methods. Specifically, it includes segment-based current change analysis, composite impedance calculation based on voltage response, and temperature-voltage differential feature extraction. These features not only reflect the static characteristics of the battery, but also reveal the battery's dynamic response under actual load.

[0143] Multi-stage model and graded assessment: By setting up a multi-level health assessment model, a preliminary health assessment value is first calculated based on basic battery operating indicators. This is then dynamically adjusted based on actual operating data to ultimately output a refined health grade. These health grades include but are not limited to "healthy," "mildly degraded," "moderately degraded," and "severely degraded," facilitating subsequent maintenance and optimization.

[0144] A battery status assessment system based on multi-dimensional data fusion and hybrid features, including a data interface and processing module, accesses multi-dimensional time series data, including voltage, current, and temperature, and pre-processes the data; used to implement the method of step S1 above;

[0145] A feature calculation engine module, comprising submodules that implement the specific calculation logic in steps S2-S8, for calculating composite impedance, pressure difference, and balance characteristics;

[0146] Configuration management module, used to store thresholds, parameters, range definitions, etc.;

[0147] The result synthesis and output module is used to output the battery health level report, corresponding to step S9 of the above method.

[0148] The present invention is not limited to the above-mentioned optimal implementation mode. Anyone can derive other forms of products under the inspiration of the present invention. However, no matter what changes are made in the shape or structure, any technical solution that is the same or similar to that of the present application falls within the scope of protection of the present invention.

Claims

1. A battery status assessment method based on multi-dimensional data fusion and hybrid features, characterized in that: The following steps are involved: S1: Data access and preprocessing: Access multi-dimensional time series data, including voltage, current, and temperature, through the battery management system and preprocess the data; S2: Feature extraction and analysis: Based on the above battery data, extract the composite impedance, voltage difference, and balance characteristics. Use the corresponding algorithm to extract the composite impedance characteristics and calculate the effective capacity of the battery. S3: Health assessment and judgment: Based on the set health level threshold model, comprehensive analysis of multiple evaluation indicators such as composite impedance, effective capacity, and balance is performed to determine the health status of the battery in real time and output a health level report.

2. The battery status assessment method based on multi-dimensional data fusion and hybrid features according to claim 1 is characterized in that: After step S3, it also includes report generation and early warning: a structured battery status report is generated based on the health assessment results, providing a prediction of the remaining battery life and maintenance recommendations. When the battery is abnormal, the system will issue an early warning.

3. The battery status assessment method based on multi-dimensional data fusion and hybrid features according to claim 1, characterized in that: In step S2, the composite impedance is calculated using the integration method, the peak method, and the area method. The impedance values obtained by the three methods are integrated to form the final composite impedance characteristics.

4. The battery status assessment method based on multi-dimensional data fusion and hybrid features according to claim 1, characterized in that: In step S2, the operating state of the battery is divided into several temperature intervals and voltage intervals, and the voltage difference value is calculated in each interval.

5. The battery status assessment method based on multi-dimensional data fusion and hybrid features according to claim 1 is characterized in that: In step S2, the capacity of the battery is estimated based on the voltage change of the battery in the constant current section.

6. The battery status assessment method based on multi-dimensional data fusion and hybrid features according to claim 1, characterized in that: In step S2, the voltage dynamic performance factor is also extracted. The dynamic performance factor calculation formula is: Where n is the total number of sampling points of the data, ΔV t is the rate of change of voltage calculated at each sampling point.

7. The battery status assessment method based on multi-dimensional data fusion and hybrid features according to claim 1, characterized in that: Before step S3, the validity of the data of the battery temperature sensor and the single cell voltage sensor is also verified.

8. The battery status assessment method based on multi-dimensional data fusion and hybrid features according to claim 7 is characterized in that: Validity verification includes data consistency check, temperature drift detection, reasonable range check and single cell voltage consistency check.

9. The battery status assessment method based on multi-dimensional data fusion and hybrid features according to claim 1, characterized in that: In step S3, based on a large amount of battery historical data and experimental data, a threshold model for each health level is established through regression analysis or machine learning methods. The threshold model will be dynamically adjusted according to the different usage conditions and working environment of the battery to ensure that the evaluation results have high accuracy and adaptability.

10. A battery status assessment system based on multi-dimensional data fusion and hybrid features, characterized by: It includes a data interface and processing module, which accesses multi-dimensional time series data, including voltage, current, and temperature, and pre-processes the data; Feature calculation engine module, used to calculate composite impedance, pressure difference, and balance characteristics; Configuration management module, used to store thresholds, parameters, and range definitions; The result synthesis and output module is used to output the battery health level report.

Images