A vehicle-mounted lithium battery life prediction and driving strategy generation method based on a temperature-mediated aging model

Abstract

The application discloses a kind of vehicle-mounted lithium battery life prediction and driving strategy generation method based on temperature-mediated aging model, belong to vehicle-mounted battery management technical field.The method obtains battery operating data and driving behavior data, generates driving behavior feature matrix and battery operating state sequence;Based on current sequence, heat generation rate, temperature evolution value and temperature consistency index are calculated;The temperature evolution value is input into the aging model containing SEI film growth, lithium extraction reaction and active material rupture, the degradation parameters are updated and SOH and RUL are calculated;Then, according to the user type, aging contribution and risk judgment result, driving strategy, charging and discharging strategy and maintenance warning are generated.

Description

Technical Field

[0001] This invention relates to the fields of vehicle lithium battery health status assessment, battery life prediction, and vehicle driving strategy generation, and particularly to a method for vehicle lithium battery life prediction and driving strategy generation based on a temperature-mediated aging model. Background Technology

[0002] Lithium-ion batteries are the core power source for new energy vehicles, and their health directly affects the vehicle's range, safety, and total lifecycle cost. In real-world operation, on-board lithium batteries are subjected to complex conditions such as charging, discharging, resting, rapid acceleration, rapid deceleration, energy recovery, high temperature, and low temperature. Capacity degradation is not only affected by the battery's own materials and structure, but is also closely related to driving behavior, current fluctuations, temperature field changes, and battery management strategies.

[0003] Existing technologies for predicting the lifespan of automotive lithium-ion batteries mainly include model-driven and data-driven methods. Model-driven methods typically use electrochemical models, equivalent circuit models, or empirical aging models to estimate the state of health. They can explain aging mechanisms such as SEI film growth, lithium plating, and loss of active materials. However, under real-world vehicle dynamic conditions, temperature is often treated as a static external input, making it difficult to characterize the impact of transient heating and temperature inconsistencies caused by driving behavior on aging side reactions. Data-driven methods can use real-vehicle operating data to predict SOH or RUL, but they rely on large amounts of high-quality data. Their generalization ability is insufficient under extreme conditions, data sparsity, or significant differences in user driving habits, and they struggle to explain the physical causes of the prediction results.

[0004] Therefore, there is an urgent need for a technical solution that can unify driving behavior, current characteristics, temperature evolution, battery aging side effects, and strategy output, so that life prediction is no longer limited to a single SOH numerical estimate, but further outputs executable driving strategies, charging and discharging strategies, and maintenance warning information. Summary of the Invention

[0005] The technical problem to be solved by this invention is that existing methods for predicting the lifespan of on-board lithium batteries are difficult to model in a closed loop the current fluctuations, temperature field changes and battery aging side reactions caused by driving behavior, resulting in insufficient adaptability to dynamic operating conditions, weak ability to explain aging mechanisms and insufficient ability to generate personalized driving strategies.

[0006] To address the aforementioned technical problems, this invention provides a method for predicting the lifespan of on-board lithium batteries and generating driving strategies based on a temperature-mediated aging model. The method includes: acquiring on-board lithium battery operating data and driving behavior data; cleaning, normalizing, and extracting features from the data to generate a driving behavior feature matrix and a battery operating state sequence; calculating the battery heat generation rate, temperature evolution value, and temperature consistency index based on the current sequence; inputting the temperature evolution value into a temperature-mediated aging model that includes SEI film growth, lithium plating reaction, and active material cracking to calculate the rates of three types of side reactions; updating the deposition layer thickness, interface resistance, cyclic lithium-ion loss, and capacity decay based on the side reaction rates; calculating the state of health (SOH) and remaining service life (RUL); and generating driving strategies, charge / discharge strategies, and maintenance warnings based on user type, aging contribution, and SOH / RUL results.

[0007] This invention also provides an on-board lithium battery life prediction and driving strategy generation system, including an on-board data acquisition terminal, a data processing and feature extraction unit, a temperature-mediated aging simulation unit, a SOH / RUL prediction unit, a driving strategy generation unit, and a BMS / user terminal feedback unit. This system can be deployed on the vehicle, in the cloud, or on a vehicle-cloud collaborative platform; the specific platform does not constitute a limitation on the scope of protection of this invention.

[0008] Compared with the prior art, the present invention has at least the following beneficial effects:

[0009] 1. This invention uses the battery heat generation and temperature evolution caused by driving behavior through current characteristics as the intermediate variable for battery aging calculation, so that temperature is no longer a static external input, but an aging driving quantity dynamically generated by the actual vehicle operation process.

[0010] 2. This invention integrates three types of side reactions—SEI film growth, lithium plating reaction, and active material rupture—into a temperature-mediated aging model, enabling SOH and RUL prediction results to correspond to specific aging paths and improving the interpretability of lifetime prediction.

[0011] 3. Based on the differences in driving behavior among users with frequent high-power driving, stable and balanced driving, and deep cycle driving, this invention outputs differentiated driving strategies, charging and discharging strategies, and maintenance warnings, enabling personalized battery life management tailored to user driving habits.

[0012] 4. This invention can form an execution feedback loop with BMS or user terminal, so that prediction results, strategy output and subsequent data collection are interconnected, which facilitates continuous correction and online application. Attached Figure Description

[0013] Figure 1 This is a diagram showing the overall architecture of the vehicle-mounted lithium battery life prediction and driving strategy generation system of the present invention.

[0014] Figure 2 This is a flowchart of the method for predicting the lifespan of an on-board lithium battery and generating a driving strategy according to the present invention.

[0015] Figure 3 This is a diagram illustrating the conduction relationship between temperature-mediated driving behavior and battery aging, as presented in this invention.

[0016] Figure 4 This is a flowchart of the multi-source vehicle data processing and user classification process of the present invention;

[0017] Figure 5 This is a simulation flowchart of the temperature-mediated aging model of the present invention;

[0018] Figure 6 A closed-loop graph is generated for the SOH / RUL prediction and driving strategy of this invention;

[0019] Figure 7 This is a comparison chart of the battery life prediction effects for three types of users in an embodiment of the present invention.

[0020] Figure labeling: 101, vehicle-mounted data acquisition terminal; 102, data processing and feature extraction unit; 103, temperature-mediated aging simulation unit; 104, SOH / RUL prediction unit; 105, driving strategy generation unit; 106, BMS / user terminal feedback unit; 107, intelligent agent platform. Detailed Implementation

[0021] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. Without departing from the concept of the present invention, those skilled in the art can make equivalent substitutions or combinations of data sources, model deployment methods, user classification algorithms, and strategy output formats.

[0022] like Figure 1 As shown, the system of the present invention includes an on-board data acquisition terminal 101, a data processing and feature extraction unit 102, a temperature-mediated aging simulation unit 103, a SOH / RUL prediction unit 104, a driving strategy generation unit 105, a BMS / user terminal feedback unit 106, and an intelligent agent platform 107.

[0023] The onboard data acquisition terminal 101 is used to acquire driving behavior data, battery operation data, ambient temperature data, and basic vehicle parameters from the onboard CAN bus, BMS, battery pack temperature sensor, vehicle networking terminal, or vehicle controller. Driving behavior data may include vehicle speed, acceleration, frequency of rapid acceleration, frequency of rapid deceleration, pedal opening, and driving period; battery operation data may include battery voltage, battery current, SOC, individual cell temperature, battery pack temperature difference, and individual cell voltage range; basic vehicle parameters may include battery rated capacity, number of individual cells, thermal resistance, thermal capacity, initial capacity, and electrode parameters.

[0024] The data processing and feature extraction unit 102 is used to perform outlier removal, missing value filling, time alignment, Min-Max normalization and feature filtering on the collected data, and output driving behavior feature matrix, battery operating state sequence, corrected temperature sequence and user type label.

[0025] The temperature-mediated aging simulation unit 103 is used to calculate the battery heat generation rate, temperature evolution value, and temperature consistency index VCPT based on the current sequence and temperature sequence, and further calculate the rates of three types of side reactions: SEI film growth, lithium plating reaction, and active material rupture.

[0026] The SOH / RUL prediction unit 104 is used to update the deposit thickness, interfacial resistance, cyclic lithium-ion loss, and capacity decay based on the side reaction rate, and to calculate the state of health (SOH) and remaining lifetime (RUL).

[0027] The driving strategy generation unit 105 is used to generate driving strategies, charging and discharging strategies, maintenance warnings, and monthly health reports based on user type, SOH / RUL prediction results, side reaction contribution, and abnormal risk assessment results. The BMS / user terminal feedback unit 106 is used to output strategies to the BMS or user terminal and collect execution feedback, which can be returned to the intelligent agent platform 107 for subsequent model correction.

[0028] like Figure 2 As shown, the method of the present invention includes the following steps.

[0029] S101 acquires onboard lithium battery operating data and driving behavior data. Specifically, data can be collected at 10 seconds / frame, 1Hz, or other sampling frequencies. For high-frequency driving behavior and current data, a higher sampling frequency can be used to capture transient conditions such as rapid acceleration, rapid deceleration, and energy recovery; for basic vehicle parameters and environmental parameters, a lower frequency can be used for updating.

[0030] S102 involves cleaning, normalizing, and extracting features from the data. The cleaning process includes removing outliers in current, voltage, temperature, and vehicle speed that clearly exceed physical boundaries, handling missing values ​​using linear interpolation, nearest neighbor interpolation, or generative imputation, and aligning multi-source data based on timestamps.

[0031] S103, construct the driving behavior feature matrix and battery operating state sequence. The driving behavior feature matrix may include features such as maximum negative acceleration, vehicle speed standard deviation, average negative acceleration, maximum vehicle speed, average vehicle speed, maximum positive acceleration, accelerator pedal travel standard deviation, frequency of rapid acceleration, frequency of rapid deceleration, and percentage of high-power operation; the battery operating state sequence may include current, SOC, temperature, voltage, cell temperature difference, and voltage range.

[0032] S104 calculates the battery heat generation rate, temperature evolution value, and temperature consistency index based on the current sequence. This step transforms driving behavior into heat generation intensity and temperature field changes through current amplitude, fluctuation frequency, and duration.

[0033] S105: Input the temperature evolution value into the aging model to calculate the SEI film growth, lithium plating reaction, and active material rupture rate.

[0034] S106 updates the deposition layer thickness, interfacial resistance, cyclic lithium-ion loss, and capacity decay based on the side reaction rate.

[0035] S107, calculate SOH and RUL.

[0036] S108 generates driving strategies, charging / discharging strategies, and maintenance warnings based on user type, aging contribution, and SOH / RUL results.

[0037] The conduction mechanism of temperature-mediated driving behavior affecting battery aging

[0038] like Figure 3 As shown, this invention sequentially correlates driving behavior parameters, current characteristic parameters, electrochemical losses, heat generation process, temperature evolution process, aging mechanism effects, aging result evolution, and performance degradation manifestations. Driving behavior parameters include vehicle speed, acceleration, and pedal opening; current characteristic parameters include current amplitude, fluctuation frequency, and duration; electrochemical losses include ohmic losses and polarization losses; the heat generation process outputs the heat generation rate; the temperature evolution process outputs the battery temperature evolution value and the temperature consistency index VCPT; the aging mechanism effects include SEI film growth, lithium plating reaction, and active material rupture; the aging result evolution includes deposit layer thickening, increased interfacial resistance, and loss of cyclic lithium; and performance degradation manifestations include capacity decay, decreased SOH, and shortened RUL.

[0039] The temperature in this invention is not a fixed external input, but is calculated by a thermal model based on current fluctuations caused by driving behavior. Aggressive driving will cause high-rate, high-frequency current fluctuations, which will increase the heat generation rate and disrupt temperature consistency; smooth driving will result in a smaller current amplitude and a stable direction, making the heat generation process relatively gentle; deep cycle driving will generate a large current load in the low SOC stage and increase the risk of lithium plating in the rapid recharge stage.

[0040] Multi-source data processing and user classification

[0041] like Figure 4 As shown, the input data sources include vehicle CAN data, BMS data, ambient temperature data, and basic vehicle parameters. The processing steps include outlier removal, missing value imputation, time alignment, Min-Max normalization, MIC feature selection, driving behavior feature matrix generation, and user type classification.

[0042] In one embodiment, the Maximum Information Coefficient (MIC) is used to analyze the nonlinear correlation between driving behavior parameters and the Temperature Consistency Index (VCPT). The MIC values ​​and grading criteria are as follows: 0 to 0.09 indicates no correlation, 0.1 to 0.3 indicates weak correlation, 0.3 to 0.5 indicates moderate correlation, and 0.5 to 1.0 indicates strong correlation. After screening, maximum negative acceleration, vehicle speed standard deviation, and average negative acceleration are identified as key features affecting temperature consistency. Furthermore, a clustering algorithm can be used to classify users into frequent high-power users, stable and balanced users, and deeply cyclical users.

[0043] Frequent high-power users exhibit frequent rapid acceleration and deceleration for energy recovery, with frequent switching of current direction and large amplitude, and a high proportion of high-power range; stable and balanced users exhibit constant speed driving and regular charging, with small current amplitude and stable direction; deep-cycle users exhibit long-distance deep discharge and concentrated fast charging, with a higher risk of lithium plating under low SOC and high rate conditions.

[0044] Temperature-mediated aging model

[0045] like Figure 5 As shown, the inputs to the temperature-mediated aging model include a standardized current sequence, a state of charge (SOC) sequence, a temperature sequence, and fundamental battery parameters. The model first calculates the temperature evolution using a first-order equivalent thermal model, then calculates the heat generation rate using the Bernardi heat generation model, and corrects this using the temperature consistency index VCPT. Subsequently, the SEI film growth rate, lithium plating reaction rate, and active material breakage rate are calculated in parallel. After correction by side reaction coupling, the aging parameters are updated, and the outputs are the SEI film thickness, lithium plating layer thickness, interfacial resistance, cyclic lithium loss, and state of equilibrium (SOH).

[0046] In one embodiment, the first-order equivalent thermal model is:

[0047]

[0048] in, Indicates the overall battery temperature. Indicates ambient temperature. This represents the equivalent thermal resistance between the battery temperature and the ambient temperature. This indicates the concentrated heat capacity of the battery. This represents the internal heat generation power of the battery. This formula is used to calculate the temperature change over time based on the balance between heat generation and dissipation.

[0049] Bernardi's heat generation model is:

[0050]

[0051] in, Indicates the rate at which the battery generates heat. Indicates battery volume, Indicates the battery operating current. Indicates open-circuit voltage. Indicates the operating voltage. Indicates battery temperature. This represents the partial derivative of the open-circuit voltage with respect to temperature. This formula is used to convert the current signal generated by driving behavior into heat intensity.

[0052] The Temperature Consistency Index (VCPT) is:

[0053]

[0054] in, The standard deviation of the mean temperature at each time point within the selected driving segment. The VCPT represents the average temperature of the battery pack segment. The larger the VCPT, the more uneven the temperature distribution inside the battery pack. When the VCPT exceeds 0.05, it can be determined that there is a risk of local hot spots or a deterioration in temperature uniformity, and it can be used as a correction factor for the side reaction rate.

[0055] SEI membrane growth calculation

[0056] The growth current density of the SEI film can be expressed as:

[0057]

[0058] in, This represents the current density during SEI film growth. Indicates the effective specific surface area. Represents the exchange current density. Indicates the transfer coefficient of the reduction reaction. Denotes Faraday's constant. Represents the ideal gas constant. Represents absolute temperature. This represents the potential difference at the solid-liquid interface. This represents the equilibrium potential of the SEI reaction.

[0059] The change of exchange current density with temperature satisfies:

[0060]

[0061] in, This represents the SEI exchange current density at the reference temperature. Indicates the activation energy of the SEI reaction. This represents the reference temperature. This formula characterizes the exponential change in SEI film growth rate with temperature following a temperature increase caused by driving behavior.

[0062] Calculation of lithium plating reaction

[0063] The current density for the lithium plating reaction can be expressed as:

[0064]

[0065] in, This represents the current density during the lithium plating reaction. Indicates the effective specific surface area. This represents the lithium plating exchange current density. Indicates the transfer coefficient of the reduction reaction. This represents the lithium plating equilibrium potential. When the potential difference at the solid-liquid interface is lower than the lithium plating trigger potential, the lithium plating reaction is triggered; when the potential difference at the solid-liquid interface is not lower than the trigger potential, the lithium plating reaction current density is zero.

[0066] In actual vehicle operation, low temperatures can lead to increased electrolyte viscosity, decreased lithium-ion diffusion coefficient, and intensified polarization of the negative electrode surface concentration, making it easier to trigger lithium plating. While high temperatures are conducive to diffusion, they can also accelerate SEI film growth and increase interfacial impedance, potentially exacerbating the risk of localized lithium plating under conditions of localized hot spots and high current density coupling.

[0067] Calculation of active material fracture

[0068] The fracture rate of the active material can be expressed as:

[0069]

[0070] in, Indicates the volume fraction of active materials. Indicates the coefficient of failure of the active material. This represents the total current density on the surface of the active particles. This represents the main reaction current density during normal charge and discharge. The greater the lithium-ion flux through the surface of the active particles, the greater the lithium concentration gradient inside the particles, and the higher the fracture rate caused by diffusion-induced stress.

[0071] Aging parameter update and capacity decay calculation

[0072] The deposition of SEI film and lithium plating layer alters the electrode microstructure. The rate of increase in lithium plating layer thickness can be expressed as:

[0073]

[0074] in, Indicates the thickness of the lithium plating layer. This indicates the initial volume fraction of the active material. This represents the initial effective specific surface area. This indicates the molar volume of the lithium plating product.

[0075] The rate of increase in SEI film thickness can be expressed as:

[0076]

[0077] in, Indicates the thickness of the SEI film. The formula represents the molar volume of the SEI product. The 2 in the denominator indicates that 2 moles of lithium ions are consumed to generate 1 mole of SEI product.

[0078] As an electronic insulator, the SEI film's increased thickness leads to an increase in the particle surface resistivity. The rate of surface resistivity increase can be expressed as:

[0079]

[0080] in, Indicates the interface resistance of the SEI film. This represents the conductivity of the SEI film. This formula is used to map an increase in SEI film thickness to an increase in interfacial resistance.

[0081] The recyclable lithium-ion loss consists of three parts. The lithium loss due to SEI film growth is as follows:

[0082]

[0083] The lithium loss due to the lithium plating reaction is as follows:

[0084]

[0085] The lithium loss due to the breakage of the active material is as follows:

[0086]

[0087] in, This represents the negative electrode cross-sectional area. Indicates the thickness of the negative electrode. This indicates the maximum lithium concentration at the negative electrode. This represents the average lithium intercalation fraction.

[0088] The total lithium-ion loss rate is the sum of the three:

[0089]

[0090] The cumulative lithium-ion loss is obtained by time integration:

[0091]

[0092] This allows us to obtain the remaining battery capacity and health status:

[0093]

[0094]

[0095] in, Indicates the initial capacity of the battery. express Remaining battery capacity at any time.

[0096] In another embodiment, the relationship between capacity decay and equivalent temperature can also be represented by the temperature-mediated Arrhenius capacity decay rate:

[0097]

[0098] in, Indicates the temperature-mediated capacity decay rate. Indicates pre-exponential factor, This represents the equivalent activation energy for capacity decay. Represents the ideal gas constant. This represents the equivalent temperature after adjusting for vehicle power load. This formula is used to convert the temperature load under high-power driving conditions into a capacity decay rate.

[0099] SOH / RUL Prediction and Strategy Generation Closed Loop

[0100] like Figure 6 As shown, the SOH prediction results, RUL prediction results, side reaction contribution, user type label, and abnormal temperature rise / lithium plating risk assessment are all input into the strategy generation unit. The strategy generation unit outputs driving strategy, charging strategy, maintenance warning, and monthly health report, and executes feedback through the user terminal or BMS back to the prediction and strategy generation stage, forming a closed loop.

[0101] For users who frequently use high-power charging, the strategy generation unit can output strategies to reduce the frequency of rapid acceleration, limit acceleration, enable economy driving mode, reduce fast charging frequency, and keep the fast charging percentage within a preset range. For users who frequently engage in deep charging cycles, the strategy generation unit can output strategies to charge when the SOC drops to 30% to 40%, avoid deep discharge, reduce full charge / discharge cycles, and plan for stable driving speeds in advance. For users who prefer a stable and balanced driving style, the strategy generation unit can output strategies to maintain existing driving habits, prioritize charging during off-peak hours, keep the SOC between 40% and 80%, and perform periodic health checks.

[0102] The system can issue a maintenance warning when the standard deviation of the single-cell voltage exceeds 0.015V, the voltage range exceeds 0.04V, the average temperature consistently exceeds 35℃, high power output occurs in low-temperature environments, the state of equilibrium (SOH) drops below 85%, or the risk of lithium plating exceeds a preset threshold. When the SOH drops to around 80%, the system can issue recommendations for battery replacement, repair, or secondary use.

[0103] Example data and technical effects

[0104] In one embodiment, approximately 350,000 frames of real-vehicle operating lithium battery data were used for verification. This data covered approximately 16,500 kilometers of driving, with battery health decreasing from 100% to approximately 85%, a temperature range of approximately 80°C, a state of charge (SOC) primarily between 20% and 80%, a peak discharge current reaching -1000A, a charging current within +200A, 96 cells connected in series, and an internal temperature difference of approximately 8°C to 15°C within the battery pack. Surface resistivity increased from approximately 0.05 mΩ to approximately 0.20 mΩ, the liquid phase volume fraction decreased from 0.40 to approximately 0.32, and the active particle volume fraction decreased from approximately 0.50 to approximately 0.46.

[0105] like Figure 7 As shown, the SOH decay and remaining capacity trends differ among the three types of users. Frequent high-power users exhibit significant current fluctuations, resulting in a mid-term accelerated SOH decay, with a predicted mileage of approximately 18,000 km. Stable and balanced users have smaller current amplitudes and more stable directions, leading to the lowest aging rate, with a predicted mileage of approximately 22,500 km, representing a lifespan extension of about 25% compared to frequent high-power users. Deep-cycle users are more prone to lithium plating reactions at the end of deep discharge and during rapid recharging, resulting in the fastest capacity decay, with a predicted mileage of approximately 17,000 km.

[0106] As can be seen from the above embodiments, the present invention can transform differences in driving behavior into differences in temperature evolution, and further map them into three types of side reaction rates, capacity decay, and lifetime prediction results, thereby generating differentiated driving strategies and charging / discharging strategies for different users.

Claims

1. A method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model, characterized in that, Includes the following steps: S101, acquire vehicle lithium battery operation data and driving behavior data. The vehicle lithium battery operation data includes at least battery voltage, battery current, state of charge, cell temperature, ambient temperature and basic battery parameters. The driving behavior data includes at least vehicle speed, acceleration, pedal opening and vehicle running time. S102, perform outlier removal, missing value filling, time alignment, normalization and feature extraction on the vehicle lithium battery operation data and driving behavior data to obtain the driving behavior feature matrix and battery operation state sequence; S103, calculate the heat generation rate, temperature evolution value, and temperature consistency index of the vehicle lithium battery based on the current amplitude, current fluctuation frequency, and charging / discharging duration in the driving behavior feature matrix. S104, input the temperature evolution value and temperature consistency index into the temperature-mediated aging model to calculate the SEI film growth rate, lithium plating reaction rate and active material rupture rate respectively; S105, based on the SEI film growth rate, lithium plating reaction rate and active material rupture rate, update the deposition layer thickness, interface resistance, cyclic lithium ion loss and capacity decay. S106, calculate the State of Health (SOH) and Remaining Service Life (RUL) of the vehicle lithium battery based on the capacity decay, and generate the contribution of three types of side reactions to the aging results; S107. Determine the user type based on the driving behavior feature matrix, health status (SOH), remaining useful life (RUL), and aging contribution, and output the corresponding driving strategy, charging / discharging strategy, and maintenance warning information according to the user type.

2. The method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model according to claim 1, characterized in that, In S102, the feature extraction includes: screening driving behavior features related to the temperature consistency index based on the maximum information coefficient (MIC), and using at least one of the following as feature variables: maximum negative acceleration, vehicle speed standard deviation, average negative acceleration, maximum vehicle speed, average vehicle speed, maximum positive acceleration, and accelerator pedal travel standard deviation; wherein, when the MIC value is between 0 and 0.09, it is determined to be uncorrelated; when the MIC value is between 0.1 and 0.3, it is determined to be weakly correlated; when the MIC value is between 0.3 and 0.5, it is determined to be moderately correlated; and when the MIC value is between 0.5 and 1.0, it is determined to be strongly correlated.

3. The method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model according to claim 1, characterized in that, In S103, the temperature evolution value and heat generation rate are calculated using the first-order equivalent thermal model and the Bernardi heat generation model, and at least the following relationships are satisfied: ； ； in, Indicates the overall battery temperature. Indicates ambient temperature. This represents the equivalent thermal resistance between the battery and the environment. Indicates the concentrated heat capacity of the battery. Indicates the rate at which the battery generates heat. Indicates the battery operating current. Indicates battery volume, Indicates open-circuit voltage. Indicates the operating voltage.

4. The method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model according to claim 3, characterized in that, In S103, the temperature consistency index is the probe temperature variation coefficient (VCPT), and it satisfies the following relationship: ； in, The standard deviation of the mean temperature at each time point within the selected driving segment. The average temperature of the driving segment is used; when the VCPT exceeds a preset threshold, it is used as a correction factor for local hot spots or uneven temperature conditions and input into the temperature-mediated aging model.

5. The method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model according to claim 4, characterized in that, In S104, the SEI film growth rate is calculated based at least on the effective specific surface area, exchange current density, reduction reaction transfer coefficient, Faraday constant, ideal gas constant, absolute temperature, solid-liquid interfacial potential difference, and equilibrium potential, and the exchange current density changes with temperature according to the Arrhenius relation: ； The temperature rise caused by driving behavior can be used to change the SEI film growth rate through the temperature term.

6. The method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model according to claim 1, characterized in that, In S104, the lithium plating reaction rate is triggered when the potential difference at the solid-liquid interface is lower than the lithium plating equilibrium potential, and is zero when the potential difference at the solid-liquid interface is not lower than the lithium plating equilibrium potential; the active material rupture rate is calculated based on the combined current density of the normal charge-discharge main reaction current density, the SEI film growth current density, and the lithium plating reaction current density, in order to characterize the effect of diffusion-induced stress caused by lithium ion flux on the rupture of active particles.

7. The method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model according to claim 5, characterized in that, In S105 and S106, the recyclable lithium-ion loss includes lithium loss due to SEI film growth, lithium loss due to lithium plating reaction, and lithium loss due to active material breakage. The State of Health (SOH) of the vehicle-mounted lithium battery must at least satisfy the following relationship: ； ； in, This indicates the total loss of recyclable lithium ions. This indicates lithium loss due to SEI film growth. This indicates the lithium loss caused by the lithium plating reaction. This indicates lithium loss due to the breakage of the active material. Indicates the initial capacity of the battery. express Remaining battery capacity at any time.

8. The method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model according to claim 1, characterized in that, In S107, the user types include frequent high-power users, stable and balanced users, and deep-cycle users; among them, frequent high-power users correspond to driving behaviors with frequent switching of current direction and a high proportion of high-power range, stable and balanced users correspond to driving behaviors with small current amplitude and stable direction, and deep-cycle users correspond to driving behaviors with long-term deep discharge and centralized fast charging.

9. The method for predicting the lifespan of an on-board lithium battery and generating driving strategies based on a temperature-mediated aging model according to claim 8, characterized in that, The driving strategy, charging / discharging strategy, and maintenance warning information include: for frequent high-power users, strategies to reduce the frequency of rapid acceleration, limit acceleration, and reduce the proportion of fast charging; for deep-cycle users, strategies to charge before the state of charge reaches a preset lower limit, reduce deep discharge, and plan a stable vehicle speed; for stable and balanced users, strategies to maintain current driving habits, prioritize off-peak charging, and conduct periodic health checks; and when voltage difference, cell temperature, SOH, or lithium plating risk exceed the corresponding threshold, maintenance warning information is output.

10. A system for predicting the lifespan of an on-board lithium battery and generating a driving strategy based on a temperature-mediated aging model, implementing the method as described in any one of claims 1-9, characterized in that, include: The system comprises an onboard data acquisition unit, a data processing and feature extraction unit, a temperature-mediated aging simulation unit, a State of Health (SOH) / Remaining Life (RUL) prediction unit, a driving strategy generation unit, and a BMS / user terminal feedback unit. The onboard data acquisition unit acquires onboard lithium battery operating data and driving behavior data. The data processing and feature extraction unit generates a driving behavior feature matrix and a battery operating state sequence. The temperature-mediated aging simulation unit calculates temperature evolution values, temperature consistency indices, SEI film growth rate, lithium plating reaction rate, and active material rupture rate. The SOH / RUL prediction unit calculates the State of Health (SOH) and Remaining Life (RUL). The driving strategy generation unit generates driving strategies, charging / discharging strategies, and maintenance warning information based on user type, aging contribution, SOH, and RUL. The BMS / user terminal feedback unit receives the strategies and warning information and provides feedback on the execution results.

Images