Storage battery health state on-line evaluation system

Abstract

The invention relates to the technical field of storage batteries, and discloses a storage battery health state on-line evaluation system, which comprises a magnetic type probe and a data acquisition device, which are used for acquiring a voltage sequence, a pole temperature sequence, an internal resistance sequence, a pole micro-vibration signal and an environment temperature and humidity sequence; the data processing module is configured to obtain an internal resistance health index according to the pole temperature sequence; obtaining a thermal health index according to the pole temperature sequence; obtaining a mechanical health index according to the pole micro-vibration signal; obtaining an environmental health index according to the environmental temperature and humidity sequence; obtaining a stability index according to the pole temperature sequence and the voltage sequence; obtaining a system consistency index according to a residual error generated by a dynamic relation model of the voltage sequence, the pole temperature sequence and the internal resistance sequence; and analyzing the indexes by adopting a D-S evidence theory to obtain a health state evaluation result. According to the method provided by the invention, the data acquisition convenience is high, and the evaluation result is more accurate.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and in particular to an online assessment system for the health status of batteries. Background Technology

[0002] In various power storage, transportation power, and backup power systems, batteries serve as the core energy storage unit, and their accurate health status assessment directly impacts the reliability and safety of the overall system operation. Precise, real-time online assessment of battery health status is crucial for enabling predictive battery maintenance, preventing sudden failures, and optimizing system operation.

[0003] In existing technologies, battery health status assessment methods have the following drawbacks: At the data acquisition level, sensors usually require wiring for fixed installation, which is inconvenient to deploy and difficult to apply quickly and flexibly in battery packs, resulting in poor data acquisition convenience; At the data processing level, they often rely on isolated monitoring and threshold judgment of single or a few state parameters such as voltage, internal resistance, or temperature, which often results in a single assessment dimension, making it difficult to comprehensively reflect the overall aging status of the battery under complex operating conditions, and failing to effectively capture early performance degradation and potential faults caused by the coupling of multiple factors, leading to a high misjudgment rate.

[0004] Therefore, improving the convenience of battery status data collection and the accuracy of battery health status assessment have become technical problems that need to be solved by those skilled in the art. Summary of the Invention

[0005] This invention provides an online battery health status assessment system to solve the technical problems of improving the convenience of battery status data collection and the accuracy of battery health status assessment. It enables convenient collection of battery status data and integrates multi-dimensional information such as battery electrical, thermodynamic, mechanical vibration and environmental factors to achieve a more comprehensive, robust and accurate assessment of battery health status.

[0006] To address the aforementioned technical problems, embodiments of the present invention provide an online battery health status assessment system, the system comprising:

[0007] A magnetic probe is used to match the terminals of the battery being evaluated.

[0008] The data acquisition device is fixedly connected to the magnetic probe and is used to acquire the voltage sequence, terminal temperature sequence, internal resistance sequence, terminal micro-vibration signal and ambient temperature and humidity sequence of the battery to be evaluated.

[0009] The data processing module is configured as follows:

[0010] Based on the pole temperature sequence, the internal resistance sequence is temperature compensated to obtain an internal resistance reference sequence. Trend analysis is then performed on the internal resistance reference sequence to obtain an internal resistance health index.

[0011] The thermal health index is obtained based on the electrode temperature sequence.

[0012] The mechanical health index is obtained based on the micro-vibration signal of the pole column;

[0013] The environmental health index is obtained based on the environmental temperature and humidity sequence.

[0014] The stability index is obtained based on the electrode temperature sequence and the voltage sequence.

[0015] The system consistency index is obtained based on the residuals generated by the dynamic relationship model constructed from the voltage sequence, the pole temperature sequence, and the internal resistance sequence.

[0016] The DS evidence theory is used to perform evidence fusion analysis on the internal resistance health index, the thermal health index, the mechanical health index, the environmental health index, the stability index, and the system consistency index to obtain the health status assessment result of the battery to be evaluated.

[0017] Preferably, the magnetic probe integrates a spring probe, a temperature sensor, a rigid metal probe, and a probe metal shell, and the data acquisition device includes a measurement circuit and an accelerometer chip;

[0018] The spring probe forms an electrical connection with the surface of the battery terminal to be evaluated.

[0019] The spring probe is fixedly connected to the measurement circuit, which includes a microcontroller, an analog-to-digital converter, and a precision power resistor controlled by a metal-oxide-semiconductor field-effect transistor to perform voltage sequence acquisition and internal resistance sequence acquisition.

[0020] The temperature sensor is connected to the metal base of the spring probe via a thermal connection structure, so that a low thermal resistance path is formed between the temperature sensing area of ​​the temperature sensor and the metal base.

[0021] One end of the rigid metal probe forms a tight mechanical contact with the surface of the battery terminal to be evaluated, and the other end is rigidly connected to the metal housing of the probe.

[0022] The accelerometer chip is rigidly mounted in the center of the main printed circuit board of the data processing module and is fixedly connected to the metal shell of the probe through a rigid support structure to perform pole micro-vibration signal acquisition.

[0023] Preferably, the spring probe is made of high-carbon spring steel, and the head of the spring probe is plated with thick gold or hard gold.

[0024] The rigid metal probe is made of titanium alloy, and the metal shell of the probe is made of aluminum alloy.

[0025] Preferably, the data acquisition device further includes several temperature and humidity sensors that are communicatively connected to the data processing module;

[0026] The temperature and humidity sensors are distributed and installed at the ventilation holes of the battery compartment of the battery to be evaluated or on the inner wall of the battery compartment.

[0027] Preferably, the data processing module includes:

[0028] The internal resistance health analysis unit is used to perform linear fitting on the internal resistance reference sequence and use the rate of change of the slope of the trend equation obtained by linear fitting as the internal resistance health index.

[0029] The thermal health analysis unit is used to extract multiple thermal characteristic parameters of the pole temperature sequence, score each thermal characteristic parameter, and fuse the corresponding multiple characteristic score results according to a preset weight to obtain a thermal health index.

[0030] The mechanical health analysis unit is used to extract the time-domain and frequency-domain features of the pole micro-vibration signal, obtain the current vibration energy based on the time-domain and frequency-domain features, and obtain the mechanical health index based on the ratio between the preset vibration energy baseline and the current vibration energy.

[0031] The environmental risk analysis unit is used to score the environmental temperature and humidity sequence according to the preset environmental safety working range, and to obtain the environmental health index based on the obtained environmental score results.

[0032] Preferably, the data processing module further includes:

[0033] The stability analysis unit is used to obtain a comprehensive entropy value based on the first entropy value of the pole temperature sequence and the second entropy value of the voltage sequence, and to map the comprehensive entropy value into a stability index.

[0034] The system consistency analysis unit is used to obtain the predicted voltage sequence corresponding to the pole temperature sequence and the internal resistance sequence based on the dynamic relationship model constructed based on the health status operation data, and to obtain the system consistency index based on the variance of the residual sequence between the predicted voltage sequence and the voltage sequence.

[0035] Preferably, the data processing module further includes:

[0036] The health status assessment unit is configured as follows:

[0037] A health status identification framework for the battery to be evaluated is established, and a state membership function is constructed based on the health status identification framework. The state membership function includes: a first state membership function corresponding to the internal resistance health index, a second state membership function corresponding to the thermal health index, a third state membership function corresponding to the mechanical health index, a fourth state membership function corresponding to the environmental health index, a fifth state membership function corresponding to the stability index, and a sixth state membership function corresponding to the system consistency index.

[0038] Based on the state membership function, the state membership distribution corresponding to the battery to be evaluated is obtained. The state membership distribution includes: a first state membership distribution corresponding to the internal resistance health index, a second state membership distribution corresponding to the thermal health index, a third state membership distribution corresponding to the mechanical health index, a fourth state membership distribution corresponding to the environmental health index, a fifth state membership distribution corresponding to the stability index, and a sixth state membership distribution corresponding to the system consistency index.

[0039] The state membership distribution is normalized and uncertainty is assigned sequentially to obtain a basic probability allocation, which includes: a first basic probability allocation corresponding to the internal resistance health index, a second basic probability allocation corresponding to the thermal health index, a third basic probability allocation corresponding to the mechanical health index, a fourth basic probability allocation corresponding to the environmental health index, a fifth basic probability allocation corresponding to the stability index, and a sixth basic probability allocation corresponding to the system consistency index.

[0040] DS evidence fusion is performed on the first basic probability allocation, the second basic probability allocation, the third basic probability allocation, the fourth basic probability allocation, the fifth basic probability allocation, and the sixth basic probability allocation to obtain the fused basic probability allocation. Based on the fused basic probability allocation, the health status assessment result of the battery to be evaluated is determined.

[0041] Preferably, the first, second, third, fourth, fifth, and sixth basic probability allocation functions are all constructed based on trapezoidal membership functions, and the shape parameters of the trapezoidal membership functions are determined by cluster analysis of historical state data.

[0042] Preferably, when performing the DS evidence fusion, the health status assessment unit also simultaneously analyzes the contribution of the basic probability allocation corresponding to each of the internal resistance health index, the thermal health index, the mechanical health index, the environmental health index, the stability index, and the system consistency index to the health status assessment result.

[0043] Preferably, the system further includes a self-calibration module, used to control the data acquisition device to acquire a reference voltage and a reference internal resistance when the battery to be evaluated is idle and without load, and to perform zero-point calibration and gain calibration on the measurement circuit based on the reference voltage and the reference internal resistance.

[0044] The online battery health status assessment system disclosed in this invention has the following advantages compared to existing technologies:

[0045] The online battery health status assessment system disclosed in this application uses a magnetic probe, eliminating the need for specialized installation tools. This results in low installation complexity, fast installation speed, and convenient maintenance. The probe can be easily removed for calibration or replacement. It penetrates the oxide layer, ensuring good contact and high reliability. The DS evidence theory effectively integrates and scientifically decides on multi-dimensional and uncertain battery health information, ultimately outputting a reliable health status assessment result. By explicitly assigning confidence to the entire set, it preserves the inherent uncertainty of each evidence source and handles conflicts through normalization during the fusion process, making the final decision more robust. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the structure of an online battery health status assessment system provided in one embodiment of the present invention; Figure 2 This is a schematic diagram of the external appearance of the magnetic probe and data acquisition device provided in one embodiment of the present invention; Figure 3 This is an installation diagram of the magnetic probe and data acquisition device provided in one embodiment of the present invention; Figure label: Among them, 1-magnetic probe, 2-data acquisition device, 3-data processing module, and 4-battery to be evaluated. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0048] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0049] In a preferred embodiment of the present invention, an online battery health status assessment system is provided, such as... Figure 1 The diagram shown is a structural schematic of an online battery health status assessment system. Figure 2 The diagram shown depicts the external shape of the magnetic probe and data acquisition device. Figure 3The diagram shows the installation of the magnetic probe and data acquisition device. The online battery health status assessment system includes: a magnetic probe 1, matched with the terminals of the battery 4 to be assessed; a data acquisition device 2, fixedly connected to the magnetic probe 1, used to acquire the voltage sequence, terminal temperature sequence, internal resistance sequence, terminal micro-vibration signal, and ambient temperature and humidity sequence of the battery 4 to be assessed; and a data processing module 3, configured to: perform temperature compensation on the internal resistance sequence based on the terminal temperature sequence to obtain an internal resistance reference sequence; perform trend analysis on the internal resistance reference sequence to obtain an internal resistance health index; obtain a thermal health index based on the terminal temperature sequence; obtain a mechanical health index based on the terminal micro-vibration signal; obtain an environmental health index based on the ambient temperature and humidity sequence; and obtain a data processing index based on the terminal temperature sequence and the ambient temperature and humidity sequence. The voltage sequence yields a stability index; the residuals generated by the dynamic relationship model constructed from the voltage sequence, the terminal temperature sequence, and the internal resistance sequence yield a system consistency index; DS evidence theory is used to perform evidence fusion analysis on the internal resistance health index, the thermal health index, the mechanical health index, the environmental health index, the stability index, and the system consistency index to obtain the health status assessment result of the battery 4 to be evaluated; in a preferred embodiment of this application, a spring probe, a temperature sensor, a rigid metal probe, and a probe metal shell are integrated within the magnetic probe 1; the data acquisition device 2 includes a measurement circuit and an accelerometer chip; voltage measurement and internal resistance measurement are implemented by the same hardware circuit, consisting of a spring probe and a measurement circuit, with the spring probe fixedly connected to the measurement circuit. During voltage and internal resistance measurements, a high-strength spring probe penetrates the oxide layer on the surface of the battery terminal to form an electrical connection with the surface of the terminal of the battery 4 to be evaluated. The spring probe is made of high-carbon spring steel with a thick gold or hard gold plating at the tip, meeting the requirements of high conductivity, high elasticity, high hardness, and corrosion resistance. A high-strength permanent magnet is integrated around the spring probe to ensure constant contact pressure under various battery terminal surface conditions, thus achieving stable contact resistance. The housing of the magnetic probe 1 is made of reinforced engineering plastic or die-cast aluminum alloy, meeting the requirements of insulation, strength, temperature resistance, corrosion resistance, and lightweight. For reinforced engineering plastic, a metal lining is required to shield electromagnetic interference; for die-cast aluminum alloy, the surface requires anodized insulation treatment. The measurement circuit includes a microcontroller, an analog-to-digital converter (ADC), and a precision power resistor controlled by a MOSFET. During voltage sequence acquisition, a high-impedance, high-precision ADC directly measures the potential difference between the two probes. Protection and filtering circuits are present at the input front end of the ADC to prevent high-voltage surges.When performing internal resistance sequence acquisition, a DC pulse discharge method is used. A MOSFET controls a precision power resistor as a load, and an MCU (microcontroller) controls the MOSFET to turn on within a very short time (e.g., 3-10 milliseconds), so that the battery discharges a known test current with a small amplitude and a very short duration through the load. At the same time, the analog-to-digital converter measures the instantaneous voltage drop value, and then obtains the internal resistance value according to Ohm's law. The formula for calculating the internal resistance is:

[0050]

[0051] in, This indicates the internal resistance of the battery to be evaluated. This represents the instantaneous voltage drop measured by the analog-to-digital converter during data acquisition. This indicates the test current value.

[0052] For temperature sensors, digital temperature sensor chips, such as the DS18B20 or the higher-precision PT1000, are used in conjunction with analog-to-digital converters. These chips are small in size, high in accuracy, directly output digital signals, and have strong anti-interference capabilities. To achieve accurate, rapid, and hysteresis-free measurement of the temperature of critical parts of the spring probe, the temperature sensor chip must achieve good thermal coupling with the spring probe. The temperature sensor is connected to the metal base of the spring probe through a thermal connection structure, so that a low thermal resistance thermal path is formed between the temperature sensing area of ​​the temperature sensor and the metal base. Specifically, the back of the temperature sensor chip package or its dedicated temperature sensing surface is bonded to the surface of the metal base of the spring probe in one of the following ways: direct close mounting, using a high thermal conductivity adhesive to directly fix the temperature sensor chip to the surface of the metal base; or the temperature sensor chip is integrated into the metal base during the manufacturing stage using a high thermal conductivity insulating material, so that the temperature sensor chip and the metal base are fixed into a rigid integral structure, achieving a near-integral thermal state. The thermal connection structure provided in this application will minimize the contact thermal resistance and spatial thermal resistance between the temperature sensor chip and the metal base surface of the spring probe, ensuring that the two are highly synchronized thermally. This will enable the temperature measured by the temperature sensor chip to accurately and instantly reflect the actual temperature of the spring probe connector, avoiding temperature measurement lag and errors caused by air gaps or poor contact.

[0053] A rigid metal probe, a probe metal housing, and an accelerometer chip constitute a micro-vibration sensor. One end of the rigid metal probe forms a tight mechanical contact with the surface of the four terminals of the battery being evaluated, while the other end is rigidly connected to the probe metal housing. The accelerometer chip is a triaxial microelectromechanical system (MEMS) accelerometer chip, which is extremely small in size and has low power consumption, such as the ADXL345, and can detect minute vibration accelerations. The core function of the triaxial MEMS accelerometer chip is to convert the physical quantities of vibration acceleration in the three mutually perpendicular directions (X, Y, and Z) into digital or analog electrical signals that can be read by the microcontroller. The installation position and mechanical coupling of the triaxial MEMS accelerometer chip are the most challenging parts of this application. The triaxial MEMS accelerometer chip must be rigidly mounted in the center of the main printed circuit board of the data processing module 3 of the battery health status online assessment system. The mechanical structure of the rigid metal probe must be designed as a rigid whole. The transmission path of the vibration signal is: battery terminal → rigid metal probe → probe metal housing → main printed circuit board fixing point → triaxial MEMS accelerometer chip. The rigid metal probe is made of titanium alloy, which is characterized by high strength and low density, making it sturdy yet lightweight. It also possesses excellent corrosion resistance, remaining virtually unaffected by battery acid mist environments. This allows it to transmit minute vibrations from the four terminals of the battery being evaluated to the triaxial microelectromechanical system (MEMS) accelerometer chip, resulting in high measurement accuracy. The probe's metal housing is made of rigid aluminum alloy, which is low in density, easy to process, and can be anodized to obtain a hard, insulating, and corrosion-resistant surface layer. High-strength permanent magnets are integrated around the rigid metal probe, providing a strong and rigid clamping force to ensure lossless transmission of vibrational energy from the terminals to the probe's metal housing. However, relying on the magnetic probe 1 to attach the data acquisition device 2 to the terminals does not guarantee synchronous vibration between the main printed circuit board inside the data acquisition device 2 and its housing. Therefore, a rigid plastic or metal bracket must be used to rigidly fix the main printed circuit board to the housing to lock the relative movement between the main printed circuit board and the housing, thus completing the final link in the vibration transmission chain and achieving lossless transmission of terminal vibration.

[0054] In a preferred embodiment of this application, the magnetic attraction force of the magnetic probe 1 is in the range of 5N-20N to ensure stable contact with the four terminals of the battery to be evaluated under vibration.

[0055] In the preferred embodiment of this application, the magnetic probe 1 does not require professional installation tools for installation, has low installation complexity, fast installation speed, convenient maintenance, and can be easily removed for calibration or replacement. The probe penetrates the oxide layer, has good contact, and high contact reliability.

[0056] In a preferred embodiment of this application, the data acquisition device 2 further includes several temperature and humidity sensors that are communicatively connected to the data processing module 3. The temperature and humidity sensors are not integrated on the magnetic probe, but are separately deployed at the ventilation holes of the battery compartment of the battery 4 to be evaluated or on the inner wall of the battery compartment to measure the ambient temperature and humidity of the battery 4 to be evaluated, obtain the ambient temperature and humidity sequence, and transmit the ambient temperature and humidity sequence to the data processing module 3 through a wireless network for subsequent data analysis and processing.

[0057] In a preferred embodiment of this application, the temperature and humidity sensor is deployed separately at the ventilation hole of the battery compartment or on the inner wall of the battery compartment. This ensures that the temperature and humidity sensor measures environmental data and is not affected by components such as the power module, power resistor, and MCU of the battery health status online assessment device that generate significant heat, as well as the temperature rise of the battery terminals to be assessed, thereby improving the accuracy of environmental temperature and humidity sequence acquisition.

[0058] In a preferred embodiment of this application, the temperature and humidity sensor can also be arranged on the top of the housing of the data acquisition device 2, with the sensor chip mounted on the edge area of ​​the main printed circuit board and its temperature and humidity sensing element aligned with the opening in the housing. Alternatively, the temperature and humidity sensor can be a separate module, connected by a cable, and fixedly installed on the outer wall inside the housing of the data acquisition device 2, and connected to the outside air.

[0059] In a preferred embodiment of this application, the voltage sequence, terminal temperature sequence, internal resistance sequence, terminal micro-vibration signal, and ambient temperature and humidity sequence of the battery 4 to be evaluated, collected by the data acquisition device 2, are processed and packaged by the microcontroller inside the data acquisition device 2, and sent to a smart gateway with edge computing capabilities through an integrated wireless module. The smart gateway communicates with all data acquisition devices 2 in the area through a low-power wireless local area network. The smart gateway is configured to: aggregate and parse the monitoring data from each data acquisition device 2; perform preliminary preprocessing and local caching of the data; upload the processed data to the data processing module 3 through an encrypted channel; and distribute instructions and time synchronization signals to the data acquisition devices 2. The data processing module 3 is deployed on a cloud server or a local server. The data processing module 3 includes an internal resistance health analysis unit, which receives the internal resistance sequence and terminal temperature sequence, and compensates the internal resistance sequences at different temperatures to the internal resistance value at a standard reference temperature based on the temperature characteristic parameters of the battery material, forming an internal resistance reference sequence. Specifically, an internal resistance-temperature relationship model is constructed based on the Arrhenius equation, and the temperature compensation coefficient is calculated based on this model. The formula for calculating the temperature compensation coefficient is as follows:

[0060]

[0061] in, Represents the polar temperature sequence. Indicates a time index. Indicates the temperature compensation coefficient. Indicates the battery activation energy. Represents the gas constant. This indicates the reference temperature, which is a standard temperature value.

[0062] Substituting each temperature value in the electrode temperature sequence into the internal resistance temperature relationship model yields the corresponding internal resistance temperature compensation coefficient. Multiplying each internal resistance value in the internal resistance sequence by the corresponding internal resistance temperature compensation coefficient converts it to a reference temperature, forming an internal resistance reference sequence. The formula for calculating the internal resistance reference sequence is as follows:

[0063]

[0064] in, Indicates the internal resistance reference sequence. This represents the internal resistance sequence.

[0065] Furthermore, trend analysis is performed on the internal reference sequence. A trend analysis time window is set, and a subset of the internal resistance reference sequence within the time window is extracted. Linear regression is then performed on this subset of the internal resistance reference sequence to obtain the trend equation. The rate of change of the slope of the trend equation is calculated. The formula for calculating the rate of change of the slope of the trend equation is as follows:

[0066]

[0067] in, This represents the rate of change of the slope of the trend equation. This indicates the slope of the current trend analysis time window. This represents the slope of the initial health status of the battery to be evaluated.

[0068] The rate of change of the trend equation slope quantifies the deviation of the current internal resistance growth trend from the initial health benchmark. Under ideal health conditions, the internal resistance reference sequence should be basically stable. Since the slope of the current trend analysis time window and the slope of the initial health state of the battery 4 under evaluation are both close to zero, the rate of change of the trend equation slope should also be close to zero. When the battery 4 under evaluation ages or fails, the internal resistance will show an upward trend, the slope of the current trend analysis time window will become a positive value, and the rate of change of the trend equation slope will also increase accordingly. Therefore, the rate of change of the trend equation slope captures the rate of change of internal resistance, which is more sensitive than simply looking at the absolute value of internal resistance and can detect potential problems of the battery 4 under evaluation earlier. Therefore, in the preferred embodiment of this application, the mapping equation between the rate of change of the trend equation slope and the internal resistance health index is:

[0069]

[0070] in, This indicates the internal resistance health index. This represents the sensitivity coefficient. The rate of change of the trend equation's slope is dimensional, and one function of the sensitivity coefficient is to standardize this rate of change, allowing for dimensionless addition and subtraction with the number 1. The unit of the sensitivity coefficient is the reciprocal of the unit of the rate of change of the trend equation's slope. Different types of batteries have different natural rates of internal resistance increase with aging. For batteries with inherently faster internal resistance growth, a smaller sensitivity coefficient can be set to prevent excessive attention to normal aging; for critical equipment requiring high sensitivity, a larger sensitivity coefficient can be set.

[0071] The mapping equation between the rate of change of the trend equation slope and the internal resistance health index in this application is a linear negative correlation model. When the rate of change of the trend equation slope is equal to zero, the internal resistance health index is equal to 1, indicating that the battery 4 to be evaluated is completely healthy. As the rate of change of the trend equation slope increases... As the value of decreases linearly, the internal resistance health index also decreases. However, under extreme fault conditions, the rate of change of the slope of the trend equation can be very large, leading to... It becomes negative, but a negative internal resistance health index is difficult to interpret physically and may cause confusion or errors in subsequent health assessments. To ensure that the internal resistance health index never falls below 0, when... When the value is negative, the internal resistance health index is clamped at 0, defining the lower limit of the internal resistance health index. The entire internal resistance health index is regulated within this range. Within this range, its value is positively correlated with the health status of the battery 4 to be evaluated.

[0072] Data processing module 3 includes a thermal health analysis unit. This unit extracts multiple thermal characteristic parameters from the pole temperature sequence, including absolute temperature, temperature rise rate, temperature standard deviation, and high-temperature duration. Specifically, it calculates the average or representative temperature within the monitoring period to obtain the absolute temperature, reflecting the baseline heat load; it calculates the temperature rise rate per unit time, a forward-looking indicator for predicting thermal runaway risk; it calculates the standard deviation of the pole temperature sequence after removing long-term trends, reflecting the stability of thermal management or external disturbances; and it statistically analyzes the percentage of time the temperature exceeds the safety threshold to obtain the high-temperature duration, reflecting accumulated thermal stress. For each thermal characteristic parameter, a safe operating range, warning threshold, and fault threshold are set. Based on the range in which the characteristic value of the thermal characteristic parameter falls, a linear function maps it to a score from 0 to 1, where 1 represents the ideal state and 0 represents the most dangerous state, resulting in a characteristic score. These multiple characteristic scores are then fused according to preset weights to obtain a thermal health index, which is also standardized within a certain range. Within this range, its value is positively correlated with the health status of the battery 4 to be evaluated.

[0073] Data processing module 3 includes a mechanical health analysis unit. This unit combines the independent X, Y, and Z-axis acceleration sequences acquired by the triaxial microelectromechanical system accelerometer chip into a single total vibration vector sequence to comprehensively capture vibrations in any direction. The mean of the total vibration vector sequence is subtracted to eliminate the influence of constant offsets such as gravitational acceleration, resulting in a pure AC vibration signal. A bandpass filter is used to retain the mid-to-high frequency vibration components caused by loose mechanical connections and mechanical impacts, while filtering out extremely low-frequency slow tilting movements and high-frequency electronic noise, yielding a time-domain signal. A fast Fourier transform is then performed on the time-domain signal to obtain the frequency-domain signal. When the battery terminal connection is loose, the vibration energy will significantly increase within a specific frequency range. In the preferred embodiment of this application, a specific frequency range is determined through experimental simulation, and the frequency-domain signal energy of the terminal micro-vibration signal within this specific frequency range is calculated as the current vibration energy characterizing the loosening fault. The calculation formula is as follows:

[0074]

[0075] in, This represents the current vibration energy, characterizing the vibration energy associated with the loosening fault. Indicates frequency index, This represents the discrete spectrum index corresponding to the lower limit of a specific frequency band. This represents the discrete spectrum index corresponding to the upper limit of a specific frequency band. The first frequency domain signal represents the second frequency domain signal. A discrete value.

[0076] The specific frequency band can be determined based on the battery model and mechanical structure through impact experiments, finite element simulations, or statistical analysis of historical fault data. In a preferred embodiment, the specific frequency band determined by experimental simulation is from 200 Hz to 500 Hz.

[0077] Furthermore, the mechanical health index is obtained based on the ratio between the vibration energy baseline and the current vibration energy. The vibration energy baseline is a reference value for vibration energy within a specific frequency range, determined through experimental measurement and statistical analysis when the battery system is in a healthy state. For example, it represents the normal vibration level under normal operating conditions without loosening, after factory inspection and installation without faults. During the experimental simulation, the operating conditions of the battery 4 to be evaluated should cover typical operating states, such as static conditions when the equipment is shut down and stationary, dynamic conditions with different charge / discharge rates and load conditions, and conditions with different ambient temperatures. The mechanical health index is also standardized when the vibration energy baseline is less than or equal to the current vibration energy. Within this range, and the larger the value, the closer the current vibration energy is to the vibration energy baseline. Its value is positively correlated with the health status of the battery 4 to be evaluated.

[0078] Data processing module 3 includes an environmental risk analysis unit. This unit scores the environmental temperature and humidity sequence based on a preset environmental safety operating range and obtains an environmental health index based on the resulting environmental scores. The preset environmental safety operating range refers to the optimal environmental temperature and humidity operating range determined based on the electrochemical characteristics of the battery 4 to be evaluated, including both temperature and humidity safety operating ranges. During the scoring process, for each individual environmental temperature or humidity value collected at any given time, a single-point environmental temperature score or a single-point environmental humidity score is calculated based on its deviation from the preset safety operating range. In a preferred embodiment of this application, a piecewise linear scoring function is used to measure the degree to which environmental temperature and humidity deviate from the preset safety operating range. The piecewise linear scoring function corresponding to the environmental temperature is shown below:

[0079]

[0080] in, This indicates the ambient temperature score. Indicates the ambient temperature value. This represents the critical value for ambient temperature. This indicates the upper limit of the safe operating temperature range. This indicates the lower limit of the safe operating temperature range. This represents the critical value at ambient temperature.

[0081] The critical values ​​for ambient temperature, ambient temperature, upper and lower limits of the safe operating temperature range are determined based on the environmental temperature requirements of the battery type.

[0082] The piecewise linear scoring function corresponding to ambient humidity is shown below:

[0083]

[0084] in, This indicates the ambient humidity score. Indicates the ambient humidity value. This indicates the critical value for ambient humidity. This indicates the upper limit of humidity within the safe operating range. This indicates the lower limit of the humidity range within the safe operating humidity zone. This indicates the critical value under ambient humidity.

[0085] The upper and lower threshold values ​​of ambient humidity, the upper limit of humidity in the safe operating range, and the lower limit of humidity in the safe operating range are determined based on the environmental humidity requirements of the battery type.

[0086] For all valid data points within a scoring period, the mean of their ambient temperature scores and the mean of their ambient humidity scores are calculated respectively, yielding the corresponding ambient temperature and humidity scores. Finally, these scores are mapped to a standardized risk scale and weighted according to preset weights to obtain the environmental health index. The formula for calculating the environmental health index is as follows:

[0087]

[0088] in, Indicates the environmental health index. and This represents the weighting coefficient. The environmental health index is also standardized within... Within this range, its value is positively correlated with the health status of the battery 4 to be evaluated.

[0089] Data processing module 3 includes a stability analysis unit. This unit calculates a comprehensive entropy value based on the first entropy value of the terminal temperature sequence and the second entropy value of the voltage sequence, and maps this comprehensive entropy value to a stability index. The stability index in this application is used to quantify the randomness and disorder of the operating state of the battery 4 under evaluation. A lower value indicates a more stable and predictable operation of the battery 4. For a healthy and stable battery, fluctuations in terminal temperature and voltage are mainly caused by predictable load changes or environmental factors. The time series corresponding to terminal temperature and voltage exhibits high regularity, resulting in a low information entropy value. Conversely, when a battery ages or experiences internal faults, its electrochemical reactions and thermodynamic states become unstable, leading to unpredictable random fluctuations in terminal temperature and voltage. The complexity of the time series corresponding to terminal temperature and voltage increases, and the corresponding information entropy value rises. Therefore, by calculating the first entropy value of the terminal temperature sequence and the second entropy value of the voltage sequence and mapping them to a standardized index, a quantitative assessment of the operating stability of the battery 4 under evaluation is achieved. Specifically, the sample entropy algorithm is used to calculate the first entropy value of the pole temperature sequence and the second entropy value of the voltage sequence. The calculation process of the sample entropy algorithm is as follows:

[0090] For a length of Time series, setting embedding dimensions Similarity tolerance The formula for calculating the similarity tolerance is:

[0091]

[0092] in, Representing time series The standard deviation.

[0093] structure indivual A dimensional vector, represented as ,in, , The first time series The numerical values ​​corresponding to each sequence.

[0094] definition For vectors With the remaining vectors The proportion of numbers whose Chebyshev distance is less than the similarity tolerance, where, Then define the following parameters:

[0095]

[0096] Increase the embedding dimension to Repeat the above calculations to obtain Then the formula for calculating the sample entropy of this time series is:

[0097]

[0098] Substituting the electrode temperature sequence as a time series into the above sample entropy calculation process yields the first entropy value. Substituting the voltage sequence as a time series into the above sample entropy calculation process yields the second entropy value. The first and second entropy values ​​are then weighted and fused according to preset weights to obtain the comprehensive entropy value. Following a preset mapping function, the comprehensive entropy value is mapped to a stability index. The formula for calculating the stability index is as follows:

[0099]

[0100] in, Indicates the stability index. This represents the overall entropy value. This represents the first sensitivity coefficient, used to adjust the sensitivity of the stability index to the entropy value. The stability index is also normalized to... Within this range, its value is positively correlated with the health status of the battery 4 to be evaluated.

[0101] In a preferred embodiment of this application, the disorder of the terminal temperature and voltage of the battery to be evaluated reflects the battery's internal health condition, thereby capturing early, gradual performance degradation of the battery to be evaluated.

[0102] Data processing module 3 includes a system consistency analysis unit. This unit obtains predicted voltage sequences corresponding to the terminal temperature sequence and internal resistance sequence based on a dynamic relationship model constructed from healthy operating data. It also obtains the system consistency index based on the variance of the residual sequence between the predicted voltage sequence and the actual voltage sequence. In a healthy battery, there is a definite physical correlation between voltage, terminal temperature, and internal resistance. When the battery ages or fails, this physical correlation is disrupted. Therefore, the stability of the dynamic relationship between battery voltage, terminal temperature, and resistance characterizes the battery's health status. Specifically, continuous voltage, terminal temperature, and resistance data of the battery under evaluation are selected during its healthy operating state, such as the first three months after factory acceptance or the stable operating period after the most recent successful verification discharge test. All data undergoes time alignment and outlier removal. A multiple linear regression model was chosen to construct the voltage prediction model. Using healthy operating data, the regression coefficients of the voltage prediction model were solved using the least squares method to minimize the mean square error between the predicted and actual voltage values, thus obtaining a dynamic relationship model between voltage, electrode temperature, and resistance. This dynamic relationship model is not limited to multiple linear regression; support vector machine regression, neural networks, and other machine learning models can also be used, as long as they can effectively predict voltage based on electrode temperature and internal resistance.

[0103] In real-time evaluation, the synchronously sampled pole temperature sequence and internal resistance sequence during the evaluation period are input into the dynamic relationship model to obtain the predicted voltage sequence. The variance of the residual sequence between the predicted voltage sequence and the synchronously sampled voltage sequence is calculated and mapped to the system consistency index. The formula for calculating the system consistency coefficient is as follows:

[0104]

[0105] in, Represents the system consistency index. This represents the second sensitivity coefficient, used to adjust the sensitivity of the consistency coefficient to the variance of the residual series. This represents the variance of the residual sequence. The value of the second sensitivity coefficient is calibrated using health-state operating data, ensuring that the average system consistency index of the battery under health conditions being evaluated is above 0.9. The system consistency index is also standardized within... Within this range, its value is positively correlated with the health status of the battery being evaluated.

[0106] In the preferred embodiment of this application, the internal resistance health index reflects the aging degree of the battery's electrochemical body, the thermal health index reflects the battery's thermodynamic state and heat dissipation, the mechanical health index reflects the tightness of the terminal mechanical connection, the environmental health index quantifies the impact of external environmental stress on the battery's state, the stability index characterizes the disorder of the battery's output, and the system consistency index reveals the stability of the synergistic relationship between the battery's internal parameters. This constructs a multi-dimensional, multi-physical field battery health assessment system to achieve a comprehensive, early, accurate, and interpretable assessment of the battery's health status.

[0107] Data processing module 3 also includes a health status assessment unit. This unit is configured to: set a health status identification framework for the battery 4 to be assessed; and construct state membership functions based on the health status identification framework. These state membership functions include: a first state membership function corresponding to the internal resistance health index, a second state membership function corresponding to the thermal health index, a third state membership function corresponding to the mechanical health index, a fourth state membership function corresponding to the environmental health index, a fifth state membership function corresponding to the stability index, and a sixth state membership function corresponding to the system consistency index. Based on these state membership functions, the unit obtains the state membership distribution corresponding to the battery to be assessed. This state membership distribution includes: a first state membership distribution corresponding to the internal resistance health index, a second state membership distribution corresponding to the thermal health index, a third state membership distribution corresponding to the mechanical health index, and a sixth state membership function corresponding to the environmental health index. The system employs a four-state membership distribution, a fifth-state membership distribution corresponding to the stability index, and a sixth-state membership distribution corresponding to the system consistency index. The state memberships are then normalized and assigned uncertainty to obtain basic probability distributions. These basic probability distributions include: a first basic probability distribution corresponding to the internal resistance health index, a second basic probability distribution corresponding to the thermal health index, a third basic probability distribution corresponding to the mechanical health index, a fourth basic probability distribution corresponding to the environmental health index, a fifth basic probability distribution corresponding to the stability index, and a sixth basic probability distribution corresponding to the system consistency index. DS evidence fusion is performed on the first, second, third, fourth, fifth, and sixth basic probability distributions to obtain a fused basic probability distribution. Based on this fused basic probability distribution, the health status assessment result of the battery to be evaluated is determined. In a preferred embodiment of this application, a health status identification framework for the battery 4 to be evaluated is established, and the identification framework is represented as follows:

[0108]

[0109] in, A framework for identifying the health status of a battery. It indicates good health. Indicates a warning. Indicates a malfunction.

[0110] State membership functions are set for the internal resistance health index, thermal health index, mechanical health index, environmental health index, stability index, and system consistency index. These state membership functions are constructed based on trapezoidal membership functions. The shape parameters of the trapezoidal membership function are determined through cluster analysis of historical state data, including historical health state data, historical fault state data, and historical warning state data. The four shape parameters of the trapezoidal membership function are defined as follows: , , , ,in The mathematical expression for the trapezoidal membership function is:

[0111]

[0112] in, Indicates data to be processed The membership values, in this application, the data to be processed include internal resistance health index, thermal health index, mechanical health index, environmental health index, stability index, and system consistency index. This represents the left boundary point where the membership degree begins to increase. This represents the left boundary point where the membership degree reaches 1. This represents the right boundary point where the membership degree remains at 1. This represents the right boundary point where the membership degree drops to 0.

[0113] For different types of data to be processed, cluster analysis is performed on their corresponding historical state data to determine their shape parameters. Specifically, a training dataset is constructed based on the historical state data, and the training dataset is represented as follows:

[0114]

[0115] in, Indicates the first The feature values ​​of each training sample include internal resistance health index, thermal health index, mechanical health index, environmental health index, stability index, and system consistency index. Indicates the first The state labels of each training sample come from historical maintenance records or experimental calibrations, and belong to a certain state in the health status identification box. This indicates the number of training samples.

[0116] For each of the six coefficients, an independent clustering analysis process is performed to determine the shape parameters of its trapezoidal membership function. Specifically, for each coefficient among the internal resistance health index, thermal health index, mechanical health index, environmental health index, stability index, and system consistency index, a corresponding training dataset is selected from historical state data. Based on the state label, the training dataset is grouped, and all samples labeled "healthy" are extracted to obtain the healthy subset of the internal resistance health index. Similarly, the early warning subset and fault subset of the internal resistance health system are obtained. For each subset, the number of clusters is set to 2, and K-means clustering is used to perform clustering analysis. The feature values ​​of each healthy state are further subdivided into typical value clusters and boundary value clusters. The typical value cluster represents the most typical and concentrated numerical range in that state, while the boundary value cluster represents the numerical range transitioning from adjacent states in that state. To determine which cluster is the typical value cluster, the cluster centers of the two clusters are calculated. The cluster whose cluster center is closer to the overall mean of the subset of that state is defined as the typical value cluster, and the other cluster is defined as the boundary value cluster. Based on the clustering results, the minimum value in the typical value cluster is taken as... The value, the maximum value of the data in the typical value cluster is The minimum value of the data in the boundary value cluster is The maximum value of the data in the boundary value cluster is... Value; further determine whether it is satisfied. To ensure the effectiveness of the trapezoidal membership function, if the clustering results do not meet this condition, the range of the boundary value clusters can be slightly expanded, such as by... The value decreased by 5%. The value increases by 5%. The determination of the shape parameters of traditional trapezoidal membership functions relies heavily on the personal experience of experts, resulting in subjectivity and inconsistency. This invention uses data-driven cluster analysis to objectively extract shape parameters from historical state data, eliminating subjective biases caused by human settings. It directly reflects the numerical distribution characteristics of the battery under various states during actual operation, enabling the constructed membership function to more accurately depict the continuous gradual change process from health to failure, thereby significantly improving the accuracy of health status assessment.

[0117] In a preferred embodiment of this application, taking the internal resistance health index as an example, the state membership function designed for the health state is as follows:

[0118]

[0119] in, The membership function of the internal resistance health index to the health status is represented.

[0120] In one specific embodiment, when hour:

[0121] because ,but .

[0122] when hour:

[0123] because ,but .

[0124] when hour:

[0125] because ,but .

[0126] Using the same method, the state membership function designed for the early warning state is as follows:

[0127]

[0128] in, The membership function represents the internal resistance health index as belonging to the warning state.

[0129] The state membership function designed for the fault state is:

[0130]

[0131] in, The internal resistance health index represents the state membership function of the fault state.

[0132] Further normalization of the obtained membership values ​​yields a preliminary basic probability assignment. The aforementioned membership function is then used to... Taking an example, the distribution of state membership degrees is obtained as follows:

[0133]

[0134]

[0135]

[0136] After normalization, the preliminary basic probability assignment (BPA) is obtained:

[0137]

[0138]

[0139]

[0140]

[0141] Furthermore, by allocating uncertainty, assuming an uncertainty factor of 0.9, we obtain the basic probability allocation:

[0142]

[0143]

[0144]

[0145]

[0146] The final basic probability distribution corresponding to the internal resistance health index is expressed as follows:

[0147]

[0148] For the thermal health index, the above method is used to determine the second health state membership function of the thermal health index. Then, based on the second health state membership function, the second state membership is obtained. The second state membership is then normalized and assigned uncertainty in sequence to obtain the second basic probability assignment corresponding to the thermal health index.

[0149] For the mechanical health index, the above method is used to determine the membership function of the third health state of the mechanical health index. Then, based on the membership function of the third health state, the membership distribution of the third state is obtained. The membership distribution of the third state is then normalized and the uncertainty of the distribution is processed in sequence to obtain the third basic probability allocation corresponding to the mechanical health index.

[0150] For the environmental health index, the above method is used to determine the membership function of the fourth health state of the environmental health index. Then, based on the membership function of the fourth health state, the membership distribution of the fourth state is obtained. The membership distribution of the fourth state is then normalized and the uncertainty of the allocation is processed in sequence to obtain the fourth basic probability allocation corresponding to the environmental health index.

[0151] For the stability index, the membership function of the fifth health state is determined using the above method. Then, based on the membership function of the fifth health state, the membership distribution of the fifth state is obtained. The membership distribution of the fifth state is then normalized and assigned uncertainty in sequence to obtain the fifth basic probability distribution corresponding to the stability index.

[0152] For the system consistency index, the above method is used to determine the membership function of the sixth health state of the system consistency index. Then, based on the membership function of the sixth health state, the membership distribution of the sixth state is obtained. The membership distribution of the sixth state is then normalized and assigned uncertainty in sequence to obtain the sixth basic probability allocation corresponding to the system consistency index.

[0153] Furthermore, the first, second, third, fourth, fifth, and sixth basic probability allocations are fused according to the DS combination rule to obtain the fused basic probability allocation. Based on the fused basic probability allocation, the health status assessment result of the battery to be evaluated is determined. In a preferred embodiment of this application, a recursive DS combination rule is used to sequentially fuse the first, second, third, fourth, fifth, and sixth basic probability allocations pairwise. First, the first and second basic probability allocations are fused, and the fusion process is described as follows:

[0154]

[0155] in, This represents the fused basic probability assignment corresponding to the first and second basic probability assignments. Indicates a fusion operation. This represents the first basic probability assignment. This represents the second basic probability assignment.

[0156] For any proposition Then we have:

[0157]

[0158] in, This represents the result of combining the first and second basic probability assignments on the proposition. probability allocation, This represents the first basic probability assignment for a specific proposition. Reliability, This indicates that in the second basic probability assignment, a specific proposition... Reliability, Indicates the conflict coefficient.

[0159] denominator This is a normalization factor used to redistribute the remaining reliability proportionally after excluding conflicting reliability, ensuring... It remains a valid BPA, and the formula for calculating the conflict factor is:

[0160]

[0161] Furthermore, the basic probability allocation resulting from the fusion of the first and second basic probability allocations is recursively fused with the third basic probability allocation. This recursive fusion operation is represented as follows:

[0162]

[0163]

[0164]

[0165]

[0166] in, This represents the fused basic probability assignment corresponding to the first, second, and third basic probability assignments. This represents the third basic probability assignment. This represents the fused basic probability allocation corresponding to the first, second, third, and fourth basic probability allocations. This represents the fourth basic probability assignment. This represents the fused basic probability allocation corresponding to the first, second, third, fourth, and fifth basic probability allocations. This represents the fifth basic probability distribution. This represents the fused basic probability allocation corresponding to the first, second, third, fourth, and fifth basic probability allocations. This represents the sixth basic probability distribution.

[0167] Further, choose The state corresponding to the single-point proposition with the highest reliability is used as the health status assessment result for the battery 4 to be evaluated. The system outputs the contribution of internal resistance health index, thermal health index, mechanical health index, environmental health index, stability index, and system consistency index to this health status assessment result, in order to accurately pinpoint the cause of the failure. If the reliability of the two single-point propositions after fusion is very close, the uncertainty of the decision result is high. In this case, the online battery health status assessment system will output an uncertainty flag and prioritize safety, tending to select a more conservative deterioration state. For example, when the reliability of health and warning are close, it will prioritize the warning state.

[0168] In a preferred embodiment of this application, the DS evidence theory is employed to effectively fuse and scientifically decide on multi-dimensional and uncertain battery health information, ultimately outputting a reliable health status assessment result. By explicitly assigning confidence to the entire set, the inherent uncertainty of each evidence source is preserved, and conflicts are handled through normalization during the fusion process, making the final decision result more robust. By integrating multi-dimensional information such as battery electrical, thermodynamic, mechanical vibration, and environmental factors, a more comprehensive, robust, and accurate assessment of battery health status is achieved.

[0169] In a preferred embodiment of this application, when performing DS evidence fusion, the health status assessment unit also simultaneously analyzes the contribution of the basic probability allocation values ​​corresponding to the internal resistance health index, thermal health index, mechanical health index, environmental health index, stability index, and system consistency index to the health status assessment result. The contribution is the degree of influence of the corresponding index on the final basic probability allocation value during the evidence fusion process. Specifically, taking the health status assessment result as a fault state and the fusion of the first basic probability allocation and the second basic probability allocation as an example, in the fusion, all those satisfying... The product values ​​of the terms are all accumulated in the final confidence score. During DS evidence fusion, a contribution record table is created for each index to record its contribution to the fault state in each fusion step. For each pairwise fusion step, the values ​​that satisfy the criteria are recorded. Analyze the sources of contribution for each item, if Then The values ​​are accumulated into the internal resistance health index. In the direct contribution, if, then The value is accumulated into the thermal health index. In the direct contributions, if ,but and Then The values ​​are proportionally allocated to the internal resistance health index. In the indirect contributions, The values ​​are proportionally allocated to the thermal health index. In the indirect contributions, the allocation ratio is based on The percentage of each item's contribution to the total contribution of all items to the failure reliability, and The contribution of each item to the fault reliability is determined as a percentage of the total contribution of all items to the fault reliability. After all fusion steps of DS evidence fusion are completed, an initial contribution is obtained based on direct and indirect contributions. The initial contribution is normalized to obtain the contribution of the basic probability allocations of the internal resistance health index, thermal health index, mechanical health index, environmental health index, stability index, and system consistency index to the health status assessment results. The influence weight of each indicator is quantified, and the root causes of the health status are traced, which greatly improves the credibility and acceptability of the assessment results.

[0170] To ensure the accuracy of the raw data relied upon by the health status assessment system, in a preferred embodiment of this application, the online battery health status assessment system further includes a self-calibration module. This module controls the data acquisition device 2 to acquire the reference voltage and reference internal resistance when the battery 4 to be assessed is stationary and unloaded. Based on the reference voltage and reference internal resistance, the measurement circuit undergoes zero-point calibration and gain calibration. Specifically, when the battery 4 to be assessed is in a stationary state, i.e., at least 2 hours after the charging and discharging operation is completed, and completely disconnected from all external loads and charging equipment, the open-circuit voltage between the positive and negative terminals of the battery is acquired and recorded as the reference voltage. The internal resistance of the battery is measured using a short-time DC pulse method and recorded as the reference internal resistance. The reference voltage and reference internal resistance reflect the inherent electrical characteristics of the battery in its current stationary and stable state. Furthermore, the measurement circuit is calibrated, including zero-point calibration and gain calibration. Zero-point calibration is used to eliminate the inherent offset voltage of the measurement circuit. The data acquisition device 2 connects the measurement circuit to a known zero-potential reference point via a relay. The reading acquired at this time is the zero-point offset value, and all subsequent voltage measurement raw values ​​must undergo zero-point compensation. Gain calibration is used to correct proportional errors in the measurement loop, ensuring that the measured value is accurately proportional to the true value. Using the reference voltage as the standard voltage value, and based on zero-point calibration, the voltage gain correction coefficient is calculated using the following formula:

[0171]

[0172] in, This represents the voltage gain correction factor. Indicates the reference voltage. This represents the reading after zero-point compensation when acquiring the reference voltage.

[0173] After zero-point compensation, the voltage value needs to be multiplied by the voltage gain correction factor for further gain correction to obtain the final corrected voltage value.

[0174] Furthermore, using a reference internal resistance as the standard internal resistance value, a test current is applied to the measurement circuit, and the voltage change caused by this test current is measured. Based on the voltage change, the uncorrected original internal resistance value of the battery to be evaluated is calculated. The calculation formula is as follows:

[0175]

[0176] in, This represents the uncorrected original internal resistance value of the battery to be evaluated. This represents the voltage change during the gain calibration process. This represents the noise offset measured without current excitation, which can be obtained by averaging multiple samples.

[0177] Furthermore, based on the original, uncorrected internal resistance value of the battery to be evaluated, the gain correction coefficient is calculated using the following formula:

[0178]

[0179] in, This represents the gain correction factor.

[0180] Subsequently, all real-time measured internal resistance values ​​need to be multiplied by the voltage gain correction factor for further gain correction to obtain the final gain-corrected internal resistance value.

[0181] In the preferred embodiment of this application, zero-point calibration and gain calibration effectively eliminate system errors introduced by factors such as sensor offset, amplifier temperature drift, and device aging in the data acquisition device, ensuring the measurement accuracy of the two key parameters, voltage and internal resistance, and laying the foundation for the accuracy of subsequent health status assessment. The self-calibration process uses the static state of the battery under evaluation as a benchmark, eliminating the need for external high-precision standard instruments, achieving self-contained calibration, and adapting to different battery cells and slowly changing environmental conditions, ensuring the stability and reliability of long-term monitoring.

[0182] The online battery health status assessment system also includes a display and alarm unit for real-time display of health status assessment results and triggering audible and visual alarms when the health status assessment results are warnings or faults. The core components of the display and alarm unit in this application include a display module, an audible and visual alarm module, and a communication interface. The display module can use an LCD screen, a light-emitting diode array, or a touch screen to achieve visual information output. The audible and visual alarm module includes multi-color LED indicators and a buzzer. The communication interface supports pushing alarm information to a remote monitoring center or mobile terminal.

[0183] In summary, the online battery health status assessment system provided in this embodiment addresses the technical issues of improving the convenience of battery status data collection and the accuracy of battery health status assessment. The system includes: a magnetic probe matched to the terminals of the battery to be evaluated; a data acquisition device fixedly connected to the magnetic probe for acquiring the voltage sequence, terminal temperature sequence, internal resistance sequence, terminal micro-vibration signal, and ambient temperature and humidity sequence of the battery to be evaluated; and a data processing module configured to: perform temperature compensation on the internal resistance sequence based on the terminal temperature sequence to obtain an internal resistance reference sequence; perform trend analysis on the internal resistance reference sequence to obtain an internal resistance health index; obtain a thermal health index based on the terminal temperature sequence; obtain a mechanical health index based on the terminal micro-vibration signal; obtain an environmental health index based on the ambient temperature and humidity sequence; obtain a stability index based on the terminal temperature sequence and voltage sequence; obtain a system consistency index based on the residual generated by the dynamic relationship model constructed from the voltage sequence, terminal temperature sequence, and internal resistance sequence; and perform evidence fusion analysis on the internal resistance health index, thermal health index, mechanical health index, environmental health index, stability index, and system consistency index using DS evidence theory to obtain the health status assessment result of the battery to be evaluated. The online battery health status assessment system disclosed in this application uses a magnetic probe, eliminating the need for specialized installation tools. This results in low installation complexity, fast installation speed, and convenient maintenance. The probe can be easily removed for calibration or replacement. It penetrates the oxide layer, ensuring good contact and high reliability. The DS evidence theory effectively integrates and scientifically decides on multi-dimensional and uncertain battery health information, ultimately outputting a reliable health status assessment result. By explicitly assigning confidence to the entire set, it preserves the inherent uncertainty of each evidence source and handles conflicts through normalization during the fusion process, making the final decision result more robust and supporting comprehensive health status assessments with early warning capabilities.

[0184] The various embodiments in this specification are described in a progressive manner. For directly identical or similar parts of the embodiments, refer to each other. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. It should be noted that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0185] The above-described embodiments are merely preferred embodiments of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various improvements and substitutions without departing from the principles of the present invention, and these improvements and substitutions should also be considered within the scope of protection of the present invention. Therefore, the scope of protection of this invention should be determined by the scope of the claims.

Claims

1. An online health status assessment system for batteries, characterized in that, The system includes: A magnetic probe is used to match the terminals of the battery being evaluated. The data acquisition device is fixedly connected to the magnetic probe and is used to acquire the voltage sequence, terminal temperature sequence, internal resistance sequence, terminal micro-vibration signal and ambient temperature and humidity sequence of the battery to be evaluated. The data processing module is configured as follows: Based on the pole temperature sequence, the internal resistance sequence is temperature compensated to obtain an internal resistance reference sequence. Trend analysis is then performed on the internal resistance reference sequence to obtain an internal resistance health index. The thermal health index is obtained based on the electrode temperature sequence. The mechanical health index is obtained based on the micro-vibration signal of the pole column; The environmental health index is obtained based on the environmental temperature and humidity sequence. The stability index is obtained based on the electrode temperature sequence and the voltage sequence. The system consistency index is obtained based on the residuals generated by the dynamic relationship model constructed from the voltage sequence, the pole temperature sequence, and the internal resistance sequence. The DS evidence theory is used to perform evidence fusion analysis on the internal resistance health index, the thermal health index, the mechanical health index, the environmental health index, the stability index, and the system consistency index to obtain the health status assessment result of the battery to be evaluated.

2. The online battery health status assessment system as described in claim 1, characterized in that, The magnetic probe integrates a spring probe, a temperature sensor, a rigid metal probe, and a metal housing for the probe. The data acquisition device includes a measurement circuit and an accelerometer chip. The spring probe forms an electrical connection with the surface of the battery terminal to be evaluated. The spring probe is fixedly connected to the measurement circuit, which includes a microcontroller, an analog-to-digital converter, and a precision power resistor controlled by a metal-oxide-semiconductor field-effect transistor to perform voltage sequence acquisition and internal resistance sequence acquisition. The temperature sensor is connected to the metal base of the spring probe via a thermal connection structure, so that a low thermal resistance path is formed between the temperature sensing area of ​​the temperature sensor and the metal base. One end of the rigid metal probe forms a tight mechanical contact with the surface of the battery terminal to be evaluated, and the other end is rigidly connected to the metal housing of the probe. The accelerometer chip is rigidly mounted in the center of the main printed circuit board of the data processing module and is fixedly connected to the metal shell of the probe through a rigid support structure to perform pole micro-vibration signal acquisition.

3. The online battery health status assessment system as described in claim 2, characterized in that, The spring probe is made of high carbon spring steel, and the head of the spring probe is plated with thick gold or hard gold. The rigid metal probe is made of titanium alloy, and the metal shell of the probe is made of aluminum alloy.

4. The online battery health status assessment system as described in claim 1, characterized in that, The data acquisition device also includes several temperature and humidity sensors that are communicatively connected to the data processing module; The temperature and humidity sensors are distributed and installed at the ventilation holes of the battery compartment of the battery to be evaluated or on the inner wall of the battery compartment.

5. The online battery health status assessment system as described in claim 1, characterized in that, The data processing module includes: The internal resistance health analysis unit is used to perform linear fitting on the internal resistance reference sequence and use the rate of change of the slope of the trend equation obtained by linear fitting as the internal resistance health index. The thermal health analysis unit is used to extract multiple thermal characteristic parameters of the pole temperature sequence, score each thermal characteristic parameter, and fuse the corresponding multiple characteristic score results according to a preset weight to obtain a thermal health index. The mechanical health analysis unit is used to extract the time-domain and frequency-domain features of the pole micro-vibration signal, obtain the current vibration energy based on the time-domain and frequency-domain features, and obtain the mechanical health index based on the ratio between the preset vibration energy baseline and the current vibration energy. The environmental risk analysis unit is used to score the environmental temperature and humidity sequence according to the preset environmental safety working range, and to obtain the environmental health index based on the obtained environmental score results.

6. The online battery health status assessment system as described in claim 1, characterized in that, The data processing module further includes: The stability analysis unit is used to obtain a comprehensive entropy value based on the first entropy value of the pole temperature sequence and the second entropy value of the voltage sequence, and to map the comprehensive entropy value into a stability index. The system consistency analysis unit is used to obtain the predicted voltage sequence corresponding to the pole temperature sequence and the internal resistance sequence based on the dynamic relationship model constructed based on the health status operation data, and to obtain the system consistency index based on the variance of the residual sequence between the predicted voltage sequence and the voltage sequence.

7. The online battery health status assessment system as described in claim 1, characterized in that, The data processing module further includes: The health status assessment unit is configured as follows: A health status identification framework for the battery to be evaluated is established, and a state membership function is constructed based on the health status identification framework. The state membership function includes: a first state membership function corresponding to the internal resistance health index, a second state membership function corresponding to the thermal health index, a third state membership function corresponding to the mechanical health index, a fourth state membership function corresponding to the environmental health index, a fifth state membership function corresponding to the stability index, and a sixth state membership function corresponding to the system consistency index. Based on the state membership function, the state membership distribution corresponding to the battery to be evaluated is obtained. The state membership distribution includes: a first state membership distribution corresponding to the internal resistance health index, a second state membership distribution corresponding to the thermal health index, a third state membership distribution corresponding to the mechanical health index, a fourth state membership distribution corresponding to the environmental health index, a fifth state membership distribution corresponding to the stability index, and a sixth state membership distribution corresponding to the system consistency index. The state membership distribution is normalized and uncertainty is assigned sequentially to obtain a basic probability allocation, which includes: a first basic probability allocation corresponding to the internal resistance health index, a second basic probability allocation corresponding to the thermal health index, a third basic probability allocation corresponding to the mechanical health index, a fourth basic probability allocation corresponding to the environmental health index, a fifth basic probability allocation corresponding to the stability index, and a sixth basic probability allocation corresponding to the system consistency index. DS evidence fusion is performed on the first basic probability allocation, the second basic probability allocation, the third basic probability allocation, the fourth basic probability allocation, the fifth basic probability allocation, and the sixth basic probability allocation to obtain the fused basic probability allocation. Based on the fused basic probability allocation, the health status assessment result of the battery to be evaluated is determined.

8. The online battery health status assessment system as described in claim 7, characterized in that, The first, second, third, fourth, fifth, and sixth basic probability allocation functions are all constructed based on trapezoidal membership functions, the shape parameters of which are determined by cluster analysis of historical state data.

9. The online battery health status assessment system as described in claim 7, characterized in that, When performing the DS evidence fusion, the health status assessment unit also simultaneously analyzes the contribution of the basic probability allocation corresponding to the internal resistance health index, the thermal health index, the mechanical health index, the environmental health index, the stability index, and the system consistency index to the health status assessment result.

10. The online battery health status assessment system as described in claim 1, characterized in that, The system also includes a self-calibration module, which controls the data acquisition device to acquire a reference voltage and a reference internal resistance when the battery to be evaluated is idle and without load, and performs zero-point calibration and gain calibration on the measurement circuit based on the reference voltage and the reference internal resistance.

Images