Thermal runaway early warning threshold calculation method and device and thermal runaway early warning method based on fuzzy logic

Abstract

The invention discloses a thermal runaway early warning threshold calculation method and device and a thermal runaway early warning method based on fuzzy logic, and the calculation method comprises the steps: representing parameters in a to-be-optimized parameter set of a fuzzy membership function through an antibody in an initial population, and carrying out the iterative optimization of the initial population through an immune clone selection algorithm, obtaining a global optimal solution of the parameter set; determining an optimal fuzzy membership function based on the global optimal solution of the parameter set; mapping different environment temperatures and charging rates to different fuzzy intervals through the optimal fuzzy membership function, outputting early warning thresholds under different environment temperature and charging rate combination conditions, and constructing corresponding relations between different combination conditions and the early warning thresholds; the problem of migration of the thermal runaway early warning threshold under different environment temperatures and charging rates is solved.

Description

Technical Field

[0001] The present invention relates to the field of battery safety technology, and in particular to a thermal runaway warning threshold calculation method and device, and a thermal runaway warning method based on fuzzy logic. Background Art

[0002] In recent years, the energy storage sector has shown a rapid growth trend, which has promoted the upgrading and expansion of the entire green energy system. Various clean technologies, especially lithium-ion battery technology, are rapidly making breakthroughs. However, the frequent occurrence of battery fires and explosions has also brought severe challenges to the healthy development of the new energy industry. The safety of lithium batteries has become the main bottleneck in the development of the industry. The most serious safety issue of lithium batteries is thermal runaway, and the causes of thermal runaway can be summarized as electrical abuse, thermal abuse, and mechanical abuse. These improper use behaviors will cause internal short circuits in the battery, causing a sharp rise in temperature, which in turn triggers a series of side reactions and ultimately triggers thermal runaway. To ensure safety, it is required to issue a warning at least five minutes before thermal runaway occurs to provide sufficient time for personnel evacuation and escape.

[0003] So far, there are many methods for thermal runaway warning of lithium-ion batteries, most of which are based on the fusion of multiple signals. The fused signals usually include temperature, voltage, gas and other signals. For example, early warning of temperature or temperature rise rate is given, and an alarm is sounded when the signal reaches the corresponding threshold. In addition, existing research has also confirmed that lithium batteries will release a large amount of gas during thermal runaway, and the first gas to appear is usually H 2 Therefore, there is also a method of using gas signals for early warning based on this characteristic. For example, in the relevant technology, a Chinese patent application document with publication number CN118472446A discloses a thermal runaway safety early warning method for lithium batteries. The scheme determines the level range of environmental data based on a preset threshold range level, and determines the early warning level based on the threshold; however, the threshold is set to static in the scheme, ignoring that the safety performance of lithium batteries under different ambient temperatures and charging rates varies greatly, and the threshold cannot be generalized. A Chinese patent application document with publication number CN117810576A discloses a comprehensive thermal runaway alarm method for lithium-ion battery energy storage power stations. In the scheme, the position of the battery in the charge and discharge interval is first determined based on the temperature, and then the level of the early warning is determined based on the preset temperature, voltage, and gas thresholds; however, the scheme also ignores that the safety performance of lithium batteries under different ambient temperatures and charging rates varies greatly, and a preset fixed threshold cannot be used to complete a static early warning.

[0004] In addition, a Chinese patent application with publication number CN116093497A discloses a method for predicting the probability of thermal runaway of a battery. In this scheme, the battery temperature and the location of thermal runaway of the battery are used as fuzzy inputs to fuzzy output the probability of thermal runaway of the battery, thereby predicting the probability of thermal runaway. In non-patent literature, "Immune clonal selection algorithm based on fuzzy sets, Computer Technology and Development, Vol. 17, No. 12, Zhang Kui et al." proposed an immune clonal selection algorithm based on fuzzy set theory, which introduced the concepts of fuzzy sets and membership, and adopted a dynamic intelligent optimization strategy, which effectively improved the characteristics of the detection element and improved the adaptability of the detection element in a complex network environment; however, this document mainly focuses on optimizing the rule weights or parameter distribution of the existing fuzzy system, and the goal is usually to enhance the prediction accuracy or classification ability of the system, but does not focus on the shape of the membership function. Summary of the invention

[0005] The technical problem to be solved by the present invention is the migration problem of thermal runaway warning threshold under different ambient temperatures and charging rates.

[0006] The present invention solves the above technical problems by the following technical means: In a first aspect, a method for calculating a thermal runaway warning threshold is proposed, the method comprising: The parameters in the parameter set to be optimized of the fuzzy membership function are represented by antibodies in the initial population, and the initial population is iteratively optimized using the immune clonal selection algorithm to obtain the global optimal solution of the parameter set; Determining an optimal fuzzy membership function based on a global optimal solution of the parameter set; Different ambient temperatures and charging rates are mapped to different fuzzy intervals through the optimal fuzzy membership function, warning thresholds under different combinations of ambient temperatures and charging rates are output, and corresponding relationships between different combinations of conditions and warning thresholds are constructed.

[0007] Furthermore, the parameters in the parameter set to be optimized of the fuzzy membership function are represented by antibodies in the initial population, and the initial population is iteratively optimized using an immune clonal selection algorithm to obtain a global optimal solution for the parameter set, including: S11, randomly generating an initial population, each antibody in the initial population represents a parameter of the fuzzy membership function; S12, using the fitness function to evaluate each antibody, and determine whether the evaluation result meets the termination condition, if not, execute step S13, if yes, execute step S16; S13, selecting high-quality antibodies based on the antibody evaluation results, and cloning the selected antibodies to obtain cloned antibodies; S14, performing a mutation operation on the cloned antibody, evaluating the mutated antibody using the fitness function, and competing with the antibodies in the initial population, retaining the current optimal antibody to the next generation, and forming the next generation population; S15, re-execute steps S12 to S14 for the next generation population; S16. Output the global optimal solution of the parameter set of the fuzzy membership function.

[0008] Furthermore, the fitness function is expressed as:

[0009]

[0010] In the formula, They represent the first The strain turning point, Voltage turning point; They represent the experimental measurements. The strain turning point, Voltage turning point; Indicates antibody Adaptability; Indicates the population size; , They represent the difference between the predicted strain turning point and the actual strain turning point, and the difference between the predicted voltage turning point and the actual voltage turning point, respectively.

[0011] Furthermore, the method further comprises: When the selected antibody is cloned, the number of clones is proportional to the fitness of the selected antibody.

[0012] Furthermore, the method further comprises: When performing mutation operations on cloned antibodies, random perturbations are introduced as follows:

[0013] In the formula, Indicates the mutated The first decision variables; Indicates the number before mutation The first decision variables; Indicates multiplication; represents a standard normal distribution random number; indicates that the strength of mutation is usually inversely proportional to the antibody fitness, , Indicates The fitness value of an antibody, It indicates the maximum fitness value in the current population and is used for normalization to make the variation range relatively uniform.

[0014] Furthermore, the termination condition is expressed as:

[0015] In the formula, represents the next generation population; It means that the antibody with the best fitness is selected from the combination of the original population and the clone population to form a new generation population; represents the initial population; represents a union; Represents the population after clonal mutation.

[0016] Furthermore, the mapping of different ambient temperatures and charging rates to different fuzzy intervals through the optimal fuzzy membership function, outputting warning thresholds under different combinations of ambient temperatures and charging rates, and constructing corresponding relationships between different combinations of conditions and warning thresholds include: Conducting an overcharge thermal runaway experiment on a lithium battery under different combinations of ambient temperature and charging rate conditions, and continuously recording experimental parameters, wherein the experimental parameters include a voltage change signal, a battery surface temperature change signal, a strain change signal, and a gas change signal; The ambient temperature and charging rate in each combination condition are mapped to different fuzzy intervals through the optimal fuzzy membership function, and the warning thresholds under different combinations of ambient temperature and charging rate are obtained based on the fuzzy rules established according to the experimental data.

[0017] Furthermore, the method further comprises: The ambient temperature and charging rate in each combination condition are normalized and then mapped to different fuzzy intervals through the optimal fuzzy membership function.

[0018] Furthermore, the warning threshold includes a voltage turning point and a strain turning point.

[0019] Furthermore, types of the fuzzy membership function include trigonometric function and trapezoidal function.

[0020] In a second aspect, the present invention provides a thermal runaway warning threshold calculation device, the device comprising: The parameter iteration optimization module is used to represent the parameters in the parameter set to be optimized of the fuzzy membership function using antibodies in the initial population, and to iteratively optimize the initial population using the immune clonal selection algorithm to obtain the global optimal solution of the parameter set; An optimal fuzzy membership function determination module, used to determine an optimal fuzzy membership function based on a global optimal solution of the parameter set; The warning threshold migration module is used to map different ambient temperatures and charging rates to different fuzzy intervals through the optimal fuzzy membership function, output the warning thresholds under different combinations of ambient temperatures and charging rates, and construct the corresponding relationship between different combinations of conditions and warning thresholds.

[0021] In a third aspect, the present invention proposes a thermal runaway early warning method based on fuzzy logic, the method comprising: Real-time collection of lithium battery status parameters under certain working conditions; Comparing the state parameter with the corresponding warning threshold under the current working condition to issue a thermal runaway warning; The warning threshold is a warning threshold calculated based on the thermal runaway warning threshold calculation method as described above and corresponding to the ambient temperature and charging rate under the current working conditions.

[0022] Further, the state parameter includes a voltage signal and a strain signal, the warning threshold includes a voltage turning point and a strain turning point, and the state parameter is compared with the corresponding warning threshold under the current working condition to perform thermal runaway warning, including: comparing the voltage signal and the voltage turning point with the strain signal and the strain turning point respectively; When the voltage signal is greater than the voltage turning point and the strain signal is greater than the strain turning point, a first-level thermal runaway warning is performed.

[0023] Furthermore, the state parameter also includes a gas signal, and the method further includes: When the strain decrease rate is greater than the set first threshold and the gas concentration increase rate is greater than the set second threshold, a second-level thermal runaway warning is issued.

[0024] Furthermore, the state parameter also includes a battery surface temperature signal, and the method further includes: When the voltage drop rate is greater than the set third threshold and the battery surface temperature rise rate is greater than the set fourth threshold, a third-level thermal runaway warning is issued.

[0025] The advantages of the present invention are: (1) The present invention first uses an immune optimization algorithm to optimize the parameters of the fuzzy membership function to obtain the optimal fuzzy membership function, and then uses the optimal fuzzy membership function to construct the corresponding relationship between different ambient temperature and charging rate combination conditions and the warning threshold. The dynamic threshold that changes dynamically with different ambient temperatures and charging rates is obtained instead of the static threshold set in the related art, which solves the migration problem of the thermal runaway warning threshold under different ambient temperatures and charging rates.

[0026] (2) The present invention optimizes the optimal fuzzy membership function through the immune optimization algorithm, which can effectively reflect the influence of ambient temperature and charging rate on the risk of thermal runaway, ensuring the accuracy of the early warning threshold calculation while also improving the accuracy and response speed of the membership function to the system.

[0027] (3) When the present invention performs thermal runaway warning for lithium batteries, the real-time collected lithium battery status parameters are compared with the warning threshold to achieve thermal runaway warning, and the warning threshold is the voltage turning point and the strain turning point calculated by the fuzzy membership function according to the ambient temperature and charging rate under the current working conditions. Therefore, the warning threshold used in the present invention is not a static threshold, but a dynamic threshold that changes dynamically with different ambient temperatures and charging rates, which solves the migration problem of the thermal runaway warning threshold under different ambient temperatures and charging rates; therefore, by selecting a warning threshold that is suitable for the ambient temperature and charging rate under the current working conditions to perform thermal runaway warning, the accuracy of the thermal runaway warning can be greatly improved.

[0028] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flow chart of a method for calculating a thermal runaway warning threshold value proposed in one embodiment of the present invention; Figure 2 It is a schematic diagram of a process of optimizing parameters by using immune clonal selection algorithm in one embodiment of the present invention; Figure 3 is a schematic diagram of antibody fitness convergence in one embodiment of the present invention; Figure 4 It is a schematic diagram of the structure of an experimental device in one embodiment of the present invention; Figure 5 is a block diagram of the fuzzy logic system in one embodiment of the present invention; Figure 6 is a schematic diagram of fuzzy logic output in one embodiment of the present invention; Figure 7 1 is a schematic diagram of a thermal runaway warning threshold calculation and verification process in one embodiment of the present invention; Figure 8 2 is a schematic diagram of the calculation and verification result of the thermal runaway warning threshold in one embodiment of the present invention; Fig. 9 It is a structural schematic diagram of a thermal runaway warning threshold calculation device proposed in one embodiment of the present invention; Fig.10 is a flow chart of a thermal runaway early warning method based on fuzzy logic proposed in one embodiment of the present invention; Fig.11It is a complete flow chart of a thermal runaway early warning method based on fuzzy logic in one embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] Example 1 like Figure 1 As shown, the first embodiment of the present invention proposes a method for calculating a thermal runaway warning threshold, the method comprising the following steps: S10, using antibodies in the initial population to represent the parameters in the parameter set to be optimized of the fuzzy membership function, and using an immune clonal selection algorithm to iteratively optimize the initial population to obtain a global optimal solution for the parameter set; It should be noted that the parameters of the fuzzy membership function include shape parameters, location parameters, width parameters, slope parameters, height parameters, etc. It is understandable that the number and meaning of specific parameters may vary depending on different types of membership functions.

[0032] This embodiment targets the lithium battery thermal runaway warning scenario, focuses on solving the shape problem of the membership function, and is oriented to the fuzzy system design problem for complex, multi-input variable scenarios. It is particularly suitable for tasks that require high reliability, such as lithium battery safety performance evaluation.

[0033] S20, determining an optimal fuzzy membership function based on the global optimal solution of the parameter set; S30, mapping different ambient temperatures and charging rates to different fuzzy intervals through the optimal fuzzy membership function, outputting warning thresholds under different combinations of ambient temperatures and charging rates, and constructing corresponding relationships between different combinations of conditions and warning thresholds.

[0034] In view of the fact that a static warning threshold is set in the existing thermal runaway warning technology and the differences in the safety performance of lithium batteries under different ambient temperatures and charging rates are ignored, this embodiment provides a thermal runaway warning method based on fuzzy logic. Firstly, the immune optimization algorithm is used to optimize the parameters of the fuzzy membership function to obtain the optimal fuzzy membership function. Then, the optimal fuzzy membership function is used to construct the correspondence between different ambient temperature and charging rate combination conditions and the warning threshold. The dynamic threshold that changes dynamically with different ambient temperatures and charging rates is obtained instead of the static threshold set in the related technology, which solves the migration problem of the thermal runaway warning threshold under different ambient temperatures and charging rates.

[0035] As a further preferred technical solution, Figure 2 As shown, the step S10: using the antibodies in the initial population to represent the parameters in the parameter set to be optimized of the fuzzy membership function, and using the immune clonal selection algorithm to iteratively optimize the initial population to obtain the global optimal solution of the parameter set, specifically includes the following steps: S11, randomly generate an initial population, each antibody represents a parameter of the fuzzy membership function; Specifically, we first randomly generate an initial population, where each antibody represents a potential solution for a parameter. , ,…, }, where each antibody represents a solution to the problem, and the population size is N. The value range of the solution is determined by the upper and lower bounds [ , ]Decide:

[0036] Wherein, D is the dimension of the solution, which is determined according to the parameters required by the fuzzy membership function. In the embodiment of the present invention, D is set to 22.

[0037] S12, using the fitness function to evaluate each antibody, and determine whether the evaluation result meets the termination condition, if not, execute step S13, if yes, execute step S16; As a further preferred technical solution, the formula of the termination condition is expressed as:

[0038] In the formula, represents the next generation population; It means that the antibody with the best fitness is selected from the combination of the original population and the clone population to form a new generation population; represents the initial population; represents a union; Represents the population after clonal mutation.

[0039] S13, selecting high-quality antibodies based on the antibody evaluation results, and cloning the selected antibodies to obtain cloned antibodies; It should be noted that this embodiment uses the fitness function to evaluate the performance of each antibody and select high-quality antibodies therefrom.

[0040] S14, performing a mutation operation on the cloned antibody, evaluating the mutated antibody using the fitness function, and competing with the antibodies in the initial population, retaining the current optimal antibody to the next generation, and forming the next generation population; It should be noted that the mutated antibodies will be evaluated again through the fitness function and compete with the antibodies in the original population to form the next generation population. In order to retain the optimal solution, the elite retention strategy is usually adopted, that is, the current optimal antibody is directly retained to the next generation. This process will be iterated continuously until the stopping condition is reached and finally converged to obtain the global optimal solution. The fitness result is as follows Figure 3 shown.

[0041] S15, re-execute steps S12 to S14 for the next generation population; S16. Output the global optimal solution of the parameter set of the fuzzy membership function.

[0042] It should be noted that the immune clone selection algorithm is an optimization algorithm inspired by the biological immune system. Its core idea comes from the mechanism of antibody recognition of antigens in the immune system. The algorithm simulates the clone selection process of the immune system and gradually finds the global optimal solution by continuously screening, cloning and mutating the solution. In this embodiment, an antibody in the initial population represents a potential solution of a parameter of the fuzzy membership function, and the initial population is iteratively optimized to determine the global optimal solution of the parameter set of the fuzzy membership function.

[0043] As a further preferred technical solution, for each antibody Calculate its fitness In this embodiment, the fitness is the error between the predicted label and the true label, that is, the difference between the strain turning point and the voltage turning point predicted by the fuzzy system and the strain turning point and the voltage turning point obtained by the experiment. The fitness function formula is expressed as:

[0044]

[0045] In the formula, They represent the first The strain turning point, Voltage turning point; They represent the experimental measurements. The strain turning point, Voltage turning point; Indicates antibody Adaptability; Indicates the population size; , They represent the difference between the predicted strain turning point and the actual strain turning point, and the difference between the predicted voltage turning point and the actual voltage turning point, respectively.

[0046] As a further preferred technical solution, the method further includes: when performing a cloning operation on the selected antibody, the number of clones is proportional to the fitness of the selected antibody. , for The number of clones is increased to ensure that the high-quality solution is further expanded.

[0047] As a further preferred technical solution, the method further includes: when performing mutation operation on the cloned antibody, introducing random perturbation, thereby increasing the diversity of the population and avoiding falling into a local optimal solution, wherein the random perturbation is:

[0048] In the formula, Indicates the mutated The first variables; Indicates the number before mutation The first decision variables; Indicates multiplication; represents a standard normal distribution random number; indicates that the strength of mutation is usually inversely proportional to the antibody fitness, , Indicates The fitness value of an antibody, It indicates the maximum fitness value in the current population and is used for normalization to make the variation range relatively uniform.

[0049] Furthermore, the global optimal solution of the parameter set determined by the immune clone selection algorithm in this embodiment is shown in Table 1, Table 2, Table 3 and Table 4 below: Table 1 Specific parameters of input charging rate

[0050] Table 2 Specific parameters of input ambient temperature

[0051] Table 3 Specific parameters of output strain turning point

[0052] Table 4 Specific parameters of output voltage turning point and charging rate

[0053] It should be noted that in a fuzzy control system, choosing a suitable membership function is crucial to the accuracy and response speed of the system. When choosing a membership function, the following factors should generally be considered: The actual needs of the system, the shape and number of membership functions should match the characteristics of the system. For example, if the system is sensitive to input changes, a narrower membership function can be selected; conversely, a wider function can be used. For the battery thermal runaway warning research, the membership function selected in this study should be able to effectively reflect the impact of temperature and charging rate on thermal runaway risk.

[0054] Different membership functions vary in computational complexity. For example, trigonometric and trapezoidal functions are simple to compute and suitable for real-time systems, while Gaussian and Bell functions are more computationally complex but can provide smoother transitions. In situations where real-time requirements are high, simple membership functions may be a better choice.

[0055] As a further preferred technical solution, the step S30: mapping different ambient temperatures and charging rates to different fuzzy intervals through the optimal fuzzy membership function, outputting warning thresholds under different combinations of ambient temperatures and charging rates, and constructing a corresponding relationship between different combinations of conditions and warning thresholds, includes the following steps: S31. Perform an overcharge thermal runaway experiment on the lithium battery under different combinations of ambient temperature and charging rate, and continuously record the experimental parameters, which include voltage change signals, battery surface temperature change signals, strain change signals, and gas change signals.

[0056] This embodiment uses a commercial soft-pack lithium iron phosphate battery as an example to conduct an overcharge thermal runaway experiment, wherein the specific parameters of the lithium battery are shown in Table 5: Table 5 Commercial soft pack lithium iron phosphate battery parameters

[0057] Specifically, the overcharge thermal runaway experiment is as follows: (1) Overcharge thermal runaway test was carried out at 0.5C, 1C and 2C charging rates at 25℃, 50℃ and 75℃ respectively: 1-1) When conducting an overcharge thermal runaway experiment, the experimental equipment should be initialized first. This includes the initialization of the charging and discharging device, explosion-proof device, voltage signal, surface temperature signal, strain signal, and gas signal acquisition equipment. The safety measures of the experimental environment should also be checked at the same time to ensure that the equipment operates stably and the data during the experiment can be collected accurately. Specific equipment such as Figure 4 shown.

[0058] 1-2) Pre-treat the battery so that the battery SOC is 100%. Before the experiment begins, use constant current and constant voltage charging to pre-treat the battery to ensure that the battery SOC reaches 100%. Specifically, first discharge at a constant current of 1C until the battery voltage reaches a cut-off voltage of 2.0V. After the constant current discharge stage is over, stand still for 10 minutes and start constant current and constant voltage charging. First, it is constant current charging, charging at a charging rate of 1C until the battery reaches a cut-off voltage of 3.65V, and then enters the constant voltage stage until the battery charging current reaches about 0.05C, and the charging process is considered to be over. During the charging process, it is necessary to monitor parameters such as voltage, current and temperature to ensure that the battery is within a safe range during the pre-treatment stage, so as to provide consistent initial conditions for subsequent overcharge thermal runaway experiments.

[0059] 1-3) Set different ambient temperatures and charging rates to conduct overcharge thermal runaway experiments. The experiment will be conducted under three typical ambient temperature conditions: 25°C, 50°C and 75°C, representing normal temperature, high temperature and extreme high temperature environments respectively. These three temperature settings are used to simulate the operating status of the battery under daily use, outdoor high temperature exposure and extreme working conditions.

[0060] The experiment also needs to be carried out at different charging rates to simulate the risk of thermal runaway of the battery under different charging conditions. The charging rate will be set to 0.5C, 1C and 2C, corresponding to slow charging, standard charging and fast charging respectively. The setting of the charging rate helps to understand the impact of the charging current on the internal reaction rate and thermal runaway tendency of the battery.

[0061] Before the experiment begins, the laboratory environment temperature must be adjusted to the set value (25°C, 50°C or 75°C). After the environment temperature stabilizes, place the battery in the experimental environment and ensure that the temperature between the battery and the environment reaches a balanced state, which usually requires a certain amount of static time (about 30 minutes). Monitor the surface temperature of the battery in real time to ensure that the battery is consistent with the experimental environment temperature.

[0062] The battery will be charged at a set rate using a constant current and constant voltage charging mode (as described in step 1-2). After reaching SOC 100%, the overcharge operation will continue at the set rate until the battery triggers thermal runaway.

[0063] (2) Collect data such as battery surface temperature, voltage, strain, gas concentration, etc. during overcharge thermal runaway like Figure 4 As shown in the figure, during the entire overcharge experiment, the data collection device, including the voltage signal, surface temperature signal, strain signal, and gas signal acquisition device, will continuously record the changes in battery surface temperature, voltage, strain, and gas concentration as the overcharge progresses. The data under each combination of temperature and charging rate are classified and stored for subsequent feature analysis, such as specifying fuzzy rules and verifying whether the designed fuzzy system is reasonable.

[0064] S32, mapping the ambient temperature and charging rate in each combination condition to different fuzzy intervals through the optimal fuzzy membership function, and obtaining the warning thresholds under different combinations of ambient temperature and charging rate based on the fuzzy rules established according to the experimental data.

[0065] In order to adapt to the differences in battery safety performance under different ambient temperatures and charging rates and the different thermal runaway thresholds, this embodiment uses the ambient temperature and charging rate as the input of the optimal fuzzy membership function, and the fuzzy rules established according to the experimental data are mapped to different fuzzy intervals through the fuzzy membership function. At the same time, in order to adapt to the different thermal runaway warning thresholds under different situations, the output is the strain turning point and the voltage turning point.

[0066] It should be noted that the fuzzy membership function is used to describe the degree of belonging of an input variable in a fuzzy set, that is, the degree to which the variable belongs to a fuzzy set. This kind of belonging is not the "all or nothing" binary belonging in the traditional set, but a continuous value between 0 and 1 to represent the degree of belonging, thereby providing the ability to handle fuzzy phenomena. The specific process of fuzzy logic is as follows Figure 5 As shown in Figure 1, it achieves precise control and decision-making in multivariable systems by dealing with uncertainty and fuzziness. The fuzzy logic system mainly includes the following key steps: fuzzification, fuzzy rule base, fuzzy reasoning, fuzzy output, and defuzzification.

[0067] Furthermore, the fuzzy rule base contains a number of "if-then" fuzzy rules that are used to describe the relationship between input and output variables. These rules are usually expressed in a language, for example:

[0068] in, is a fuzzy rule, is a set of fuzzy inputs, representing the ambient temperature and charging rate, is a set of fuzzy outputs, representing strain turning points and voltage turning points respectively. The fuzzy rules established based on the experimental data are shown in Table 6 and Table 7: Table 6 Fuzzy rules of strain turning point

[0069] Table 7 Fuzzy rules for voltage turning points

[0070] In a fuzzy control system, fuzzy rules are used to convert the relationship between input and output into conditional statements in natural language, thereby achieving control of complex or uncertain systems. The main functions of fuzzy rules are as follows: establishing the relationship between input and output, such as in this system, establishing the relationship between different charging rates and ambient temperatures of lithium iron phosphate batteries and the corresponding strain turning points and voltage turning points. It should be understood that this embodiment can also use the fuzzy control system proposed in this embodiment to establish the relationship between different charging rates and ambient temperatures and the corresponding strain turning points and voltage turning points for other types of lithium batteries such as ternary lithium batteries.

[0071] Specifically, LC, MC, HC, LT, MT, and HT represent low, medium, and high charging rates and low, medium, and high ambient temperatures, respectively; LS, MS, and HS represent low, medium, and high strain turning points, respectively; and LV, MV, and HV represent low, medium, and high voltage turning points, respectively.

[0072] Furthermore, the corresponding fuzzy outputs, namely the strain turning point and the voltage turning point, are obtained according to the established fuzzy rules and fuzzy inputs: This embodiment is based on the Mamdani system, which is based on fuzzy logic rules and input fuzzy sets, where the "Min" operation is used to determine the membership value of the rule antecedent, that is, the minimum membership value of all input conditions in the rule is selected as the activation strength of the rule. The "Max" operation is used to aggregate the outputs of multiple rules. If different rules output the same fuzzy set, the maximum value of the membership output of these rules is selected.

[0073] After obtaining the corresponding fuzzy output according to the MIN-MAX method, defuzzification is required. Here, this embodiment adopts the centroid method, which is as follows:

[0074] in, Represent the fuzzy membership function and output variable of the output variable respectively. The output is the strain turning point and voltage turning point. Figure 6 As shown, Figure 6 (a) is the output voltage turning point result. Figure 6 (b) in the figure is the output strain turning point result. It can be seen that it is consistent with the analysis of this implementation, that is, higher ambient temperature and charging rate will reduce the safety performance of the battery, that is, the threshold is reduced.

[0075] As a further preferred technical solution, in order to verify the effectiveness and generalization of the proposed thermal runaway warning threshold calculation method, this embodiment selected the same battery at a 1.5C charging rate, 50°C ambient temperature and the same type of 17.8AH capacity battery, and conducted an overcharge thermal runaway experiment at a 1C charging rate and 30°C ambient temperature. Then verify whether the method of the embodiment of the present invention can provide effective warning. The entire verification process is as follows: Figure 7 As shown, Figure 7 (a) is the strain turning point calculation process. Figure 7 (b) in the figure is the voltage turning point calculation process: (1) Conduct an overcharge thermal runaway experiment under the above conditions, including initializing the experimental equipment, setting the initial SOC of the battery to 100, and conducting an overcharge thermal runaway experiment on the battery at a preset ambient temperature and charging rate. At the same time, collect signals such as surface temperature, voltage, strain, and gas concentration during the battery thermal runaway process.

[0076] (2) Normalize the input. First, normalize the input ambient temperature and charging rate. The method proposed in the embodiment of the present invention studies the range of ambient temperature of 25°C-75°C and charging rate of 0.5C-2C, and can be normalized according to the following formula:

[0077] in, represents the input fuzzy variable, They represent the upper and lower limits of the input fuzzy variables, respectively. The specific ranges have been introduced above. For example, after normalization, the input of the 1.5C charging rate is 2 / 3.

[0078] (3) The normalized input is mapped through the fuzzy membership function. The input values ​​of ambient temperature and charging rate are mapped to different stages according to the membership function.

[0079] The normalized input data is mapped through a fuzzy membership function to convert it into a fuzzy set suitable for fuzzy reasoning. Specifically, the membership function maps the input values ​​of ambient temperature and charging rate to different stages or levels according to the preset membership range and membership level. For example, by dividing the ambient temperature into stages such as "low temperature", "medium temperature" and "high temperature", and dividing the charging rate into levels such as "slow", "medium speed" and "fast", the fuzziness of the input variables can be more finely characterized. This process helps the fuzzy logic system to more accurately understand and reason about the state of the battery at different temperatures and charging rates, thereby supporting the subsequent fuzzy control decision-making process.

[0080] For example, after the charging rate of 1.5C is normalized, the input is 2 / 3, which corresponds to a combination of half MC and half HC in the membership function, that is, a combination of half medium speed and half high speed.

[0081] (4) The fuzzy output is obtained through the established fuzzy rules. The specific formula is as follows:

[0082] j This is the definition of a fuzzy rule R(s) where , and It represents the fuzzy membership function, which corresponds to the membership of the input variables; It is a logical "and" operation, which usually corresponds to the operation of taking the minimum value in fuzzy logic; represents the final fuzzy output, Represents a union operation, which represents the union of multiple results; j =1 to j =9 means there are 9 different fuzzy rules; A represents the set of input fuzzy variables, Indicates the corresponding activation The fuzzy rules The vector representing the complement of the input fuzzy variable set, Representation and operation, T Represents the transpose of a matrix.

[0083] Finally, the corresponding output strain turning point and voltage turning point are obtained. The specific test results are as follows: Figure 8 As shown, Figure 8 (a) is the test result of the same battery at a charging rate of 1.5C and an ambient temperature of 50°C. Figure 8 (b) is the test result of a battery of the same type with a larger capacity at a charging rate of 1C and an ambient temperature of 30°C. The test results successfully predicted the voltage turning point and the strain turning point, effectively proving the effectiveness of the method proposed in this embodiment.

[0084] Example 2 like Fig. 9 As shown, the second embodiment of the present invention provides a thermal runaway warning threshold calculation device, the device comprising: The parameter iteration optimization module 10 is used to represent the parameters in the parameter set to be optimized of the fuzzy membership function using antibodies in the initial population, and to iteratively optimize the initial population using an immune clonal selection algorithm to obtain a global optimal solution for the parameter set; An optimal fuzzy membership function determination module 20, used to determine an optimal fuzzy membership function based on the global optimal solution of the parameter set; The warning threshold migration module 30 is used to map different ambient temperatures and charging rates to different fuzzy intervals through the optimal fuzzy membership function, output warning thresholds under different combinations of ambient temperatures and charging rates, and construct a corresponding relationship between different combinations of conditions and warning thresholds.

[0085] As a further preferred technical solution, the parameter iteration optimization module 10 is specifically used to perform the following steps: S11, randomly generate an initial population, each antibody represents a parameter of the fuzzy membership function; S12, using the fitness function to evaluate each antibody, and determine whether the evaluation result meets the termination condition, if not, execute step S13, if yes, execute step S16; Specifically, the fitness function is expressed as:

[0086]

[0087] In the formula, They represent the first The strain turning point, Voltage turning point; They represent the experimental measurements. The strain turning point, Voltage turning point; Indicates antibody Adaptability; Indicates the population size; , They represent the difference between the predicted strain turning point and the actual strain turning point, and the difference between the predicted voltage turning point and the actual voltage turning point, respectively.

[0088] S13, selecting high-quality antibodies based on the antibody evaluation results, and cloning the selected antibodies to obtain cloned antibodies; Furthermore, when the selected antibody is cloned, the number of clones is proportional to the fitness of the selected antibody.

[0089] S14, performing a mutation operation on the cloned antibody, evaluating the mutated antibody using the fitness function, and competing with the antibodies in the initial population, retaining the current optimal antibody to the next generation, and forming the next generation population; Furthermore, when performing mutation operations on the cloned antibodies, random perturbations are introduced as follows:

[0090] In the formula, Indicates the mutated The first variables; Indicates the number before mutation The first decision variables; Indicates multiplication; represents a standard normal distribution random number; indicates that the strength of mutation is usually inversely proportional to the antibody fitness, , Indicates The fitness value of an antibody, It indicates the maximum fitness value in the current population and is used for normalization to make the variation range relatively uniform.

[0091] S15, re-execute steps S12 to S14 for the next generation population; S16. Output the global optimal solution of the parameter set of the fuzzy membership function.

[0092] As a further preferred technical solution, the warning threshold migration module 30 specifically includes: An experimental unit, used to perform an overcharge thermal runaway experiment on a lithium battery under different combinations of ambient temperature and charging rate, and continuously record experimental parameters, wherein the experimental parameters include a voltage change signal, a battery surface temperature change signal, a strain change signal, and a gas change signal; The fuzzy logic unit is used to map the ambient temperature and charging rate in each combination condition to different fuzzy intervals through the optimal fuzzy membership function, and obtain the warning thresholds under different combinations of ambient temperature and charging rate based on the fuzzy rules established by experimental data, wherein the warning thresholds include voltage turning points and strain turning points.

[0093] It should be noted that other embodiments of the thermal runaway warning threshold calculation device or implementation methods of the present invention can refer to the above-mentioned method embodiments and will not be described in detail here.

[0094] Example 3 like Fig.10 As shown, the third embodiment of the present invention proposes a thermal runaway early warning method based on fuzzy logic, and the method includes the following steps: S1. Real-time collection of lithium battery status parameters under certain working conditions; S2. Compare the state parameter with the corresponding warning threshold under the current working condition to perform thermal runaway warning; The warning threshold is a warning threshold calculated based on the thermal runaway warning threshold calculation method described in the first embodiment and corresponding to the ambient temperature and charging rate under the current working conditions.

[0095] When performing thermal runaway warning for lithium batteries in this embodiment, the real-time collected lithium battery status parameters are compared with the warning threshold to achieve thermal runaway warning, and the warning threshold is the voltage turning point and the strain turning point calculated by the fuzzy membership function according to the ambient temperature and charging rate under the current working conditions. Therefore, the warning threshold used in the present invention is not a static threshold, but a dynamic threshold that changes dynamically with different ambient temperatures and charging rates, which solves the migration problem of the thermal runaway warning threshold under different ambient temperatures and charging rates; therefore, by selecting a warning threshold that is suitable for the ambient temperature and charging rate under the current working conditions to perform thermal runaway warning, the accuracy of the thermal runaway warning can be greatly improved. As a further preferred technical solution, Fig.11 As shown, the state parameter includes a voltage signal and a strain signal, the warning threshold includes a voltage turning point and a strain turning point, and the state parameter is compared with the corresponding warning threshold under the current working condition to perform thermal runaway warning, including: comparing the voltage signal and the voltage turning point with the strain signal and the strain turning point respectively; When the voltage signal is greater than the voltage turning point and the strain signal is greater than the strain turning point, a first-level thermal runaway warning is performed.

[0096] As a further preferred technical solution, the state parameter further includes a gas signal, and the method further includes: When the strain decrease rate is greater than the set first threshold and the gas concentration increase rate is greater than the set second threshold, a second-level thermal runaway warning is issued.

[0097] As a further preferred technical solution, the state parameter also includes a battery surface temperature signal, and the method further includes: When the voltage drop rate is greater than the set third threshold and the battery surface temperature rise rate is greater than the set fourth threshold, a third-level thermal runaway warning is issued.

[0098] It should be noted that the first threshold, the second threshold, the third threshold and the fourth threshold in this embodiment are respectively , , , .

[0099] It should be understood that the first threshold, the second threshold, the third threshold and the fourth threshold in the present embodiment are empirical values ​​obtained by technicians in this field through a large number of experiments for comparison with the strain decrease rate, the gas concentration increase rate, the voltage decrease rate and the battery surface temperature increase rate, respectively.

[0100] It should be noted that in order to explore the characteristics of battery overcharge thermal runaway under different ambient temperatures and charging rates, the following key warning signal characteristics need to be extracted from the collected experimental parameters during the overcharge thermal runaway experiment. These characteristics will provide a reliable basis for the establishment of a battery thermal runaway warning system: Maximum voltage drop rate: The maximum voltage drop rate indicates the speed at which the battery voltage begins to drop sharply. It is a key indicator for evaluating the acceleration of the electrochemical reaction inside the battery. By analyzing the voltage change curve, the peak value of the voltage drop rate and the time point when it occurs can be determined.

[0101] Maximum temperature rise rate: The maximum temperature rise rate refers to the maximum rate of increase of the battery surface or internal temperature per unit time. Generally, as the intensity of the internal reaction of the battery increases, the temperature rise rate will increase significantly. When the temperature rise rate reaches a certain level, it indicates that thermal runaway inside the battery has begun or is about to occur. This feature can help predict the occurrence time of thermal runaway and provide a basis for taking protective measures in advance.

[0102] Strain turning point: The strain turning point refers to the critical moment when the battery shell or internal structure undergoes significant deformation during overcharging. During the process of overcharging the battery, which leads to internal gas generation and pressure increase, the strain of the shell will turn, indicating a change in the battery structure and potential risk of rupture. The strain sensor can monitor this turning point in real time and, by analyzing its changing trend, give early warning that the battery is about to enter a runaway state.

[0103] Voltage turning point: The voltage turning point refers to the key node where the battery voltage curve changes significantly in the early stage of the overcharging process. Usually, a turning point occurs after the voltage reaches a certain local peak, indicating that the battery's electrochemical reaction has changed, which may be accompanied by internal overheating, gas generation, etc. The extraction of the voltage turning point helps to identify the reaction stage inside the battery and provides important information for judging the risk of thermal runaway.

[0104] Exhaust SOC: Exhaust SOC refers to the state of charge when the battery begins to emit obvious gas. As the overcharge reaction proceeds, the electrolyte inside the battery decomposes to produce gas, and the battery's SOC is usually close to or exceeds 100% at this time. Exhaust SOC can reflect the charge accumulation of the battery before thermal runaway. By monitoring the relationship between the gas emission signal and the battery SOC, the critical point of the battery entering runaway can be more accurately determined.

[0105] Furthermore, by extracting these features (maximum voltage drop rate, maximum temperature rise rate, strain turning point, voltage turning point, exhaust SOC), the overcharge thermal runaway characteristics of the battery under different temperature and charging rate conditions can be fully captured, and the battery thermal runaway can be divided into three different stages: The judgment basis for the first stage is that the battery voltage and strain reach the strain and voltage turning points respectively, which indicates that the electrolyte inside the battery has reacted and started to produce gas, corresponding to the strain turning point and voltage turning point of the extracted features.

[0106] The judgment basis for the second stage is that the battery strain drops sharply and the gas concentration rises sharply. This is because the gas accumulated inside the battery continues to increase, and the excessive pressure eventually causes the battery to rupture and exhaust, which corresponds to the exhaust SOC of the extracted features.

[0107] The judgment basis for the third stage is that the battery voltage drops sharply and the battery surface temperature rises sharply, which indicates that a large-scale internal short circuit has occurred in the battery and thermal runaway is about to be triggered, corresponding to the maximum voltage drop rate and maximum temperature rise rate of the extracted features.

[0108] Under different ambient temperatures and charging rates, the voltage and strain turning points of the battery are different, which are determined by adopting the above-mentioned thermal runaway warning threshold calculation method.

[0109] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0110] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0111] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for calculating a thermal runaway warning threshold, characterized in that: include: The parameters in the parameter set to be optimized of the fuzzy membership function are represented by antibodies in the initial population, and the initial population is iteratively optimized using the immune clonal selection algorithm to obtain the global optimal solution of the parameter set; Determining an optimal fuzzy membership function based on the global optimal solution of the parameter set; Different ambient temperatures and charging rates are mapped to different fuzzy intervals through the optimal fuzzy membership function, warning thresholds under different combinations of ambient temperatures and charging rates are output, and corresponding relationships between different combinations of conditions and warning thresholds are constructed.

2. The thermal runaway warning threshold calculation method according to claim 1, characterized in that: The parameters in the parameter set to be optimized of the fuzzy membership function are represented by antibodies in the initial population, and the initial population is iteratively optimized using an immune clonal selection algorithm to obtain a global optimal solution of the parameter set, including: S11, randomly generate an initial population, each antibody represents a parameter of the fuzzy membership function; S12, using the fitness function to evaluate each antibody, and determine whether the evaluation result meets the termination condition, if not, execute step S13, if yes, execute step S16; S13, selecting high-quality antibodies based on the antibody evaluation results, and cloning the selected antibodies to obtain cloned antibodies; S14, performing a mutation operation on the cloned antibody, evaluating the mutated antibody using the fitness function, and competing with the antibodies in the initial population, retaining the current optimal antibody to the next generation, and forming the next generation population; S15, re-execute steps S12 to S14 for the next generation population; S16. Output the global optimal solution of the parameter set of the fuzzy membership function.

3. The thermal runaway warning threshold calculation method according to claim 2, characterized in that: The fitness function is expressed as: In the formula, They represent the first The strain turning point, Voltage turning point; They represent the experimental measurements. The strain turning point, Voltage turning point; Indicates antibody Adaptability; Indicates the population size; , They represent the difference between the predicted strain turning point and the actual strain turning point, and the difference between the predicted voltage turning point and the actual voltage turning point, respectively.

4. The thermal runaway warning threshold calculation method according to claim 2, characterized in that: The method further comprises: When the selected antibody is cloned, the number of clones is proportional to the fitness of the selected antibody.

5. The thermal runaway warning threshold calculation method according to claim 2, characterized in that: The method further comprises: When performing mutation operations on cloned antibodies, random perturbations are introduced as follows: In the formula, Indicates the mutated The first variables; Indicates the number before mutation The first decision variables; Indicates multiplication; represents a standard normal distribution random number; indicates that the strength of mutation is usually inversely proportional to the antibody fitness, , Indicates The fitness value of an antibody, Indicates the maximum fitness value in the current population and is used for normalization to make the variation range relatively uniform.

6. The thermal runaway warning threshold calculation method according to claim 2, characterized in that: The formula of the termination condition is expressed as: In the formula, represents the next generation population; It means that the antibody with the best fitness is selected from the combination of the original population and the clone population to form a new generation population; represents the initial population; represents a union; Represents the population after clonal mutation.

7. The thermal runaway warning threshold calculation method according to claim 1, characterized in that: The mapping of different ambient temperatures and charging rates to different fuzzy intervals through the optimal fuzzy membership function, outputting warning thresholds under different combinations of ambient temperatures and charging rates, and constructing corresponding relationships between different combinations of conditions and warning thresholds include: Conducting an overcharge thermal runaway experiment on a lithium battery under different combinations of ambient temperature and charging rate conditions, and continuously recording experimental parameters, wherein the experimental parameters include a voltage change signal, a battery surface temperature change signal, a strain change signal, and a gas change signal; The ambient temperature and charging rate in each combination condition are mapped to different fuzzy intervals through the optimal fuzzy membership function, and the warning thresholds under different combinations of ambient temperature and charging rate are obtained based on the fuzzy rules established according to the experimental data.

8. The thermal runaway warning threshold calculation method according to claim 7, characterized in that: The method further comprises: The ambient temperature and charging rate in each combination condition are normalized and then mapped to different fuzzy intervals through the optimal fuzzy membership function.

9. The thermal runaway warning threshold calculation method according to claim 7, characterized in that: The warning threshold includes a voltage turning point and a strain turning point.

10. The thermal runaway warning threshold calculation method according to any one of claims 1 to 9, characterized in that: The types of the fuzzy membership function include trigonometric function and trapezoidal function.

11. A thermal runaway warning threshold calculation device, characterized in that: The device comprises: The parameter iteration optimization module is used to represent the parameters in the parameter set to be optimized of the fuzzy membership function using antibodies in the initial population, and to iteratively optimize the initial population using the immune clonal selection algorithm to obtain the global optimal solution of the parameter set; An optimal fuzzy membership function determination module, used to determine an optimal fuzzy membership function based on a global optimal solution of the parameter set; The warning threshold migration module is used to map different ambient temperatures and charging rates to different fuzzy intervals through the optimal fuzzy membership function, output the warning thresholds under different combinations of ambient temperatures and charging rates, and construct the corresponding relationship between different combinations of conditions and warning thresholds.

12. A thermal runaway early warning method based on fuzzy logic, characterized in that: The method comprises: Real-time collection of lithium battery status parameters under certain working conditions; Comparing the state parameter with the corresponding warning threshold under the current working condition to issue a thermal runaway warning; Among them, the warning threshold is a warning threshold calculated based on the thermal runaway warning threshold calculation method described in any one of claims 1 to 10 and corresponding to the ambient temperature and charging rate under the current operating conditions.

13. The thermal runaway early warning method based on fuzzy logic according to claim 12, characterized in that: The state parameter includes a voltage signal and a strain signal, the warning threshold includes a voltage turning point and a strain turning point, and the state parameter is compared with the corresponding warning threshold under the current working condition to perform thermal runaway warning, including: comparing the voltage signal and the voltage turning point with the strain signal and the strain turning point respectively; When the voltage signal is greater than the voltage turning point and the strain signal is greater than the strain turning point, a first-level thermal runaway warning is performed.

14. The thermal runaway early warning method based on fuzzy logic according to claim 13, characterized in that: The state parameter also includes a gas signal, and the method further includes: When the strain decrease rate is greater than the set first threshold and the gas concentration increase rate is greater than the set second threshold, a second-level thermal runaway warning is issued.

15. The thermal runaway early warning method based on fuzzy logic according to claim 13, characterized in that: The state parameter also includes a battery surface temperature signal, and the method further includes: When the voltage drop rate is greater than the set third threshold and the battery surface temperature rise rate is greater than the set fourth threshold, a third-level thermal runaway warning is issued.

Images