Method for predicting battery capacity attenuation in charge-discharge cycle

Abstract

The invention provides a battery capacity attenuation prediction method in charge and discharge circulation, which comprises the following steps: continuously monitoring a contact resistance amplification value of an interface to obtain non-uniform contact resistance distribution data, and determining a time point and amplitude of abnormal resistance rise according to the non-uniform data; analyzing the voltage drop amplitude according to the distribution rule of the voltage abnormal signal to obtain the potential drop trend of the battery capacity retention ratio, and determining the associated data of the battery capacity attenuation and the voltage abnormal signal distribution rule according to the potential drop trend of the battery capacity retention ratio; and based on the correlation characteristics, the moisture permeation rate and the coating durability, constructing a battery capacity attenuation evaluation system, inputting non-uniform contact resistance distribution and current loss, outputting a capacity attenuation trend, drawing a curve, and predicting the battery capacity according to the curve.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method for predicting battery capacity attenuation during charge and discharge cycles. Background Art

[0002] In the field of electric vehicle charging technology, stable charging station operation and battery life prediction are crucial for promoting the popularization and sustainable development of new energy vehicles. Research in this area is directly related to energy efficiency and user experience, and is an indispensable component of a green transportation system. However, many current charging station systems lack adaptability in complex environments, especially under extreme weather conditions. Existing methods often overlook the impact of environmental factors on device performance and data collection, resulting in inaccurate prediction results. Specifically, existing solutions generally lack the ability to dynamically correct for environmental interference in the high-salt and humid environment of typhoon season, making it difficult to address the performance degradation caused by moisture and salt corrosion. This significantly reduces the accuracy of battery capacity decay prediction, and new technical solutions are urgently needed to fill this gap. The core challenge faced by research is the accumulation of salt crystals in the charging interface due to moisture penetration. This phenomenon gradually intensifies over time, blocking the contact surface and causing an abnormal increase in contact resistance. This increased contact resistance directly interferes with the stability of the charging voltage, causing deviations in battery cycle data collection, and ultimately affecting the reliability of capacity decay prediction. This series of problems, from environmental erosion to device performance degradation and then to compromised data accuracy, constitutes a complete chain of technical challenges. Therefore, a key issue that needs to be addressed in this research is how to improve the accuracy of battery capacity decay prediction by integrating meteorological warning information with real-time interface impedance detection data in the high-salt and humid environment of typhoon season and dynamically correcting environmental interference parameters. Summary of the Invention

[0003] The present invention provides a method for predicting battery capacity attenuation during charge and discharge cycles, which mainly includes:

[0004] Real-time data collection of high-salt and humid environments is used to calculate moisture penetration rate and residence time. Combined with the surface roughness of the charging interface material and the durability of the protective coating, salt deposition concentration and crystallization distribution are analyzed to identify surface contact resistance increases and material corrosion characteristics.

[0005] Continuously monitor the contact resistance increase value of the interface to obtain the uneven distribution data of the contact resistance, and determine the time point and amplitude of the abnormal resistance increase based on the uneven data;

[0006] By integrating the contact resistance increase and uneven distribution data with the high-salt humidity environment data, the moisture retention time and salt crystal morphology were analyzed to obtain the contact resistance change trend and voltage fluctuation characteristics. The voltage drop amplitude and fluctuation frequency were calculated based on the voltage fluctuation characteristics.

[0007] Based on the voltage drop amplitude and voltage fluctuation frequency, combined with the charging process transmission loss calculated from impedance detection data, as well as the salt deposition concentration and temperature fluctuation range, the abnormal voltage signal characteristics of the charging interface in a high-salt and humid environment are extracted, thereby determining the distribution pattern of the abnormal voltage signal;

[0008] Analyze the voltage drop amplitude based on the distribution law of the voltage abnormality signal to obtain the potential downward trend of the battery capacity retention rate, and determine the correlation data between the battery capacity attenuation and the distribution law of the voltage abnormality signal based on the potential downward trend of the battery capacity retention rate;

[0009] Based on the distribution characteristics of salt deposition concentration and salt crystal morphology, the correlation between material surface roughness and protective coating durability and the accumulation of battery electrode material deposits is analyzed. Based on this correlation, the material corrosion correlation characteristics of battery capacity decay are determined;

[0010] Based on the correlation characteristics, moisture permeation rate and coating durability, a battery capacity attenuation evaluation system is constructed. The uneven contact resistance distribution and current loss are input, the capacity attenuation trend is output and a curve is drawn, and the battery capacity is predicted based on the curve.

[0011] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0012] The present invention discloses a method for predicting battery capacity attenuation during charge and discharge cycles. The method collects high-salt and humid environment data in real time, monitors changes in the contact resistance of the charging interface, analyzes salt deposition and crystallization distribution, and identifies voltage anomaly signal characteristics. A battery capacity attenuation evaluation system is established by combining the surface characteristics of the material and the durability of the protective coating. The present invention integrates multi-dimensional data, including moisture penetration rate, salt crystallization morphology, voltage fluctuation characteristics, etc., to analyze the correlation between voltage anomalies and capacity attenuation. By constructing an evaluation model and inputting contact resistance distribution and current loss data, the battery capacity attenuation trend can be predicted, providing a new technical solution for battery performance evaluation and life prediction in high-salt and humid environments, which helps to improve the reliability and service life of battery systems in harsh environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The present invention is a flow chart of a method for predicting battery capacity attenuation during charge and discharge cycles.

[0014] Figure 2 Schematic diagram of a method for predicting battery capacity attenuation during charge and discharge cycles according to the present invention. DETAILED DESCRIPTION

[0015] To help those skilled in the art better understand the technical solutions in this specification, the following will provide a clear and complete description of the technical solutions in the embodiments of this specification, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. All other embodiments derived by those skilled in the art based on the embodiments in this specification without creative effort shall fall within the scope of protection of this specification.

[0016] like Figure 1-2 In this embodiment, a method for predicting battery capacity attenuation during charge and discharge cycles may specifically include:

[0017] Step S101 collects high-salt and humid environment data in real time, calculates the moisture penetration rate and residence time, analyzes the salt deposition concentration and crystallization distribution based on the surface roughness of the charging interface material and the durability of the protective coating, and identifies the surface contact resistance increase and material corrosion characteristics.

[0018] Real-time temperature, humidity, and air pressure data are collected around the charging interface. The concentration of salt particles, such as sodium chloride and sodium sulfate, in the air is also measured. The water vapor partial pressure is calculated based on the temperature and humidity data. The salt spray density coefficient is obtained by multiplying the salt spray density coefficient by the ambient humidity gradient to determine the moisture permeation rate. Based on the determined moisture permeation rate and the porosity and surface roughness parameter Ra of the charging interface material, the percolation channel characteristic value is calculated from the ratio of the cross-sectional area to the depth of the microscopic grooves on the material surface. The average residence time of the salt solution in the grooves is calculated from the ratio of the percolation channel characteristic value to the moisture permeation rate, thereby determining the salt residence time distribution in different regions of the charging interface. Based on the obtained salt residence time distribution and the ambient temperature and humidity curve, the rate of change of the salt solution concentration is calculated when the temperature rises and water evaporates. Crystallization occurs when the solution concentration exceeds the saturation concentration. The amount of crystallization in each region is determined by multiplying the residence time by the rate of change of concentration, thus obtaining a salt deposition concentration distribution map on the charging interface surface. The contact resistance values ​​of different areas on the charging interface surface are measured using the four-probe method. Based on the concentration values ​​at each point in the salt deposition concentration distribution diagram, the percentage increase in contact resistance relative to the initial measurement value is calculated. If the increase in contact resistance in a certain area exceeds a preset threshold, the protective coating in that area is judged to have failed. The material corrosion characteristics of the charging interface are identified by combining the deposition concentration and resistance increase data.

[0019] Specifically, in a high-salt and humid environment, corrosion monitoring of the charging interface needs to start with environmental data collection.

[0020] Specifically, environmental monitoring sensors need to simultaneously measure three basic parameters: temperature, humidity, and air pressure. Temperature sensors typically use platinum resistance thermometers with a measurement range of -40°C to 85°C. Humidity sensors use capacitive hygrometers, capable of maintaining accurate measurements within a relative humidity range of 10% to 95%. Salt particle concentration is measured using laser scattering. When a laser beam passes through air containing salt particles, the particles scatter the light. By analyzing the scattered light intensity distribution, the concentration distribution of salt particles of different sizes, such as sodium chloride and sodium sulfate, can be determined.

[0021] It should be noted that the calculation process of moisture permeability involves the determination of water vapor partial pressure. When the ambient temperature is 25°C and the relative humidity is 80%, the saturated water vapor pressure at this temperature is first queried to be 3.17 kPa. Multiplying it by the relative humidity of 0.8 gives the actual water vapor partial pressure of 2.54 kPa. The salt spray density coefficient is obtained by dividing the total mass concentration of salt particles by the water vapor partial pressure. If the measured salt concentration is 35 mg / m 3 , then the salt spray density coefficient is 13.8 mg·m -3 kPa -1 The ambient humidity gradient is obtained by dividing the humidity difference between two adjacent measuring points by the distance. The moisture penetration rate calculated in this way can reflect the speed at which salt mist moves to the surface of the charging interface.

[0022] In one possible implementation, the microstructure of the surface of the charging interface material has an important influence on salt retention. The surface roughness parameter Ra represents the arithmetic mean deviation of the profile. When the Ra value is 1.6μm, it means that there are a large number of micron-sized grooves and protrusions on the surface. The calculation of the characteristic value of the seepage channel needs to take into account the geometry of these grooves. Through microscopic observation, it was found that the cross-section of the groove is approximately V-shaped, with an average width of 10μm and a depth of 5μm. The ratio of cross-sectional area to depth is the characteristic value of the seepage channel, 2μm. The flow rate of the salt solution in these tiny channels is much lower than the flow rate on the smooth surface, resulting in a significant extension of the retention time.

[0023] Preferably, the crystallization process is monitored using real-time microscopy. When the ambient temperature rises from 25°C to 35°C, the water in the salt solution in the groove evaporates rapidly, and the concentration gradually increases from the initial 3.5%. The saturation concentration is about 26%. Once this critical value is exceeded, sodium chloride crystals begin to precipitate at the bottom of the groove, forming cubic crystals. The longer the residence time in the area, the greater the amount of crystallization. By using image processing technology to count the number and size of crystals per unit area, a salt deposition concentration distribution map of the entire charging interface surface can be drawn to provide data support for subsequent corrosion assessments.

[0024] For example, when the four-probe method is used to measure contact resistance, four probes are arranged at equal intervals on the surface of the charging interface. A constant current is passed through the two outer probes, and the voltage drop is measured by the two inner probes. The contact resistance is measured to be 10mΩ in the initial state. After salt spray corrosion, the contact resistance in the high-concentration salt deposition area increases to 15mΩ, an increase of 50%. This increase in resistance is mainly caused by two reasons: first, the insulating properties of the salt crystals themselves increase the contact resistance; second, the salt promotes electrochemical corrosion of the metal surface, and the generated corrosion products further deteriorate the conductive properties. By establishing a corresponding relationship between deposition concentration and resistance increase, the area where the protective coating fails can be accurately identified, providing precise guidance for the maintenance of the charging interface.

[0025] Step S102 : continuously monitoring the contact resistance increase value of the interface to obtain contact resistance uneven distribution data, and determining the time point and amplitude of the abnormal resistance increase based on the uneven data.

[0026] The four-probe method is used to perform grid resistance measurement on the surface of the charging interface. The contact resistance value of each measuring point is obtained according to the preset sampling interval. The local resistance gradient is obtained by dividing the difference between the resistance values ​​of adjacent measuring points by the distance between the measuring points. The resistance distribution unevenness coefficient is determined by taking the square root of the sum of the squares of the deviations of the resistance values ​​of all measuring points from the average value and dividing it by the number of measuring points to obtain the contact resistance distribution unevenness data. Based on the obtained contact resistance distribution unevenness data, when the unevenness coefficient exceeds the preset threshold, the resistance value sequence of each measuring point at that moment is compared point by point, and the difference between the resistance value at the current moment and the resistance value at the previous sampling moment is calculated. If the difference exceeds the preset change threshold, the moment is recorded as the time point of abnormal resistance increase. According to the recorded time point of abnormal resistance increase, the resistance measurement value at that time point and the initial resistance value at the start of continuous monitoring are extracted. The magnitude of the abnormal resistance increase is determined by subtracting the initial resistance value from the resistance value at that time point and dividing it by the initial resistance value.

[0027] Specifically, the four-probe method is a classic method for contact resistance measurement, and its working principle is based on the separate measurement of current and voltage.

[0028] Specifically, four probes are arranged in a straight line with equal spacing on the surface of the charging interface. A constant DC current is applied to the two outer probes, while the two inner probes measure the resulting voltage drop. This configuration eliminates the influence of contact resistance between the probes and the sample, making the measurement results more accurate. Grid measurement involves creating a virtual grid on the surface of the charging interface, with each grid intersection serving as a measurement point. The probe group then moves to each measurement point in turn for measurement.

[0029] In one possible implementation, the charging interface surface is divided into a 5×5 measurement grid with 25 measurement points. After recording the contact resistance value at each measurement point, the resistance gradient between adjacent measurement points is calculated. Assuming the resistance value at measurement point A is 12mΩ, the resistance value at the adjacent measurement point B is 15mΩ, and the distance between the two points is 2mm, the local resistance gradient is 1.5mΩ / mm. This gradient reflects the rate of change of resistance over space; a larger gradient indicates a more uneven resistance distribution in that area.

[0030] It should be noted that the calculation of the resistance distribution nonuniformity coefficient involves statistical principles. First, calculate the average resistance value of all 25 measurement points, assuming it is 13mΩ. Then, calculate the deviation of the resistance value of each measurement point from the average value. Add the squares of all deviations, divide by the total number of measurement points (25), and take the square root to obtain the standard deviation. If the standard deviation is 2.5mΩ, the resistance distribution nonuniformity coefficient can be defined as the ratio of the standard deviation to the average value, which is 0.192. A larger coefficient indicates a more nonuniform resistance distribution on the interface surface, indicating the presence of localized corrosion or contamination.

[0031] Optimally, the sampling interval setting is crucial during continuous monitoring. Too short a sampling interval will generate a large amount of redundant data, while too long an interval may miss rapidly changing anomalies. In practice, a 5-minute sampling interval is typically used for routine monitoring. When the nonuniformity coefficient exceeds the preset threshold of 0.2, the system enters intensive monitoring mode, shortening the sampling interval to 30 seconds. This dynamic adjustment ensures timely monitoring while avoiding the pressure of data storage and processing.

[0032] For example, during one monitoring session, the nonuniformity coefficient suddenly increased from 0.15 to 0.25 at 2:30 PM, triggering intensive monitoring. A point-by-point comparison revealed that the resistance value at measuring point C7 jumped from 10 mΩ to 18 mΩ at 2:32 PM, a change of 8 mΩ, exceeding the preset change threshold of 3 mΩ. The system immediately recorded 2:32 PM as the time of the abnormal resistance increase at this measuring point. This precise time location provides important information for subsequent fault analysis and preventive maintenance.

[0033] In one embodiment, the calculation of the abnormal increase in resistance is expressed in relative values. The initial resistance value of measuring point C7 at the beginning of monitoring is 10mΩ, and the resistance value at the abnormal moment is 18mΩ. The amplitude is calculated as follows: 18 minus 10 and then divided by 10, which is 0.8, or an increase of 80%. This percentage expression method facilitates horizontal comparison between different measuring points and is also convenient for setting a unified alarm threshold. When the resistance increase of multiple measuring points exceeds 50%, it usually indicates that the charging interface has a serious overall corrosion problem and requires timely maintenance.

[0034] Step S103 , integrating the contact resistance increase and distribution unevenness data with the high-salt humid environment data, analyzing the moisture retention time and salt crystal form, obtaining the contact resistance change trend and voltage fluctuation characteristics, and calculating the voltage drop amplitude and fluctuation frequency based on the voltage fluctuation characteristics.

[0035] The system obtains contact resistance increase data, distribution unevenness data, ambient humidity values, and salt concentration data. The resistance increase value is multiplied by the humidity value at the corresponding moment to obtain the moisture impact factor. The crystallization tendency coefficient is calculated based on the ratio of salt concentration to saturation concentration. The moisture residence time is determined by dividing the moisture impact factor by the crystallization tendency coefficient. An optical microscope is used to observe the salt crystal morphology and identify the distribution of cubic or needle-shaped morphologies. Based on the identified crystal morphology distribution and residence time, the average resistance increase rate of the cubic crystal region and the average resistance increase rate of the needle-shaped crystal region are calculated. A weighted average method is used to calculate the combined resistance increase rate based on the area ratio of the cubic and needle-shaped morphologies. A contact resistance trend curve is obtained through time series fitting, and the slope of the trend curve is determined. Based on the slope of the contact resistance trend curve, voltage monitoring is triggered when the slope exceeds a preset threshold. Real-time voltage data is collected across the charging port to form a voltage time series. The voltage difference between adjacent sampling points is calculated to identify fluctuation events. The voltage drop amplitude is calculated by dividing the difference between the current voltage and the voltage at the start of monitoring by the starting voltage. The voltage fluctuation frequency is determined by counting the number of fluctuation events per unit time.

[0036] Specifically, the calculation of the moisture impact factor reflects the amplification effect of ambient humidity on resistance change.

[0037] In one possible implementation, when the contact resistance increases by 50% and the relative humidity at the corresponding moment is 85%, the moisture impact factor is calculated as 0.5 multiplied by 0.85, which is 0.425. This value represents the degree of resistance deterioration in a humid environment. The larger the value, the more significant the impact of moisture on resistance. The crystallization tendency coefficient is measured by the ratio of salt concentration to saturation concentration. When the salt concentration in the environment reaches 20g / L and the saturation concentration at this temperature is 26g / L, the crystallization tendency coefficient is 0.77, indicating that the solution is close to saturation and is prone to crystallization.

[0038] It should be noted that the determination of moisture residence time is crucial for subsequent analysis. By dividing the moisture impact factor of 0.425 by the crystallization tendency coefficient of 0.77, the relative residence time value of 0.55 is obtained. This relative value needs to be multiplied by the reference time constant to obtain the actual residence time. The longer the residence time, the more salt accumulates on the interface surface and the greater the risk of corrosion. Optical microscopy observations found that cubic sodium chloride crystals mainly appear deep in the grooves, while needle-shaped crystals are mostly distributed on the surface protrusions. This distribution difference is related to local humidity and evaporation rate.

[0039] Specifically, the resistance increase rate needs to be calculated by region. The cubic crystallization region, due to its large crystal volume and dense packing, has an average resistance increase rate of 8% per hour. The needle-shaped crystallization region, despite having a large number of crystals but a small volume, has an average resistance increase rate of 5% per hour. When calculating the weighted average, assuming that the cubic crystallization region accounts for 40% of the total area and the needle-shaped crystallization region accounts for 60%, the combined resistance increase rate is 0.4 × 8% + 0.6 × 5% = 6.2% per hour. This combined rate value directly reflects the corrosion development speed of the entire interface.

[0040] In one embodiment, time series fitting uses the least squares method to process 24 hours of continuous resistance measurement data. Data is collected once an hour, resulting in 24 data points. The resulting trend curve exhibits a near-linear growth pattern with a slope of 0.15 mΩ / h. When the slope exceeds a preset threshold of 0.1 mΩ / h, the system determines that the corrosion rate has reached a level requiring attention and immediately initiates voltage monitoring.

[0041] For example, after voltage monitoring is enabled, the voltage values ​​across the charging interface are collected once per second. During normal charging, the voltage is stable at around 5V, but when the contact resistance increases, resulting in an increased voltage drop, the measured voltage will fluctuate. In one monitoring session, the starting voltage was 5.0V, and after two hours of operation, the voltage dropped to 4.7V. The voltage drop was calculated as 0.3V divided by 5.0V, resulting in a 6% drop rate. Furthermore, during these two hours, 15 incidents of instantaneous voltage fluctuations exceeding 0.1V were detected, an average of once every eight minutes, or a fluctuation frequency of 7.5 times per hour.

[0042] It's important to note that the frequency of voltage fluctuations is closely related to the degree of poor contact. A higher frequency indicates more unstable contact, likely due to corrosion products causing alternating conduction and insulation under mechanical vibration. By comprehensively analyzing the voltage drop amplitude and fluctuation frequency, we can accurately assess the health of the charging port, providing a quantitative basis for preventive maintenance.

[0043] In step S104, based on the voltage drop amplitude and voltage fluctuation frequency, combined with the charging process transmission loss calculated from the impedance detection data, as well as the salt deposition concentration and temperature fluctuation range, the voltage abnormality signal characteristics of the charging interface in a high-salt and humid environment are extracted to determine the distribution pattern of the voltage abnormality signal.

[0044] The voltage drop amplitude, fluctuation frequency data, and charging circuit impedance are collected. The transmission loss power is calculated by multiplying the impedance value by the square of the charging current. The ratio of the transmission loss power to the rated transmission power is used as the loss rate. Combined with salt deposition concentration and temperature fluctuation range data, the environmental impact comprehensive index is calculated by weighted summing the loss rate, normalized salt concentration, and normalized temperature fluctuation range. For periods where the environmental impact comprehensive index exceeds a preset threshold, voltage time series data is collected during these periods. The stable voltage value at the start of charging is set as the baseline value. Data segments where the voltage deviates from the baseline by an allowable deviation are identified as abnormal segments. The maximum voltage deviation within the abnormal segment is extracted as the abnormal signal amplitude. The duration of the abnormal segment is recorded, and the rate of change is calculated by dividing the amplitude by the duration to obtain the abnormal signal feature vector. Based on the obtained abnormal signal feature vector, three-dimensional statistical histograms are constructed based on amplitude, duration, and rate of change. The main distribution range of each feature is determined by the position of the histogram peak. The three-dimensional feature space is divided into several subregions. The frequency and time interval of abnormal signal occurrence in each subregion are counted to determine the distribution pattern of the voltage abnormal signal in the feature space.

[0045] Specifically, the calculation of transmission loss power involves the analysis of the electrical characteristics of the charging circuit.

[0046] In one possible implementation, an impedance analyzer is used to measure the complex impedance of the charging port, including the real resistance and imaginary reactance. When the charging current is 2A, if the measured total impedance is 50mΩ, the transmission loss power is the current squared multiplied by the real part of the impedance, that is, 4×0.05=0.2W. The rated transmission power is determined by the charger specifications. For example, the rated transmission power of a 10W charger is 10W, and the loss rate is calculated as

[0047] 0.2 / 10=0.02, which is a 2% power loss.

[0048] It should be noted that the comprehensive environmental impact index is calculated using a weighted sum method to reflect the combined effects of multiple environmental factors. Normalization of salt concentration is performed by dividing the measured concentration by the reference concentration of 35 g / L to obtain a normalized value between 0 and 1. Normalization of temperature fluctuation amplitude is performed by dividing the daily temperature difference by 30°C. Assuming a weight of 0.4 for the loss rate, 0.4 for the normalized salt concentration value, and 0.2 for the normalized temperature fluctuation amplitude value, when the three are 0.02, 0.6, and 0.5, respectively, the comprehensive environmental impact index is calculated as 0.4 × 0.02 + 0.4 × 0.6 + 0.2 × 0.5 = 0.348.

[0049] Specifically, the setting of the reference value is crucial. Within the first 10 seconds after charging begins, the system continuously collects voltage data and calculates the average value, and uses this stable average value as the reference value. The allowable deviation is usually set at 2% of the reference value. If the reference value is 5V, the allowable deviation is 0.1V. When the measured voltage is lower than 4.9V or higher than 5.1V, it is judged to be abnormal. The abnormal signal feature vector contains three components. The amplitude indicates the maximum deviation, such as a deviation of 0.3V; the duration indicates how long the abnormality lasts, such as 5 seconds; and the rate of change is 0.3V / 5s=0.06V / s.

[0050] In one embodiment, the establishment of a statistical histogram requires a large amount of data support. After one week of continuous monitoring, 200 sets of abnormal signal data were collected. The amplitude histogram shows that the 0.1-0.2V range has the highest frequency, reaching 80 times; the 0.0.3V range has 60 times; and those exceeding 0.3V have only 20 times. The duration histogram shows that short-term anomalies of 1-3 seconds account for 50%, medium anomalies of 3-10 seconds account for 35%, and long-term anomalies exceeding 10 seconds account for 15%. The change rate histogram shows a bimodal distribution, with slow changes and rapid mutations each accounting for a certain proportion.

[0051] For example, the three-dimensional feature space is divided into eight subregions, with each dimension divided into two levels of high and low. A combination of low amplitude, short duration, and slow rate corresponds to mild contact failure, typically seen in the early stages of charging; a combination of high amplitude, long duration, and slow rate indicates severe corrosion, often seen in interfaces after several months of use. By calculating the abnormal signal density in each subregion, it was found that 70% of the anomalies were concentrated in three specific subregions, showing a clear clustering pattern.

[0052] It's important to note that time interval analysis reveals a periodic pattern in abnormal signals. Statistics on the time intervals between adjacent abnormal events show that 40% of them are between 1 and 5 minutes, indicating short-term repetitiveness. 30% are between 1 and 2 hours, potentially related to diurnal temperature fluctuations. And 30% are between 24 hours and longer, reflecting the progressive development of corrosion. This distribution pattern provides an important basis for predictive maintenance, enabling maintenance personnel to take action before anomalies develop into failures.

[0053] Step S105 , analyzing the voltage drop amplitude according to the distribution law of the voltage abnormality signal to obtain the potential downward trend of the battery capacity retention rate, and determining the correlation data between the battery capacity attenuation and the distribution law of the voltage abnormality signal according to the potential downward trend of the battery capacity retention rate.

[0054] Based on the distribution of abnormal voltage signals, the cumulative value of the voltage drop amplitude during periods of high-frequency abnormal signal occurrence is extracted. This cumulative value is divided by the total number of corresponding charging cycles to obtain the average voltage drop per cycle. The voltage decay ratio is calculated by the ratio of the average voltage drop per cycle to the battery's nominal voltage. The voltage decay ratio for each charging cycle is recorded to form time series data. At each time point in the generated time series data, the actual discharge capacity of the battery at the end of the corresponding charging cycle is measured. The capacity retention rate is calculated by the ratio of the actual discharge capacity to the battery's rated capacity. A least squares method is used to fit a curve representing the relationship between the capacity retention rate and the number of charging cycles. The absolute value of the slope of the fitted curve is used to determine the rate of decrease in the battery capacity retention rate. Based on the determined rate of decrease in the capacity retention rate, three characteristic values ​​of the voltage abnormal signal within each charging cycle are extracted: amplitude, duration, and frequency of occurrence. The correlation coefficient between each characteristic value and the capacity retention rate decrease in the corresponding cycle is calculated. The abnormal feature strength index is established based on features with a correlation coefficient greater than a preset threshold. The abnormal feature strength index is then regressed against the capacity retention rate decrease rate to determine the correlation data between battery capacity decay and the distribution of abnormal voltage signals.

[0055] Specifically, the analysis of the distribution patterns of abnormal voltage signals requires an in-depth understanding from the time dimension.

[0056] In one possible implementation, continuous monitoring revealed that abnormal signals were most frequent at the charging port in a high-salt, humid environment at 4 a.m. and 2-4 p.m., which is closely related to the daily variation in ambient temperature and humidity. After compiling a month's worth of data, it was found that the voltage drop during these high-frequency periods reached a cumulative 0.8V. Dividing 0.8V by 30 days (a total of 120 charging cycles) yields an average voltage drop of 6.7mV per cycle. For a charging system with a nominal voltage of 5V, the voltage attenuation ratio is calculated to be 0.0067 / 5 = 0.134% per cycle.

[0057] It should be noted that the actual discharge capacity of the battery is measured using the constant current discharge method. After each charge cycle, the battery is discharged at a constant current rate of 0.2C to the cutoff voltage, and the discharged capacity is integrated to calculate the actual capacity. Assuming a battery with a rated capacity of 3000mAh, the actual capacity measured in the first cycle is 2970mAh, with a capacity retention rate of 99%. The actual capacity measured in the 100th cycle is 2820mAh, with a capacity retention rate of 94%. By collecting data points for 100 cycles, a data set is formed to show how the capacity retention rate varies with the charge cycle.

[0058] Specifically, the least squares fitting process seeks a straight line that minimizes the sum of the squared vertical distances from all data points to the line. Calculations show a slope of -0.05% / cycle, indicating a 0.05% decrease in capacity retention per charge cycle. This rate of decline reflects the accelerated aging of batteries in a high-salt, humid environment, representing a 2.5-fold deterioration compared to the 0.02% / cycle decay rate under normal conditions.

[0059] In one embodiment, the three key feature value extraction processes for the voltage anomaly signal each have their own focus. The anomaly amplitude is obtained by recording the maximum degree to which the voltage deviates from the normal value during each anomaly event. For example, in one case, the voltage dropped to 4.7V with an amplitude of 0.3V. The duration counts the time from when the voltage deviates to when it returns to normal, typically ranging from 1 to 10 seconds. The frequency of occurrence counts the total number of anomalies during each charging cycle; for interfaces with high corrosion levels, this could reach 20 to 30 occurrences per cycle.

[0060] For example, the calculation of correlation coefficients reveals the degree of influence of different features on capacity decay. The correlation coefficient between abnormal amplitude and capacity drop reached 0.82, indicating a strong correlation; the correlation coefficient between duration was 0.65, indicating a moderate correlation; and the correlation coefficient between frequency of occurrence was 0.73. Setting the correlation coefficient threshold to 0.6, all three features were included in the calculation of the abnormal feature strength index. This index is obtained by weighted summation, with weights of 0.45, 0.25, and 0.30, respectively, reflecting the relative importance of each feature.

[0061] It should be noted that the regression analysis established a quantitative relationship between the abnormal characteristic intensity index and the capacity decay rate. By performing linear regression on 50 sets of data, the determination coefficient R of the regression equation was obtained. 2 The regression coefficient shows that for every 0.1 increase in the abnormal feature intensity index, the capacity decay rate increases by 0.008% per cycle. This correlation data provides a quantitative basis for assessing the impact of charging interface corrosion on battery life, making the development of preventive maintenance strategies more accurate and effective.

[0062] Step S106 , analyzing the correlation between the material surface roughness and the durability of the protective coating and the accumulation of battery electrode material deposits based on the distribution characteristics of the salt deposition concentration and the salt crystal morphology, and determining the material corrosion correlation characteristics of the battery capacity decay based on the correlation relationship.

[0063] Electron microscopy scans are used to obtain microscopic image data of the battery electrode material surface. Salt deposits are identified based on changes in the image's grayscale value. The geometric dimensions and distribution density of each deposit are measured, and the volume concentration of the salt deposits per unit area is calculated. For areas where the volume concentration of salt deposits exceeds a preset threshold, X-ray diffractometers are used to measure the crystal structure parameters of the deposits in these areas. The position and intensity of the diffraction peaks are used to determine the type of salt crystal. Sharp diffraction peaks indicate a cubic structure, while broadened peaks indicate an amorphous structure. This generates a distribution map of the salt crystal morphology. Based on the distribution map, atomic force microscopy is used to measure surface height fluctuations in areas with cubic and amorphous structures, and the surface roughness parameter Ra is calculated. Electrochemical impedance spectroscopy is also used to determine the impedance modulus Z of the protective coating in these areas. The corresponding Ra and Z values ​​for different crystal morphologies are recorded. Based on the correspondence between Ra value and Z value, linear regression analysis was used to determine the correlation between the corrosion rate V, which is directly proportional to the Ra value and inversely proportional to the Z value. The corrosion rate value was obtained. The deposition thickness was calculated based on the corrosion rate and the volume concentration of salt deposits. The length and number of cracks in the surface microscopic image were identified through image processing to determine the degree of surface damage. The material corrosion correlation characteristics including corrosion rate, deposition thickness and surface damage degree were obtained.

[0064] Specifically, electron microscopy scanning technology plays a key role in the surface analysis of battery electrode materials.

[0065] Specifically, microscopic images obtained through a scanning electron microscope can clearly show the morphological characteristics of salt deposits. Changes in the grayscale value of the image reflect the surface undulations and compositional differences of the material. Salt deposit areas exhibit different grayscale characteristics due to differences in electron density with the substrate material. This grayscale recognition method can accurately locate the boundaries of each deposit area and then measure its geometric dimensions. By summing the volumes of all deposit areas per unit area, the volume concentration of the salt deposit can be calculated. This value directly reflects the degree of salt accumulation on the electrode surface.

[0066] In one possible implementation, when the volume concentration of salt deposits in a certain area exceeds a preset threshold, it indicates that a relatively serious accumulation of salt has occurred in that area. At this time, it becomes particularly important to use X-ray diffraction technology for in-depth analysis. The X-ray diffractometer determines the crystal structure by emitting X-rays of a specific wavelength to the sample based on the diffraction phenomenon produced by Bragg's law. The position of the diffraction peak corresponds to the interplanar spacing, while the intensity of the peak reflects the atomic arrangement density of the crystal plane. Sharp diffraction peaks indicate that the atomic arrangement is highly ordered, corresponding to regular structures such as the cubic crystal system; while broadened diffraction peaks indicate that the atomic arrangement is highly disordered, corresponding to an amorphous structure. This difference in crystal form directly affects the mechanical properties and chemical activity of salt deposits.

[0067] It should be noted that there are significant differences in the effects of salt deposits of different crystal forms on the surface of electrode materials. Salts with cubic crystal structures tend to form relatively hard deposits due to their regular atomic arrangement, while amorphous structures are relatively loose. Atomic force microscopy can measure the surface morphology of these different areas with nanometer-level precision. The surface roughness parameter Ra value is obtained by calculating the arithmetic mean deviation of the surface height relative to the average line. The larger the value, the rougher the surface. The electrochemical impedance spectroscopy test carried out at the same time measures the impedance response of the protective coating by applying an alternating current signal. The impedance modulus Z reflects the ability of the coating to hinder charge transfer and is an important indicator for evaluating the integrity of the coating.

[0068] Preferably, a quantitative relationship between the corrosion rate and the surface roughness and coating impedance is established through linear regression analysis. This analysis method is based on a large amount of experimental data and found that the corrosion rate is positively correlated with the surface roughness and negatively correlated with the coating impedance. This law is consistent with the principle of corrosion dynamics: the rough surface provides more active sites, accelerating the corrosion reaction; while the complete coating forms an effective barrier to slow down the corrosion process. Based on this correlation, combined with the volume concentration value of the salt deposit, the growth trend of the deposition thickness can be accurately calculated. The degree of surface damage is obtained by identifying crack characteristics through image processing technology. The length and number of cracks directly reflect the degree of material deterioration. These parameters together constitute the material corrosion correlation characteristics, providing a reliable basis for evaluating battery capacity attenuation.

[0069] Step S107: Based on the correlation characteristics, moisture permeation rate and coating durability, a battery capacity attenuation evaluation system is constructed, the uneven contact resistance distribution and current loss are input, the capacity attenuation trend is output and a curve is drawn, and the battery capacity is predicted based on the curve.

[0070] The corrosion rate, deposition thickness, and surface damage values ​​associated with material corrosion were obtained. A humidity sensor was used to measure the moisture permeation rate inside the battery package. Coating durability parameters were extracted based on electrochemical impedance spectroscopy test results. These three parameters were normalized using a normalization process to obtain a basic evaluation parameter set. Based on the corrosion rate and surface damage values ​​in the basic evaluation parameter set, the locations of measurement points on the electrode surface were determined. The contact resistance at each point was measured using a four-probe method. The standard deviation of the resistance values ​​across all measurement points was calculated as the contact resistance distribution nonuniformity. An ammeter was used to measure the difference between the actual and theoretical current values ​​during charge and discharge to obtain the current loss rate. A multivariate linear regression method was used, using the basic evaluation parameter set, contact resistance distribution nonuniformity, and current loss rate as independent variables, and the measured battery capacity decay values ​​as the dependent variable. The regression coefficients were calculated using the least squares method to establish a quantitative relationship between capacity decay and each parameter. Based on this quantitative relationship, a capacity decay trend data series was calculated over time intervals. According to the time points and corresponding capacity decay values ​​in the capacity decay trend data series, a time and capacity decay percentage curve is drawn. The curve is fitted using an exponential fitting method to obtain a decay trend function. The remaining capacity of the battery at a future time point is predicted based on the decay trend function and the initial capacity value of the battery.

[0071] Specifically, the material corrosion correlation characteristics serve as the basic input of the evaluation system, including three key parameters: corrosion rate, deposition thickness and surface damage degree.

[0072] Specifically, the corrosion rate reflects the rate of degradation of electrode materials under specific conditions and is typically expressed as the reduction in material thickness per unit time. Deposition thickness represents the accumulation of corrosion products, such as salts, on the electrode surface. These deposits can obstruct electron transport. The extent of surface damage is quantified by crack density and depth, which directly impacts the effective contact area of ​​the electrode.

[0073] In one possible implementation, moisture permeation rate is measured using a high-precision humidity sensor to monitor the internal battery environment in real time. Moisture penetrates the battery through tiny gaps in the packaging material, reacting with the electrolyte to produce corrosive substances. The higher the permeation rate, the more severe the corrosion reaction, and the more significant the impact on battery performance. Coating durability parameters are obtained through electrochemical impedance spectroscopy testing, which reflects the protective coating's ability to resist erosion by corrosive media. Normalization converts these parameters of different dimensions to a unified numerical range to facilitate subsequent comprehensive evaluation.

[0074] It should be noted that the principle of the four-probe method for measuring contact resistance is to apply a constant current through the two outer probes, measure the voltage drop through the two inner probes, and calculate the resistance value according to Ohm's law. Multiple measurement points are selected on the electrode surface to form a resistance distribution map. The standard deviation is calculated by first finding the average value of the resistance values ​​of all measurement points, then calculating the sum of the squares of the differences between each point and the average value, dividing it by the number of measurement points and taking the square root. The larger the standard deviation, the more uneven the contact resistance distribution. Local high-resistance areas will cause current concentration and accelerate aging in that area. The current loss rate is determined by comparing the theoretical current value with the actual measured value. The theoretical current value is calculated based on the rated capacity and discharge rate of the battery, and the actual current value is directly measured by a high-precision ammeter. The loss mainly comes from contact resistance, increased internal resistance and loss of active materials. These parameters together reflect the health of the battery.

[0075] Preferably, the multivariate linear regression method establishes a quantitative relationship between each parameter and capacity decay. This method assumes that capacity decay is linearly related to each influencing factor, and minimizes the sum of squared errors between the predicted value and the measured value through the least squares method to determine the regression coefficient. Each regression coefficient represents the contribution weight of the corresponding parameter to the capacity decay. Based on the established quantitative relationship, the parameter change value is substituted according to the set time interval to calculate the capacity decay trend data sequence. The exponential fitting method is particularly suitable for describing the decay law of battery capacity, because battery aging usually shows the characteristics of rapid decay in the early stage and flattening in the later stage. The general form of the fitting function is that the capacity retention rate decreases exponentially with time, and the decay coefficient in the function is determined by historical data points. With the decay trend function, combined with the initial capacity value of the battery, the remaining capacity at any future time point can be predicted, providing a scientific basis for battery maintenance and replacement.

[0076] It should be noted that the above examples are merely specific embodiments of the present invention. Obviously, the present invention is not limited to the above examples and is subject to numerous variations. All variations that can be directly derived or conceived by a person skilled in the art from the disclosure of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for predicting battery capacity attenuation during charge and discharge cycles, characterized in that: The method comprises: Collect high-salt, humid environment data, calculate moisture penetration rate and residence time, combine the surface roughness of the charging interface material and the parameters of the protective coating, analyze the salt deposition concentration and crystallization distribution, and determine the surface contact resistance increase and material corrosion distribution; Monitor the contact resistance increase value of the charging interface, obtain contact resistance uneven distribution data, and determine the time point and amplitude of abnormal resistance increase based on the contact resistance uneven distribution data; The contact resistance increase value and the contact resistance uneven distribution data are integrated with the high-salt humidity environment data to analyze the moisture retention time and salt crystal morphology to determine the contact resistance change trend and voltage fluctuation parameters; Extracting a voltage abnormality signal distribution of the charging interface in a high-salt and humid environment based on the voltage fluctuation parameter and the charging process transmission loss calculated using the impedance detection data, as well as the salt deposition concentration and temperature fluctuation data; Analyze the voltage drop amplitude according to the voltage abnormality signal distribution to determine the battery capacity retention rate decline trend; Analyzing the relationship between the surface roughness of the charging interface material and the protective coating parameters and the accumulation of electrode material deposits based on the salt deposition concentration and salt crystal morphology distribution to determine the material corrosion distribution of the battery capacity decay; Based on the material corrosion distribution, the moisture permeation rate and the protective coating parameters, a battery capacity decay evaluation system is constructed, the contact resistance uneven distribution data and the current loss data are input, and a capacity decay trend curve is output. The battery capacity is predicted according to the capacity decay trend curve.

2. The method for predicting battery capacity attenuation during charge and discharge cycles according to claim 1, wherein: The method involves collecting high-salt and humid environment data, calculating the moisture penetration rate and residence time, analyzing the salt deposition concentration and crystallization distribution based on the surface roughness of the charging interface material and the parameters of the protective coating, and determining the surface contact resistance increase and material corrosion distribution. The method collects temperature, humidity, and salt particle concentration data from the environment surrounding the charging interface, calculates the water vapor partial pressure based on the temperature and humidity data, determines the salt mist density coefficient based on the ratio of the salt particle concentration to the water vapor partial pressure, and calculates the moisture permeation rate based on the salt mist density coefficient and the ambient humidity gradient; calculates the seepage channel characteristic value of the material surface microstructure based on the moisture permeation rate and the surface roughness parameter of the charging interface material, and determines the salt solution residence time distribution based on the ratio of the seepage channel characteristic value to the moisture permeation rate; calculates the salt solution concentration change rate based on the salt solution residence time distribution and changes in ambient temperature and humidity, determines the crystallization amount in each region based on the salt solution concentration change rate and the residence time distribution, and obtains the salt deposition concentration distribution; measures the contact resistance value of the charging interface surface based on the salt deposition concentration distribution, calculates the increase in the contact resistance value relative to the initial value, and determines the material corrosion distribution based on the salt deposition concentration distribution.

3. The method for predicting battery capacity attenuation during charge and discharge cycles according to claim 1, wherein: The monitoring of the contact resistance increase value of the charging interface, obtaining contact resistance uneven distribution data, and determining the time point and amplitude of abnormal resistance increase according to the contact resistance uneven distribution data include: Measure the contact resistance value of each area on the surface of the charging interface, calculate the deviation of the contact resistance value from the average value, and determine the contact resistance distribution unevenness coefficient; compare the difference between the contact resistance value of each area and the previous measurement value based on the contact resistance distribution unevenness coefficient to determine the time point of abnormal resistance increase; calculate the amplitude of the abnormal resistance increase based on the difference between the contact resistance value at the time point of abnormal resistance increase and the initial value.

4. The method for predicting battery capacity attenuation during charge and discharge cycles according to claim 1, wherein: The step of fusing the contact resistance increase value and the contact resistance uneven distribution data with the high-salt moisture environment data, analyzing the moisture retention time and the salt crystal form, and determining the contact resistance change trend and voltage fluctuation parameters includes: Combining the contact resistance increase value and the humidity value in the high-salt humid environment data, the moisture impact factor is calculated, and the crystallization tendency coefficient is determined based on the ratio of the salt deposition concentration to the saturation concentration. The moisture residence time is calculated using the moisture impact factor and the crystallization tendency coefficient. The salt deposition concentration distribution is analyzed to determine the salt crystal form distribution. Based on the salt crystal form distribution, the contact resistance increase rate of each region is calculated to determine the contact resistance change trend. Based on the contact resistance change trend, the charging interface voltage data is collected to calculate the voltage fluctuation parameters.

5. The method for predicting battery capacity attenuation during charge and discharge cycles according to claim 1, wherein: The extracting of abnormal voltage signal distribution of the charging interface in a high-salt and humid environment based on the voltage fluctuation parameter and the transmission loss of the charging process calculated through the impedance detection data, as well as the salt deposition concentration and temperature fluctuation data, includes: The transmission loss rate of the charging process is calculated based on the voltage fluctuation parameters and the impedance detection data, and the environmental impact index is calculated in combination with the salt deposition concentration and the temperature fluctuation data. Based on the environmental impact index, voltage time series data is collected to determine the characteristic vector of the voltage anomaly signal. Based on the characteristic vector, the distribution law of the voltage anomaly signal is statistically analyzed.

6. The method for predicting battery capacity attenuation during charge and discharge cycles according to claim 1, wherein: Analyzing the voltage drop amplitude according to the voltage abnormality signal distribution to determine the battery capacity retention rate decrease trend includes: According to the voltage abnormality signal distribution, the voltage drop amplitude cumulative value is calculated to determine the single-cycle voltage attenuation ratio; according to the single-cycle voltage attenuation ratio, the battery capacity retention rate is calculated to determine the capacity retention rate decline trend.

7. The method for predicting battery capacity attenuation during charge and discharge cycles according to claim 1, wherein: The method of analyzing the relationship between the surface roughness of the charging interface material and the protective coating parameters and the accumulation of electrode material deposits based on the salt deposition concentration and the salt crystal morphology distribution to determine the material corrosion distribution of the battery capacity decay includes: Obtain regional data on the salt deposition concentration distribution and measure the volume concentration of the deposit on the surface of the electrode material; analyze the crystal structure of the salt crystals based on the deposit volume concentration and determine the salt crystal morphology distribution; measure the surface roughness of the charging interface material and protective coating parameters based on the salt crystal morphology distribution and calculate the corrosion rate; determine the material corrosion distribution based on the corrosion rate and the deposit volume concentration.

8. The method for predicting battery capacity attenuation during charge and discharge cycles according to claim 1, wherein: The method comprises: constructing a battery capacity decay evaluation system based on the material corrosion distribution, the moisture permeation rate, and the protective coating parameters, inputting the contact resistance uneven distribution data and the current loss data, outputting a capacity decay trend curve, and predicting the battery capacity according to the capacity decay trend curve, including: Obtain the material corrosion distribution and the moisture permeation rate, and combine them with the protective coating parameters to generate an evaluation parameter set; calculate the contact resistance distribution unevenness data and the current loss rate based on the evaluation parameter set; generate capacity decay trend data based on the contact resistance distribution unevenness data and the current loss rate; draw a capacity decay trend curve based on the capacity decay trend data to predict the battery capacity.

Images