Battery temperature regulation and control method and system

Abstract

The invention belongs to the technical field of battery temperature regulation and control, and particularly relates to a battery temperature regulation and control method and system, and the method comprises the steps: S1, obtaining an initial lithium deposition risk cloud picture, and calculating an initial lithium deposition risk index; s2, when the initial lithium deposition risk index is greater than a set lithium deposition risk threshold value, diagnosing a battery fault type based on the initial lithium deposition risk cloud picture and determining a temperature regulation and control scheme; s3, executing the temperature regulation and control scheme, and obtaining a regulated and controlled lithium deposition risk index and a regulation and control scheme feedback index after regulation and control: if the regulated and controlled lithium deposition risk index is not greater than a set lithium deposition risk threshold value, ending regulation and control, and determining whether to correct the temperature regulation and control scheme based on the regulation and control scheme feedback index; and if the regulated lithium deposition risk index is greater than the set lithium deposition risk threshold value, temperature regulation is performed again after the temperature regulation scheme is corrected. The safety, the service life and the operation efficiency of the battery are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery temperature regulation, and particularly relates to a battery temperature regulation method and system. Background Art

[0002] In the application of lithium-ion batteries, battery performance and safety have always been the focus of research. Although the current battery management system can monitor basic parameters such as battery voltage, current, and temperature, it has deficiencies in predicting and controlling the risk of lithium deposition; the risk of lithium deposition is a key factor affecting battery performance and safety, and is closely related to the temperature distribution inside the battery. Existing temperature regulation methods mostly focus on maintaining the battery within an appropriate operating temperature range, but fail to fully consider the complex interaction between the risk of lithium deposition and temperature; in addition, when diagnosing and warning battery failures, existing technologies often lack accurate assessment and timely response to the risk of lithium deposition, which may lead to irreversible decline in battery performance and even cause safety accidents.

[0003] Another major drawback of the existing technology is the lack of dynamic monitoring and intelligent regulation capabilities for the risk of lithium deposition. In actual applications, the working environment and load conditions of the battery are complex and variable, making it difficult to predict the dynamic changes in the risk of lithium deposition; existing temperature regulation schemes are usually relatively fixed and cannot be adjusted in real time according to the complex physical and chemical processes inside the battery. This makes it impossible to take effective temperature regulation measures in a timely manner when the risk of lithium deposition is high; moreover, existing technologies also have deficiencies in multi-physical field coupling analysis and are difficult to comprehensively reflect the interaction mechanism between the risk of lithium deposition and multiple fields such as temperature and stress. Summary of the Invention

[0004] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide a battery temperature regulation method and system, which can accurately monitor and intelligently regulate the risk of lithium deposition in the battery through multi-physical field coupling analysis, and significantly improve the safety, service life, and operating efficiency of the battery by combining a closed-loop feedback optimization mechanism.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A battery temperature regulation method, comprising the following steps: S1. Obtain the initial lithium deposition risk cloud map and calculate the initial lithium deposition risk index; S2. When the initial lithium deposition risk index is greater than the set lithium deposition risk threshold, diagnose the type of battery failure based on the initial lithium deposition risk cloud map and determine the temperature regulation plan; S3. Execute the temperature regulation plan, and obtain the lithium deposition risk index after regulation and the feedback index of the regulation plan after regulation: If the lithium deposition risk index after regulation is not greater than the set lithium deposition risk threshold, the regulation ends, and it is determined whether to modify the temperature regulation plan based on the feedback indicators of the regulation plan; If the lithium deposition risk index after regulation is greater than the set lithium deposition risk threshold, the temperature regulation plan is modified and then the temperature regulation is carried out again.

[0006] Preferably, in the above S1, obtaining the initial lithium deposition risk cloud map includes the following steps: Obtain the heat generation rate driven by the lithium ion concentration: ; ; In the formula, is the lithium ion concentration at position at time t, is the time step, is the change amount of the lithium ion concentration at position within the time step ; is the heat generation rate driven at position , , and are both transfer functions, is the open circuit voltage at position , is the terminal voltage, is the charge and discharge current, is the temperature at position at time t, is the overall average temperature of the battery, is the thermal resistance coefficient of the SEI film side reaction, V is the volume of the battery; Construct a temperature field model through the heat generation rate driven and output the temperature gradient: ; ; In the formula, is the equivalent average density of the whole battery, is the equivalent average specific heat capacity of the whole battery, is the equivalent thermal conductivity at position , is the temperature gradient, is the basic thermal conductivity, is the temperature gradient attenuation factor, e is the natural constant, is the divergence operator; Calculate the thermal stress field according to the temperature field and obtain the concentration gradient additional stress field based on the lithium ion concentration: ; ; In the formula, is the thermal stress field, E is the elastic modulus, is the coefficient of thermal expansion, is the additional stress field of the concentration gradient, is the reference temperature, is the concentration-stress coupling coefficient, is the second-order spatial derivative of the lithium ion concentration; Adding the thermal stress field and the additional stress field of the concentration gradient gives the total stress field : ; Based on the temperature gradient and the total stress field, an initial lithium deposition risk contour map is obtained.

[0007] Preferably, an initial lithium deposition risk contour map is obtained based on the temperature gradient and the total stress field , and the obtaining method is: ; ; ; ; In the formula, is the electrochemical polarization-concentration field coupling term, is 's weight factor, is the temperature gradient-stress field coupling term, is 's weight factor, is the SEI film-side reaction coupling term, is 's weight factor, is the local overpotential at position , is the critical overpotential, is the second-order spatial derivative of the saturation concentration, is the critical temperature gradient, is the critical total stress, is the SEI film thickness at position , is the critical SEI film thickness, is the side reaction rate at position , is the reference side reaction rate, is the temperature gradient-stress field coupling factor; The calculation process of the initial lithium deposition risk index is to accumulate the initial lithium deposition risk values corresponding to all coordinate points in the initial lithium deposition risk cloud map and then calculate the average value.

[0008] Preferably, in S2, diagnosing the battery failure type based on the initial lithium deposition risk cloud map and determining the temperature control scheme includes the following steps: Obtain the SEI film thickness data; Construct the battery temperature characteristic data by combining the SEI film thickness data, the initial lithium deposition risk cloud map, the thermal stress field, and the temperature gradient; Input the battery temperature characteristic data into the trained random forest model to output the battery failure type; Determine the thermal runaway probability based on the initial lithium deposition risk index: ; In the formula, is the thermal runaway probability, is the initial lithium deposition risk index; If the thermal runaway probability is greater than the set thermal runaway probability threshold, determine the temperature control scheme based on the characteristic data before battery runaway, where the characteristic data before battery runaway includes the lithium ion concentration gradient data , the SEI film thickness data , the local overpotential data , the temperature gradient data and the side reaction rate data ; If the thermal runaway probability is not greater than the set thermal runaway probability threshold, conduct a fault warning based on the set warning mechanism.

[0009] Preferably, determining the temperature control scheme based on the characteristic data before battery runaway includes the following steps: Obtain the scheme pointing data set corresponding to the battery failure type stored in the database. The scheme pointing data set includes several scheme pointing data, and each scheme pointing data corresponds to a temperature control scheme; Compare the characteristic data before battery runaway with each scheme pointing data one by one to obtain the pointing coefficient , determine the scheme pointing data corresponding to the largest pointing coefficient, and obtain the temperature control scheme corresponding to this scheme pointing data, where the scheme pointing data includes the lithium ion concentration gradient pointing data , the SEI film thickness pointing data , the local overpotential pointing data , the temperature gradient pointing data and the side reaction rate pointing data , and i is the number of the scheme pointing data.

[0010] Preferably, the method for obtaining the pointing coefficient is as follows: ; In the formula, is the cosine similarity function.

[0011] Preferably, in the said S3, the process of obtaining the feedback index of the regulation scheme is as follows: After the temperature regulation scheme is executed, obtain the characteristic data of the battery temperature after regulation, including the local overpotential data CU after regulation, the side reaction rate data CF after regulation, and the optimized amount CY of the lithium deposition risk index after regulation; Obtain the optimal performance data in the historical execution record of the temperature regulation scheme, including the optimal local overpotential data TU after regulation and the optimal side reaction rate data TF after regulation; Obtain the reference deviation data based on the remaining service life of the battery, including the reference deviation value SU of the local overpotential after regulation and the reference deviation value SF of the side reaction rate after regulation; Determine the feedback index FK of the regulation scheme based on the characteristic data of the battery temperature after regulation, the optimal performance data, and the reference deviation data. The calculation formula is as follows: ; In the formula, is the cosine similarity function.

[0012] Preferably, the steps of obtaining the reference deviation data based on the remaining service life of the battery include: Obtain the historical lithium deposition risk index data of the battery, including the number C of times when the historical lithium deposition risk index is greater than the set lithium deposition risk threshold, the historical maximum lithium deposition risk index Z, the lithium deposition risk index interval value Q1, the average value H of the SEI film thickness, and the SEI film thickness interval value Q2; Input the historical lithium deposition risk index data of the battery into the trained remaining life prediction model to obtain the remaining service life of the battery, where the remaining life prediction model is as follows: ; In the formula, is the remaining life, is the rated service life, is the loss of life, is the intercept, and a, b, c, d, f are all regression coefficients; Compare the remaining service life of the battery with the mapping relationship between the remaining life-reference deviation data stored in the database to determine the reference deviation data corresponding to the remaining service life of the battery.

[0013] Preferably, in the said S3, determine whether to correct the temperature regulation scheme based on the feedback index of the regulation scheme. If the feedback index of the regulation scheme is less than the set index threshold, correct the temperature regulation scheme. The correction process is as follows: Establish a battery model of the current state based on simulation software; Establish a multi-objective fitness function DM: ; In the formula, is the initial lithium deposition risk index, and FK is the feedback index of the regulation scheme; Construct multiple candidate temperature control schemes, regard each candidate temperature control scheme as an individual, and screen the generated multiple candidate temperature control schemes based on the genetic algorithm, and only retain one individual in each generation; After repeating multiple generations, output the optimal individual, and use the candidate temperature regulation scheme corresponding to the optimal individual to correct the temperature regulation scheme.

[0014] A battery temperature regulation system for implementing the above method, including: An initial lithium deposition risk index acquisition module for acquiring an initial lithium deposition risk cloud map and calculating the initial lithium deposition risk index; A fault type and regulation scheme determination module for diagnosing the battery fault type and determining the temperature regulation scheme based on the initial lithium deposition risk cloud map when the initial lithium deposition risk index is greater than the set lithium deposition risk threshold; A regulation scheme feedback module for acquiring the regulated lithium deposition risk index and the regulation scheme feedback index after regulation; A regulation scheme correction module ends the regulation when the regulated lithium deposition risk index is not greater than the set lithium deposition risk threshold, and determines whether to correct the temperature regulation scheme based on the regulation scheme feedback index; or when the regulated lithium deposition risk index is greater than the set lithium deposition risk threshold, correct the temperature regulation scheme and then re-perform temperature regulation.

[0015] The present invention has the following beneficial effects: The present invention realizes the real-time and accurate monitoring of the lithium deposition risk inside the battery by accurately constructing the lithium deposition risk cloud map and quantitatively evaluating the lithium deposition risk in combination with the temperature gradient and the total stress field. In the risk diagnosis link, the random forest model is used to intelligently diagnose the battery fault type, and the regulation strategy is flexibly determined according to the thermal runaway probability threshold. When the risk is high, based on the multi-dimensional characteristic data before battery runaway, the optimal temperature regulation scheme is intelligently matched to ensure the pertinence and effectiveness of the scheme. After the regulation is executed, the effect of the scheme is quantitatively evaluated through the feedback index, and the regulation scheme is dynamically corrected based on the genetic algorithm, forming a closed-loop optimization mechanism to continuously improve the regulation accuracy and effect.

[0016] The present invention has achieved a breakthrough in multi - physical - field coupling analysis at the technical level, deeply integrating multi - field information such as the electrochemical field, temperature field, and stress field, comprehensively capturing the complex relationship between lithium deposition risk and battery operating status, and effectively solving the problem of insufficient multi - field coupling analysis ability in the existing technology. By introducing a remaining life prediction model, the regulation scheme is combined with the full - life - cycle management of the battery, and the regulation strategy is dynamically adjusted based on the battery health state. This not only extends the battery life but also improves the safety and reliability of the battery throughout its life cycle, providing a strong guarantee for the long - term stable operation of the battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flowchart of the battery temperature regulation method of the present invention; Figure 2 is a flowchart for obtaining the initial lithium deposition risk cloud map of the battery temperature regulation method of the present invention; Figure 3 is a schematic diagram of the modules of the battery temperature regulation system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0019] In the calculation of each formula in this embodiment, the dimension - less processing can be carried out as needed to simplify the calculation.

[0020] Embodiment 1: As Figure 1 shown, a battery temperature regulation method includes the following steps: S1. Obtain the initial lithium deposition risk cloud map and calculate the initial lithium deposition risk index; S2. When the initial lithium deposition risk index is greater than the set lithium deposition risk threshold, diagnose the battery failure type based on the initial lithium deposition risk cloud map and determine the temperature regulation scheme; S3. Execute the temperature regulation scheme, and obtain the regulated lithium deposition risk index and the regulation scheme feedback index after regulation: If the regulated lithium deposition risk index is not greater than the set lithium deposition risk threshold, the regulation ends, and it is determined whether to correct the temperature regulation scheme based on the regulation scheme feedback index; If the regulated lithium deposition risk index is greater than the set lithium deposition risk threshold, the temperature regulation scheme is corrected and then the temperature regulation is carried out again.

[0021] As Figure 2 shown, in S1, obtaining the initial lithium deposition risk cloud map includes the following steps: Obtain the driving heat generation rate based on the lithium ion concentration: ; ; In the formula, is the lithium ion concentration at position at time t, is the time step, is the time step within the position the change in lithium ion concentration at position is the driving heat generation rate at position ; , and are both transfer functions, is the open circuit voltage at position ; is the terminal voltage, is the charge and discharge current, is the temperature at position at time t, is the overall average temperature of the battery, is the thermal resistance coefficient of the SEI film side reaction, and V is the volume of the battery; Incorporating the lithium ion concentration related parameters into the calculation of the driving heat generation rate can accurately reflect the heat change in the battery due to lithium ion activity. The lithium ion concentration changes continuously during the charge and discharge process of the battery. Through this correlation calculation, the heat generation mechanism inside the battery can be grasped more accurately, providing a key heat data basis for subsequent temperature control and making the control more targeted.

[0022] By constructing a temperature field model through the driving heat generation rate and outputting the temperature gradient; by constructing a temperature field model using the driving heat generation rate, the heat distribution inside the battery can be visually presented in the form of a temperature gradient. The temperature gradient is an important indicator describing the temperature difference inside the battery. Clearly understanding the temperature difference of each part of the battery helps to determine the heat concentration area and the weak heat dissipation link, thus pointing out the key area for subsequent temperature control and improving the control efficiency.

[0023] By constructing a temperature field model through the driving heat generation rate and outputting the temperature gradient: ; ; In the formula, is the equivalent average density of the whole battery, is the equivalent average specific heat capacity of the whole battery, is the equivalent thermal conductivity at position ; is the temperature gradient, is the base thermal conductivity, is the temperature gradient attenuation factor, and e is the natural constant, is the divergence operator; Calculate the thermal stress field based on the temperature field, and obtain the additional stress field of concentration gradient based on the lithium ion concentration; calculating the thermal stress field according to the temperature field can clarify the stress situation of the battery due to temperature changes, while obtaining the additional stress field of concentration gradient based on the lithium ion concentration further considers the influence of lithium ion concentration changes on the battery stress. The acquisition of these two stress fields is crucial for evaluating the stability of the internal structure of the battery. During the use of the battery, the change of internal stress may lead to the decline or even damage of the battery performance. By accurately mastering these stress fields, the possible structural problems of the battery can be predicted in advance, providing a more comprehensive basis for temperature regulation, so that the regulation can not only control the temperature, but also ensure the stability of the battery structure.

[0024] Calculate the thermal stress field based on the temperature field, and obtain the additional stress field of concentration gradient based on the lithium ion concentration: ; ; In the formula, is the thermal stress field, E is the elastic modulus, is the coefficient of thermal expansion, is the additional stress field of concentration gradient, is the reference temperature, is the concentration-stress coupling coefficient, is the second-order spatial derivative of the lithium ion concentration; Add the thermal stress field and the additional stress field of concentration gradient to obtain the total stress field : ; Combining the thermal stress field and the additional stress field of concentration gradient into the total stress field comprehensively considers the combined influence of temperature and lithium ion concentration changes on the battery stress.

[0025] Obtain the initial lithium deposition risk cloud map based on the temperature gradient and the total stress field , and the acquisition method is: ; ; ; ; In the formula, is the electrochemical polarization-concentration field coupling term, is 's weighting factor, is the temperature gradient-stress field coupling term, is 's weighting factor, is the SEI film-side reaction coupling term, is The weight factor is the local overpotential at position ; is the critical overpotential; is the second-order spatial derivative of the saturation concentration; is the critical temperature gradient; is the critical total stress; is the position where the SEI film thickness is; is the critical SEI film thickness; is the position where the side reaction rate is; is the reference side reaction rate; is the temperature gradient-stress field coupling factor; The above formula comprehensively considers multi-factors such as the electrochemical polarization-concentration field coupling term, the temperature gradient-stress field coupling term, and the SEI film-side reaction coupling term, comprehensively covering the key factors affecting the lithium deposition risk inside the battery, and can more accurately reflect the lithium deposition risk distribution under the actual state of the battery. For example, considering the local overpotential, critical overpotential, saturation concentration, etc., the electrochemical conditions of lithium deposition can be grasped at the microscopic level, providing an accurate basis for subsequent fault diagnosis and formulating control schemes.

[0026] The calculation process of the initial lithium deposition risk index is to accumulate the initial lithium deposition risk values corresponding to all coordinate points in the initial lithium deposition risk cloud map and then calculate the average value. As a quantitative index, this index can intuitively measure the overall lithium deposition risk degree of the battery. This helps to quickly judge whether the battery is in a high-risk state, providing a clear decision-making criterion for whether temperature regulation is needed and what regulation measures to take.

[0027] When the initial lithium deposition risk index is greater than the set lithium deposition risk threshold, diagnose the battery fault type based on the initial lithium deposition risk cloud map and determine the temperature regulation scheme; In S2, diagnose the battery fault type based on the initial lithium deposition risk cloud map and determine the temperature regulation scheme, including the following steps: Obtain the SEI film thickness data; Combine the SEI film thickness data, the initial lithium deposition risk cloud map, the thermal stress field, and the temperature gradient to construct the battery temperature characteristic data; Input the battery temperature characteristic data into the trained random forest model to output the battery fault type; Determine the thermal runaway probability based on the initial lithium deposition risk index: ; In the formula, is the thermal runaway probability, is the initial lithium deposition risk index; If the thermal runaway probability is greater than the set thermal runaway probability threshold, a temperature control scheme is determined based on the characteristic data before battery runaway, where the characteristic data before battery runaway includes lithium ion concentration gradient data , SEI film thickness data , local overpotential data , temperature gradient data and side reaction rate data (each data in the characteristic data before battery runaway is a set of values corresponding to all coordinate points); If the thermal runaway probability is not greater than the set thermal runaway probability threshold, a fault warning is carried out based on the set warning mechanism.

[0028] The random forest model has powerful classification ability, can handle complex data relationships, quickly and accurately identify the fault types of batteries, help technicians clarify the battery problems, and provide directions for formulating targeted temperature control schemes.

[0029] Determining the temperature control scheme based on the characteristic data before battery runaway includes the following steps: Obtain the scheme pointing data set corresponding to the battery fault type stored in the database. The scheme pointing data set includes several scheme pointing data, and each scheme pointing data corresponds to a temperature control scheme; Compare the characteristic data before battery runaway with each scheme pointing data one by one to obtain the pointing coefficient , determine the scheme pointing data corresponding to the largest pointing coefficient, and obtain the temperature control scheme corresponding to the scheme pointing data, where the scheme pointing data includes lithium ion concentration gradient pointing data , SEI film thickness pointing data , local overpotential pointing data , temperature gradient pointing data and side reaction rate pointing data , and i is the number of the scheme pointing data.

[0030] The method for obtaining the pointing coefficient is as follows: ; In the formula, is the cosine similarity function.

[0031] Through this comparison method, the matching degree between the current state of the battery and each alternative control scheme can be quantified. Screening based on the pointing coefficient can accurately find the scheme that best fits the actual situation of the battery from many schemes, make the temperature control scheme more targeted, avoid blind control, and thus more effectively solve the battery temperature problem and ensure the battery performance.

[0032] In S3, the process of obtaining the feedback index of the control scheme is as follows: After the execution of the temperature control scheme is completed, obtain the characteristic data of the battery after temperature control, including the post-control local overpotential data CU, the post-control side reaction rate data CF, and the optimized amount CY of the lithium deposition risk index after control (the difference between the initial lithium deposition risk index and the lithium deposition risk index after control). These data intuitively reflect the changes in the battery state after the implementation of the temperature control scheme. By monitoring these data, the impact of the control scheme on the key performance indicators of the battery can be directly evaluated, and it can be understood whether the control has achieved the expected effect, providing first-hand information for subsequent judgments.

[0033] Obtain the optimal performance data in the historical execution record of this control scheme, including the post-control optimal local overpotential data TU and the post-control optimal side reaction rate data TF; taking the historical optimal data as a reference, an ideal benchmark is set for the current control effect, which helps to clarify the gap between the current control and the best level, so as to find the improvement direction and continuously optimize the control scheme. Obtain the reference deviation data based on the remaining service life of the battery, including the reference deviation value SU after local overpotential control and the reference deviation value SF after side reaction rate control; based on the characteristic data of the battery after temperature control, the optimal performance data and the reference deviation data, determine the control scheme feedback index FK, and the calculation formula is as follows: ; In the formula, is the cosine similarity function.

[0034] Obtain the reference deviation data based on the remaining service life of the battery, including the following steps: Obtain the historical lithium deposition risk index data of the battery, including the number C of times when the historical lithium deposition risk index is greater than the set lithium deposition risk threshold, the historical maximum lithium deposition risk index Z, the lithium deposition risk index interval value Q1, the average value H of the SEI film thickness, and the SEI film thickness interval value Q2; Input the historical lithium deposition risk index data of the battery into the trained remaining life prediction model to obtain the remaining life of the battery, where the remaining life prediction model is as follows: ; In the formula, is the remaining life, is the rated service life, is the loss of life, is the intercept, and a, b, c, d, f are all regression coefficients; Compare the remaining life of the battery with the mapping relationship between the remaining life-reference deviation data stored in the database to determine the reference deviation data corresponding to the remaining life of the battery.

[0035] Based on regression analysis, this model comprehensively considers the impact of various factors on battery life and can accurately estimate the remaining duration that the battery can be used. This is crucial for reasonably arranging battery maintenance, replacement plans, and adjusting temperature control strategies, avoiding losses caused by sudden battery failure. Since the battery has different sensitivities to temperature control and acceptable deviations at different remaining life stages, the reference deviation data obtained in this way is more in line with the current actual situation of the battery, making the feedback indicators of the subsequent determined control scheme more scientific and reasonable. Thus, it provides a more accurate basis for optimizing the temperature control scheme, ensuring that the battery can receive appropriate temperature control at different usage stages and extending the battery life.

[0036] Determine whether to correct the temperature control scheme based on the feedback indicators of the control scheme. If the feedback indicators of the control scheme are less than the set indicator threshold, then correct the temperature control scheme. The correction process is as follows: Establish a battery model of the current state based on simulation software (simulation software includes but is not limited to tools such as COMSOL, MATLAB, etc.); Establish a multi-objective fitness function DM: ; In the formula, is the initial lithium deposition risk index, and FK is the feedback indicator of the control scheme; Construct multiple candidate temperature control schemes. Treat each candidate temperature control scheme as an individual, and screen the generated multiple candidate temperature control schemes based on the genetic algorithm. Only one individual is retained in each generation (retain the individual with the largest multi-objective fitness function value); After repeating multiple generations, output the optimal individual, and use the candidate temperature control scheme corresponding to this optimal individual to correct the temperature control scheme. Construct multiple candidate temperature control schemes and screen them using the genetic algorithm. Only one individual with the largest multi-objective fitness function value is retained in each generation. The genetic algorithm simulates the natural evolution process, efficiently searches for the optimal solution among numerous candidate schemes, avoids falling into local optima, and enables the selected scheme to achieve a better balance in aspects such as reducing lithium deposition risk and improving control effect, enhancing the quality and effectiveness of the temperature control scheme, and ensuring the stable operation of the battery.

[0037] If the lithium deposition risk index after regulation is greater than the set lithium deposition risk threshold, then correct the temperature control scheme and re-perform temperature control. Here, re-performing temperature control after correcting the temperature control scheme is the same as the above method.

[0038] This embodiment provides a battery temperature control method, which is divided into four stages: 1. Risk assessment: Construct a three-dimensional lithium deposition risk cloud map and calculate the comprehensive risk index; 2. Fault diagnosis: Identify the fault type in combination with the random forest model; 3. Scheme matching: Match the historical optimal temperature control scheme from the database; 4. Closed-loop optimization: Iteratively optimize the control parameters through the genetic algorithm.

[0039] For calculation needs, list some parameters related to battery attributes, as shown in Table 1: Table 1 Some parameters related to battery attributes

[0040] In the first stage, construct a three-dimensional lithium deposition risk cloud map and calculate the comprehensive risk index, as shown in the data of Table 2: Table 2 Comprehensive risk index

[0041] Table 2 shows the parameters related to the lithium deposition risk at different positions inside the battery (dimensionless, only showing the values). The risk cloud map is generated through multi-physics field coupling calculations, providing a key basis for subsequent fault diagnosis and temperature control. The following is a specific analysis: Each row of data in Table 2 represents the three-dimensional coordinates of a certain position inside the battery and the corresponding risk-related parameters, including: lithium ion concentration, which reflects the distribution of lithium ions in the electrode / electrolyte. Excessive concentration is likely to cause deposition. Temperature, local overheating will cause the SEI film to rupture and accelerate side reactions. Driving heat generation rate, which combines Joule heat, side reaction heat, etc., directly affects the temperature field distribution. Total stress, the superposition of thermal stress and concentration gradient stress. Excessive mechanical stress will damage the electrode structure. Local overpotential, the higher the overpotential, the stronger the driving force for lithium ion deposition. SEI film thickness, too thick (>60nm) indicates that side reactions intensify, and too thin (<40nm) is prone to rupture. Risk value, the weighted calculation result of the above parameters, quantifying the lithium deposition risk.

[0042] The data in Table 2 reveals the following rules, verifying the logical rigor of the technical solution: For example, the characteristics of high-risk areas (such as (1, 2, 1)): the highest lithium ion concentration (1350 mol / m³), large concentration gradient, strong deposition driving force; the highest temperature (310K), high temperature intensifies side reactions and reduces the stability of the SEI film. The maximum total stress (15.8 MPa), mechanical stress accelerates electrode damage. The thickest SEI film (65nm), continuous side reactions lead to film thickening and further deteriorate performance. The risk value is 1.12, far exceeding the threshold (the set value is usually 1.0), and priority control is required.

[0043] The characteristics of low-risk areas (such as (2, 1, 1)): low lithium ion concentration (950 mol / m³), low deposition risk; low temperature (298K), small stress (8.2 MPa), thin SEI film (40nm), and the parameters are all within the safe range. The risk value is 0.42, and no additional control is required.

[0044] Correlation between parameters: The lithium-ion concentration is positively correlated with the temperature, and the heat generation rate is higher in the high-concentration region. The SEI film thickness is positively correlated with the side reaction rate, which is consistent with the mechanism that the side reaction causes the film to thicken.

[0045] The high-risk areas (such as (1, 2, 1)) can be identified through Table 2 to guide the temperature control plan for priority cooling or uniform temperature treatment. Parameters such as the driving heat generation rate (2000 W / m³) and temperature gradient (100 K / m) in Table 1 are used to match the historical optimal control plan from the database.

[0046] The data in Table 2 verifies the accuracy of the multi-physical field coupling model, reveals the spatial distribution law of lithium deposition risk, and directly supports the core steps such as fault diagnosis, plan matching, and dynamic optimization. By quantifying the risk level, this table provides the operable control priority and parameter adjustment direction for engineering applications and is the data cornerstone of the whole set of methods.

[0047] In the second stage, the random forest model is combined to identify the fault types.

[0048] Definition of fault types: Type A (local overheating), Type B (SEI film rupture), Type C (lithium dendrite growth), Type D (normal state).

[0049] Taking the characteristic parameters in Table 3 as an example, they are used as training parameters: Table 3 Range of Characteristic Parameters

[0050] Among them, the range of the side reaction rate uses scientific notation.

[0051] Taking 100 decision trees and a maximum tree depth of 10 layers as an example; the feature selection criterion is Gini impurity, and the data is divided into 5-fold cross-validation.

[0052] Total number of samples: 400 groups (4 categories × 100 groups); Training set: 70% (280 groups); Test set: 30% (120 groups).

[0053] Taking Type A (local overheating) as an example, the test results are shown in Table 4: Table 4 Test Results

[0054] From the content of Table 4, calculate the following parameters: True positive (TP): 27; False positive (FP): 0 (the number of times other types are misjudged as A); False negative (FN): 1 (A misjudged as B) + 1 (A misjudged as C) + 1 (A misjudged as D) = 3; Precision (A): 1; Global accuracy: 0.9.

[0055] Statistical results of other types of tests are shown in Table 5 as follows: Table 5 Diagnostic performance data

[0056] The overall accuracy of the model > 85%, indicating that the random forest can effectively integrate multi-dimensional features such as temperature gradient and SEI film thickness to support the scheme for diagnosing battery fault types.

[0057] In the third stage, match the historical optimal temperature control scheme in the database.

[0058] Taking Scheme 1 (high-gradient cooling + SEI film repair), Scheme 2 (isothermal control), and Scheme 3 (medium-gradient cooling + side reaction inhibition) as examples, the implementation effects are shown in Table 6 as follows: Table 6 Implementation effect data

[0059] As can be seen from Table 6, Scheme 1 has the highest risk reduction rate (60%), indicating that its parameter setting is highly matched with the historical optimal scheme, verifying the effectiveness of the feature matching logic. Scheme 2 has the lowest risk reduction rate (25%), and due to conservative parameters (such as the isothermal range of ±2°C), it cannot effectively suppress local risks. Scheme 3 balances cooling and side reaction inhibition, and the risk reduction rate is in the middle (40%).

[0060] In the fourth stage, iteratively optimize the control parameters through the genetic algorithm.

[0061] The parameter settings of the genetic algorithm are shown in Table 7 as follows: Table 7 Genetic algorithm parameter settings

[0062] The influence of the number of iterations on the optimal solution (taking the cooling intensity and isothermal range as examples) is shown in Table 8: Table 8 Influence of the number of iterations on the optimal solution

[0063] After 50 iterations, the fitness growth slows down (from 1.50 to 1.64), indicating that it is approaching the optimal solution. The fitness DM increases as the risk index decreases.

[0064] Example 2: As Figure 3 shown, a battery temperature control system for implementing the method in Example 1 includes: An initial lithium deposition risk index acquisition module for obtaining an initial lithium deposition risk cloud map and calculating an initial lithium deposition risk index; A fault type and control scheme determination module, configured to diagnose the battery fault type and determine a temperature control scheme based on an initial lithium deposition risk cloud map when the initial lithium deposition risk index is greater than a set lithium deposition risk threshold; A control scheme feedback module, configured to obtain a post-regulation lithium deposition risk index and a control scheme feedback index after regulation; A control scheme correction module, configured to end the regulation when the post-regulation lithium deposition risk index is not greater than the set lithium deposition risk threshold, and determine whether to correct the temperature control scheme based on the control scheme feedback index; or when the post-regulation lithium deposition risk index is greater than the set lithium deposition risk threshold, correct the temperature control scheme and then perform temperature regulation again.

Claims

1. A battery temperature control method, characterized in that: The following steps are involved: S1. Obtain an initial lithium deposition risk cloud map and calculate an initial lithium deposition risk index; S2. When the initial lithium deposition risk index is greater than the set lithium deposition risk threshold, diagnose the battery failure type and determine the temperature control scheme based on the initial lithium deposition risk cloud map; S3. Execute the temperature control plan, and obtain the lithium deposition risk index after control and the feedback index of the control plan: If the lithium deposition risk index after regulation is not greater than the set lithium deposition risk threshold, the regulation ends, and whether to modify the temperature regulation scheme is determined based on the regulation scheme feedback index; If the lithium deposition risk index after control is greater than the set lithium deposition risk threshold, the temperature control scheme is corrected and the temperature control is performed again.

2. A battery temperature control method according to claim 1, characterized in that: In the above-mentioned S1, obtaining the initial lithium deposition risk cloud map comprises the following steps: The driving heat generation rate is obtained based on the lithium ion concentration: ； ； In the formula, is the position at time t The lithium ion concentration at is the time step, is the time step Internal position The change of lithium ion concentration at For location The driving heat generation rate at , and are all transfer functions. For location The open circuit voltage at is the terminal voltage, is the charge and discharge current, is the position at time t The temperature at is the average temperature of the battery as a whole, is the thermal resistance coefficient of the SEI film side reaction, V is the battery volume; The temperature field model is constructed by driving the heat generation rate and outputting the temperature gradient: ； ； In the formula, is the equivalent average density of the entire battery, is the equivalent average specific heat capacity of the battery as a whole, For location The equivalent thermal conductivity at is the temperature gradient, is the basic thermal conductivity, is the temperature gradient attenuation factor, e is a natural constant, is the divergence operator; The thermal stress field is calculated based on the temperature field, and the concentration gradient additional stress field is obtained based on the lithium ion concentration: ； ； In the formula, is the thermal stress field, E is the elastic modulus, is the coefficient of thermal expansion, Add the stress field for the concentration gradient, is the reference temperature, is the concentration-stress coupling coefficient, is the second-order spatial derivative of lithium ion concentration; The total stress field is obtained by adding the thermal stress field to the concentration gradient additional stress field. : ； The initial lithium deposition risk cloud map was obtained based on the temperature gradient and total stress field.

3. A battery temperature control method according to claim 2, characterized in that: Obtaining initial lithium deposition risk cloud map based on temperature gradient and total stress field , and the way to obtain it is: ； ； ； ； In the formula, is the electrochemical polarization-concentration field coupling term, for The weight factor of is the temperature gradient-stress field coupling term, for The weight factor of is the SEI film-side reaction coupling term, for The weight factor of For location At the local overpotential, is the critical overpotential, is the second-order spatial derivative of the saturation concentration, is the critical temperature gradient, is the critical total stress, For location SEI film thickness at is the critical SEI film thickness, For location The side reaction rate, is the benchmark side reaction rate, is the temperature gradient-stress field coupling factor; The calculation process of the initial lithium deposition risk index is to accumulate the initial lithium deposition risk values ​​corresponding to all coordinate points in the initial lithium deposition risk cloud map and then calculate the average value.

4. A battery temperature control method according to claim 2, characterized in that: In the above S2, based on the initial lithium deposition risk cloud map, the battery failure type is diagnosed and the temperature control scheme is determined, including the following steps: Obtain SEI film thickness data; The SEI film thickness data, initial lithium deposition risk cloud map, thermal stress field and temperature gradient are combined to construct battery temperature characteristic data; Input the battery temperature feature data into the trained random forest model and output the battery fault type; Determine the probability of thermal runaway based on the initial lithium deposition risk index: ； In the formula, is the probability of thermal runaway, is the initial lithium deposition risk index; If the thermal runaway probability is greater than the set thermal runaway probability threshold, the temperature control scheme is determined based on the battery runaway characteristic data before the runaway, where the battery runaway characteristic data before the runaway includes the lithium ion concentration gradient data. , SEI film thickness data , local overpotential data , temperature gradient data and side reaction rate data ; If the thermal runaway probability is not greater than the set thermal runaway probability threshold, a fault warning is performed based on the set warning mechanism.

5. A battery temperature control method according to claim 4, characterized in that: Determining a temperature control scheme based on the characteristic data before the battery runs out of control includes the following steps: Obtain a solution pointing data set corresponding to the battery fault type stored in the database, where the solution pointing data set includes a plurality of solution pointing data, and each solution pointing data corresponds to a temperature control solution; Compare the characteristic data before the battery loses control with the pointing data of each solution one by one to obtain the pointing coefficient , determine the scheme pointing data corresponding to the maximum pointing coefficient, and obtain the temperature control scheme corresponding to the scheme pointing data, wherein the scheme pointing data includes the lithium ion concentration gradient pointing data , SEI film thickness pointing data , local overpotential pointing data , Temperature gradient pointing data and side reaction rate pointing data , i is the number of the data pointed to by the scheme.

6. A battery temperature control method according to claim 5, characterized in that: The method to obtain the pointing coefficient is as follows: ； In the formula, is the cosine similarity function.

7. A battery temperature control method according to claim 1, characterized in that: In S3, the process of obtaining the feedback indicator of the control scheme is as follows: After the temperature control scheme is executed, characteristic data of the battery after temperature control is obtained, including local overpotential data CU after control, side reaction rate data CF after control, and lithium deposition risk index optimization value CY after control; Obtain the optimal performance data in the historical execution record of the temperature control scheme, including the optimal local overpotential data TU after control and the optimal side reaction rate data TF after control; Obtain reference deviation data based on the remaining service life of the battery, including a reference deviation value SU after local overpotential regulation and a reference deviation value SF after side reaction rate regulation; Based on the characteristic data after battery temperature control, the optimal performance data and the reference deviation data, the control scheme feedback index FK is determined. The calculation formula is as follows: ； In the formula, is the cosine similarity function.

8. A battery temperature control method according to claim 7, characterized in that: Obtaining reference deviation data based on the remaining battery life includes the following steps: Obtain the historical lithium deposition risk index data of the battery, including the number of times C that the historical lithium deposition risk index is greater than the set lithium deposition risk threshold, the historical maximum lithium deposition risk index Z, the lithium deposition risk index interval value Q1, the SEI film thickness average value H and the SEI film thickness interval value Q2; The battery historical lithium deposition risk index data is input into the trained remaining life prediction model to obtain the remaining battery life, where the remaining life prediction model is as follows: ； In the formula, For the remaining life, is the rated service life, For loss of life, is the intercept, a, b, c, d, and f are all regression coefficients; The remaining life of the battery is compared with the remaining life-reference deviation data mapping relationship stored in the database to determine the reference deviation data corresponding to the remaining life of the battery.

9. A battery temperature control method according to claim 1, characterized in that: In S3, whether to modify the temperature control scheme is determined based on the control scheme feedback index. If the control scheme feedback index is less than the set index threshold, the temperature control scheme is modified. The modification process is as follows: Establish a battery model of the current state based on simulation software; Establish a multi-objective fitness function DM: ； In the formula, is the initial lithium deposition risk index, and FK is the feedback indicator of the control scheme; Construct multiple candidate temperature control schemes, regard each candidate temperature control scheme as an individual, screen the generated multiple candidate temperature control schemes based on the genetic algorithm, and only retain one individual in each generation; After repeating for multiple generations, the optimal individual is output, and the temperature control plan is corrected using the candidate temperature control plan corresponding to the optimal individual.

10. A battery temperature control system, used to implement the method according to any one of claims 1 to 9, characterized in that: include: An initial lithium deposition risk index acquisition module is used to acquire an initial lithium deposition risk cloud map and calculate an initial lithium deposition risk index; A fault type and control scheme determination module is used to diagnose the battery fault type and determine the temperature control scheme based on the initial lithium deposition risk cloud map when the initial lithium deposition risk index is greater than the set lithium deposition risk threshold; A control scheme feedback module is used to obtain a post-control lithium deposition risk index and a control scheme feedback indicator after control; A control scheme correction module, which ends the control when the lithium deposition risk index after control is not greater than the set lithium deposition risk threshold, and determines whether to correct the temperature control scheme based on the control scheme feedback index; Or when the lithium deposition risk index after regulation is greater than the set lithium deposition risk threshold, the temperature control scheme is corrected and the temperature control is performed again.

Images