Self-adaptive evaluation method, device and equipment for operation safety state of electrochemical energy storage battery and medium

Abstract

The invention provides a self-adaptive evaluation method, device and equipment for an operation safety state of an electrochemical energy storage battery and a medium, and the method comprises the steps: obtaining historical operation data of a target battery in different working stages; determining a battery health index of the target battery based on the historical operation data in each working stage; determining a safety assessment method of the target battery based on the battery health index; and evaluating the target battery based on the safety evaluation method to obtain a risk evaluation level of the target battery. By adopting the method, the accuracy and the reliability of an evaluation result are improved.

Description

Technical Field

[0001] The present invention relates to the field of battery safety management, and more particularly, to an adaptive evaluation method, device, equipment and medium for the operating safety state of electrochemical energy storage batteries. Background Art

[0002] In today's society, electrochemical energy storage batteries play a crucial role in many fields such as renewable energy storage, electric vehicles, and smart grids. With the rapid development of these fields, the performance and safety requirements for electrochemical energy storage batteries are also getting higher and higher. Accurately evaluating the operating safety state of batteries and timely detecting potential health risks are crucial for ensuring the reliable operation of batteries, extending the battery life, and ensuring the safety and stability of related systems. Currently, when evaluating the health risks of electrochemical energy storage batteries, a specific unified evaluation method is usually adopted, that is, regardless of the working state or stage of the battery, the same evaluation method is used to evaluate its risks.

[0003] However, it has been found in the research that there are differences in the internal physical and chemical processes of different batteries in different working stages and different health states, and the requirements for safety evaluation and the adaptability to the evaluation method are also different. If the same evaluation method is used for batteries in different state stages, it may occur that the evaluation method is not suitable for the current battery state, thus reducing the accuracy and reliability of the evaluation results. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide an adaptive evaluation method, device, equipment and medium for the operating safety state of electrochemical energy storage batteries, so as to improve the accuracy and reliability of the evaluation results.

[0005] In a first aspect, an embodiment of the present application provides an adaptive evaluation method for the operating safety state of an electrochemical energy storage battery, the method comprising: Obtaining historical operation data of a target battery in different working stages; Determining a battery health index of the target battery based on the historical operation data in each working stage; Determining a safety evaluation method for the target battery based on the battery health index; Evaluating the target battery based on the safety evaluation method to obtain a risk evaluation level of the target battery.

[0006] Optionally, the determining a battery health index of the target battery based on the historical operation data in each working stage includes: Determining a health degree feature of the target battery in each working stage based on the historical operation data in each working stage; Standardize the health characteristics in each working stage respectively, and assign weights to the standardized health characteristics. Determine the battery health index based on the standardized health characteristics and their weights.

[0007] Optionally, the working stages include charging state, discharging state and standing state; the health characteristics in the charging state include the slope of the capacity increment curve and the retention degree of constant current charging duration; the health state in the discharging state includes dynamic internal resistance and the retention degree of discharging platform; the health characteristics in the standing state include voltage relaxation rate. The determination of the health characteristics of the target battery in each working stage based on the historical operation data in each working stage includes: Determine the slope of the capacity increment curve according to the ratio of the voltage change amount to the capacity change amount. Determine the retention degree of the constant current charging duration according to the ratio of the constant current charging duration of the aged battery to that of the initial battery. Determine the dynamic internal resistance according to the ratio of the voltage drop during the discharge pulse to the discharge current. Determine the retention degree of the discharging platform according to the ratio of the voltage integral within the preset platform voltage range to the duration. Determine the voltage relaxation rate according to the exponential fitting parameter of the voltage relaxation process.

[0008] Optionally, the method for determining the safety assessment of the target battery based on the battery health index includes: Classify the health level of the target battery according to the battery health index, where the health levels are low health, medium health and high health. When the target battery is of low health, determine the safety assessment method as the abnormal fluctuation analysis method based on information entropy. When the target battery is of medium health, determine the safety assessment method as the multi-parameter fuzzy logic fusion analysis method. When the target battery is of high health, determine the safety assessment method as the time series prediction analysis method using long short-term memory network.

[0009] Optionally, the assessment of the target battery based on the safety assessment method to obtain the risk assessment level of the target battery includes: When the safety assessment method is the abnormal fluctuation analysis method based on information entropy, discretize the voltage signal into multiple intervals and calculate the probability distribution of each interval. Calculate the normalized information entropy value of the voltage signal according to the probability distribution. Determine the risk assessment level according to the normalized information entropy value.

[0010] Optionally, the step of evaluating the target battery based on the safety evaluation method to obtain the risk evaluation level of the target battery includes: When the safety evaluation method is a multi-parameter fuzzy logic fusion analysis method, extract the voltage fluctuation coefficient, current peak-to-average ratio, SOC change rate, and temperature gradient as input parameters; Perform fuzzy processing on each parameter through a predefined membership function, and perform fuzzy inference based on an expert rule base; Use the area centroid method to defuzzify the inference result to obtain a risk prediction value; Determine the risk evaluation level according to the risk prediction value.

[0011] Optionally, the step of evaluating the target battery based on the safety evaluation method to obtain the risk evaluation level of the target battery includes: When the safety evaluation method is determined to be a long short-term memory network for time series prediction analysis method, construct a time series feature input including voltage fluctuation, power peak-to-average ratio, and temperature rise rate based on a sliding window; Extract the time series dependence relationship through the gating mechanism of the LSTM network, and predict the risk prediction value within the future time window; Determine the risk evaluation level according to the risk prediction value.

[0012] In a second aspect, an embodiment of the present application provides an adaptive evaluation device for the operating safety state of an electrochemical energy storage battery, where the device includes: An operating data acquisition module, configured to acquire historical operating data of a target battery in different working stages; A battery health index determination module, configured to determine the battery health index of the target battery based on the historical operating data in each working stage; An evaluation method determination module, configured to determine a safety evaluation method for the target battery based on the battery health index; An evaluation result determination module, configured to evaluate the target battery based on the safety evaluation method to obtain the risk evaluation level of the target battery.

[0013] In a third aspect, an embodiment of the present application provides a computer device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are executed by the processor, the steps of the adaptive evaluation method for the operating safety state of the electrochemical energy storage battery in any optional implementation manner in the first aspect are executed.

[0014] Fourthly, an embodiment of the present application provides a computer - readable storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the steps of the adaptive evaluation method for the operating safety state of the electrochemical energy storage battery described in any optional implementation manner of the first aspect above.

[0015] The technical solutions provided by the present application include but are not limited to the following beneficial effects: Firstly, obtaining the historical operation data of the target battery at different working stages provides a solid foundation for subsequent analysis, ensuring that the evaluation can fully consider the performance of the battery under different working conditions and avoiding one - sided evaluation due to data lack. Then, determining the battery health index of the target battery based on the historical operation data at each working stage comprehensively considers the performance changes of the battery at different stages, can accurately quantify the battery health status, provides an accurate basis for subsequent evaluation, and helps users timely grasp the degree of battery aging. Next, determining the safety evaluation method of the target battery based on the battery health index adaptively selects to ensure that the evaluation method is adapted to the actual situation of the battery, gives full play to the advantages of each method, improves the evaluation accuracy, and avoids errors caused by a "one - size - fits - all" evaluation method. Finally, evaluating the target battery based on the safety evaluation method to obtain the risk evaluation level of the target battery can effectively capture potential safety hazards during battery operation, presents them in an intuitive risk level, facilitates users to quickly judge the battery safety state, take measures in advance, ensure the safe use of the battery, and reduce the occurrence of potential safety accidents.

[0016] The present application constitutes a complete evaluation system through the above steps. From data collection to risk level output, each step closely cooperates to ensure data comprehensiveness, evaluation accuracy, method adaptability, and result intuitiveness, can effectively improve the accuracy and reliability of the evaluation results, and further helps to improve battery use safety, extend service life, and reduce operation and maintenance costs.

[0017] To make the above - mentioned objects, features, and advantages of the present invention more obvious and understandable, the following specific preferred embodiments are given, and detailed descriptions are made in conjunction with the accompanying drawings as follows. Description of the Drawings

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention, and thus should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 Shows a flowchart of an adaptive evaluation method for the operating safety state of an electrochemical energy storage battery provided by Embodiment 1 of the present invention; Figure 2 The flowchart of a method for determining the battery health index provided in the first embodiment of the present invention is shown; Figure 3 The flowchart of a method for determining the risk assessment level provided in the first embodiment of the present invention is shown; Figure 4 The flowchart of the second method for determining the risk assessment level provided in the first embodiment of the present invention is shown; Figure 5 The flowchart of the third method for determining the risk assessment level provided in the first embodiment of the present invention is shown; Figure 6 The schematic structural diagram of an adaptive evaluation device for the operating safety state of an electrochemical energy storage battery provided in the second embodiment of the present invention is shown; Figure 7 The schematic structural diagram of a computer device provided in the third embodiment of the present invention is shown. Detailed implementation manners

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are only some of the embodiments of the present invention, rather than all of them. Usually, the components of the embodiments of the present invention described and illustrated herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] Embodiment 1 For the convenience of understanding the present application, the following Figure 1 describes in detail Embodiment 1 of the present application in combination with the content described in the flowchart of a method for optimizing the output power of a vehicle battery provided in the first embodiment of the present invention shown.

[0022] See Figure 1 as shown, Figure 1 The flowchart of an adaptive evaluation method for the operating safety state of an electrochemical energy storage battery provided in the first embodiment of the present invention is shown, wherein the method includes steps S101 to S106: S101: Obtain the historical operation data of the target battery at different working stages.

[0023] Specifically, the working phases of the battery include the charging phase, the discharging phase, and the standby phase. The historical operation data covers various operation parameters of the battery in these phases, such as the data of voltage, current, temperature, capacity, etc. changing with time, and these data can reflect the performance and change trend of the battery in the actual use process.

[0024] S102: Determine the battery health index of the target battery based on the historical operation data in each working phase.

[0025] Specifically, a method combining physical model-driven and data-driven is used to extract features from the data in different working phases respectively. For example, in the charging phase, features such as the slope of the capacity increment IC curve and the retention degree of constant current charging duration are extracted; in the discharging phase, features such as the voltage drop - current ratio (dynamic internal resistance characterization) and the retention degree of the discharging platform are extracted; in the standby phase, features such as the voltage relaxation rate are extracted. Then these features are standardized, and the entropy value method is used to assign weights to different features, and the battery health index (HI) is obtained by integrating the features in different phases. The HI value ranges from 0 to 1, and the larger the value, the healthier the battery.

[0026] S103: Determine the safety assessment method of the target battery based on the battery health index.

[0027] Specifically, according to the size of the battery health index (HI), the health state of the battery is divided into three intervals: low, medium, and high. When HI < 0.3, the battery is in the low health degree interval, and irreversible physical and chemical changes usually occur inside. At this time, the information entropy analysis method is used; when 0.3 ≤ HI < 0.7, the battery is in the medium health degree interval, and its aging process shows the characteristics of multi-physical field coupling. The multi-parameter fusion assessment method based on fuzzy logic is used; when HI ≥ 0.7, the battery is in the high health degree interval, the parameter changes are stable, and it is suitable for time series modeling. The long short-term memory network is used for time series prediction method.

[0028] S104: Evaluate the target battery based on the safety assessment method to obtain the risk assessment level of the target battery.

[0029] Specifically, according to the selected safety assessment method, a detailed assessment and analysis of the battery is carried out. For example, for batteries with low health degree, the complexity characteristics of the operation data are quantified by the information entropy analysis method to evaluate the safety state; for batteries with medium health degree, a non-linear safety evaluation model is constructed based on fuzzy logic by fusing multiple parameters such as voltage, current, SOC, and temperature; for batteries with high health degree, the LSTM network architecture is used for time series feature extraction and dynamic weight adjustment to realize the prediction of the safety state. Finally, according to the evaluation results, the risk assessment level of the battery is determined, generally divided into three levels: high risk, medium risk, and low risk.

[0030] In an alternative embodiment, referring to Figure 2 as shown Figure 2 FIG. 5 shows a flowchart of a method for determining a battery health index provided in the first embodiment of the present invention. Among them, determining the battery health index of the target battery based on the historical operation data in each working stage includes steps S201 to S203: S201: Determine the health characteristics of the target battery in each working stage based on the historical operation data in each working stage.

[0031] Specifically, the performance and aging characteristics of the battery are different in different working stages. By analyzing the historical operation data of each working stage, the key characteristics reflecting the battery health status can be extracted.

[0032] S202: Standardize the health characteristics in each working stage respectively, and assign weights to the standardized health characteristics.

[0033] Specifically, different health characteristics may have different dimensions and value ranges. For example, the slope of the IC curve may be a decimal, while the retention of the constant current charging duration may be a percentage. The purpose of the standardization process is to convert these characteristics to a unified scale, usually the [0,1] interval. This can eliminate the influence of dimensions and make different characteristics comparable in subsequent calculations. Common normalization methods include max-min normalization, Z-score normalization, etc.

[0034] Specifically, the standardized health characteristics are determined by the following expression :

[0035] where is the original value of the j-th feature of the i-th sample, represents the j-th feature value of all samples, represents the minimum value among the j-th feature values of all samples, represents the maximum value among the j-th feature values of all samples. By standardization, the unit differences of different features are eliminated, making all features comparable.

[0036] The contribution degree of each health feature to the battery health index is different. Therefore, it is necessary to assign weights to the normalized health features. Here, the entropy method is used to determine the weights. The entropy method is a method based on the variability of the data itself to determine the weights. Information entropy reflects the degree of chaos or uncertainty of the data. If the data of a feature changes greatly, it means that the feature contains more information and has a stronger ability to distinguish the battery health status, so its weight will be relatively large; on the contrary, if the data of a feature changes little, it means that the feature provides less information and the weight will be relatively small. The specific steps include calculating the information entropy of each health feature, then calculating the information utility value according to the information entropy, and finally determining the weights of each health feature according to the information utility value.

[0037] Specifically, the feature information entropy of each health feature is determined through the following expression:

[0038] Among them, is the information entropy of the j-th health feature, which is used to measure the degree of data chaos of this health feature; is the total number of samples; is the proportion of the i-th sample in the j-th health feature, and all add up to 1.

[0039] The weights of each health feature are determined through the following expression:

[0040] Among them, is the weight of the j-th health feature, m is the total number of health features, represents the information entropy of the k-th health feature, represents the information entropy of the j-th health feature, represents the information utility value of the health feature. The smaller the entropy value (the more orderly the data), the greater the utility value.

[0041] S203: Determine the battery health index based on each health feature after standardization processing and its weight.

[0042] Specifically, after obtaining the normalized values and weights of each health feature, first determine the health sub-index of a single working stage (charging / discharging / standing) according to each health feature after standardization processing and its weight, and then determine the battery health index according to the health sub-index of a single working stage (charging / discharging / standing).

[0043] Specifically, first, the health sub-index of each working stage is determined through the following expression :

[0044] Among them, is the weight of the j-th feature in this working stage (calculated by the entropy weight method), is the normalized value of the j-th feature in this working stage (the original feature value is processed by the maximum and minimum values), and k represents the number of health features in this working stage.

[0045] Then, the final battery health index is determined through the following expression :

[0046] Among them, The value range is between 0 and 1, and the larger the value, the healthier the battery; and are the historical maximum and minimum values of the stage index; is the stage set, ∈{charge, discharge, rest}; HI charge 、HI discharge 、HI rest represent the health sub-indicators of the charging stage, discharging stage, and resting stage respectively, the index of the stage, ={ c 、 d 、 r}, representing the weight coefficients of the health sub-indicators of the charging stage, discharging stage, and resting stage respectively, and the weight coefficients are dynamically adjusted according to the charging duration ratio:

[0047] Among them, is the charging stage duration; is the discharging stage duration; is the total duration.

[0048] In summary, the health index is obtained through scientific calculations with multi-dimensions and dynamic weights, which not only considers the characteristics of different working states of the battery, but also automatically adjusts important features through a data-driven method, and is more in line with the actual battery attenuation law than the traditional single-index method.

[0049] In an optional embodiment, the working stages include charging state, discharging state and static state; the health characteristics in the charging state include the slope of the capacity increment curve and the constant current charging time retention; the health state in the discharging state includes the dynamic internal resistance and the discharge platform retention; the health characteristics in the static state include the voltage relaxation rate.

[0050] Specifically, the slope of the incremental capacity curve (IC curve slope): During the charging process, as the battery charge increases, the voltage will also rise accordingly. The slope of the IC curve reflects the ratio of the voltage change to the capacity change. When a healthy battery is charging, the slope of the IC curve will show a specific pattern. As the battery ages, factors such as the decrease in the activity of the electrode material and the increase in internal resistance will cause the slope of the IC curve to change. For example, structural damage to the electrode material may make it difficult for lithium ions to embed and extract, making the voltage change more sensitive to the capacity change, resulting in an increase in the slope of the IC curve.

[0051] Constant current charging duration retention: New batteries usually have a relatively stable duration during the constant current charging stage. As the battery ages, the internal resistance increases, the charging efficiency decreases, and the constant current charging duration gradually decreases. The constant current charging duration retention can intuitively reflect the degree of attenuation of the battery charging performance by comparing the constant current charging duration of the aged battery with the initial battery under the same charging conditions.

[0052] Dynamic internal resistance: During the discharge pulse, the battery voltage will drop due to the presence of internal resistance. The dynamic internal resistance is calculated by calculating the ratio of the voltage drop during the discharge pulse to the discharge current. Battery aging will cause structural changes in the internal electrode materials, decomposition of the electrolyte, etc., which will increase the internal resistance. The increase in dynamic internal resistance will affect the output power and discharge efficiency of the battery, so it is one of the important indicators to measure the health of the battery.

[0053] Discharge platform retention: During the discharge process, the battery usually has a relatively stable voltage platform. The discharge platform retention is evaluated by calculating the ratio of the voltage integral to the duration within the preset platform voltage range. A healthy battery can maintain this voltage platform for a long time during discharge. As the battery ages, factors such as the reduction of active substances in the electrode material and the increase of internal resistance will cause the discharge platform to shorten and the discharge platform retention to decrease.

[0054] Voltage relaxation rate: When the battery stops charging and discharging and enters a static state, its voltage will gradually approach the equilibrium voltage. This process is called voltage relaxation. The voltage relaxation rate is obtained by exponential fitting the voltage relaxation process. The chemical reaction and ion diffusion process inside the battery will affect the voltage relaxation rate. As the battery ages, the structural changes of the electrode material and the decrease in ion diffusion capacity will cause the voltage relaxation rate to change.

[0055] Determining the health characteristics of the target battery in each working stage based on the historical operation data in each working stage includes: Determining the slope of the capacity increment curve according to the ratio of the voltage change amount to the capacity change amount.

[0056] Specifically, the slope of the capacity increment curve is determined through the following expression (mV / mAh):

[0057] where ΔV is the voltage change amount (mV) within the voltage sampling interval; is the battery capacity; n is the data length, that is, the number of sampling points; and are the voltage and capacity at the i-th sampling point moment; ΔQ is the corresponding capacity change amount (mAh); m is the sliding window width (it is recommended to take 5 - 10 sampling points); , is the feature extraction interval (usually take the capacity segment corresponding to 3.6 - 3.8V).

[0058] Determining the constant current charging duration retention according to the ratio of the constant current charging duration of the aged battery to that of the initial battery.

[0059] Specifically, the constant current charging duration retention is determined through the following expression (%): ; where is the duration of the constant current charging stage of the aged battery (s); is the constant current charging duration of the new battery under the same working conditions (s); the charging termination condition is that the single - cell voltage V cutoff reaches 4.2V ± 10mV.

[0060] Determining the dynamic internal resistance according to the ratio of the voltage drop during the discharge pulse to the discharge current.

[0061] Specifically, the dynamic internal resistance, that is, the voltage drop - current ratio, is determined through the following expression (mΩ):

[0062] where is the voltage drop during the discharge pulse (mV); is the pulse discharge current (A); is the summary of the pulse period, k represents the current pulse period number; is the open - circuit voltage before the pulse (V); It is the voltage at the t-th second after the start of discharge (take t = 5k).

[0063] The discharge platform retention is determined according to the ratio of the voltage integral to the duration within the preset platform voltage range.

[0064] Specifically, the discharge platform retention is determined by the following expression (%):

[0065] where is the nominal platform voltage (such as 3.7V); , is the platform duration range (the section where the automatic detection voltage change rate < 0.1mV / s); represents the voltage monitoring value when the voltage is within the discharge platform.

[0066] The voltage relaxation rate is determined according to the exponential fitting parameter of the voltage relaxation process.

[0067] Specifically, the voltage relaxation rate is determined by the following expression (s):

[0068] where is the instantaneous voltage at the start of relaxation (V); is the voltage at the end of relaxation (take the value at t = 300s); is the theoretical equilibrium voltage, which can be obtained by exponential fitting.

[0069] In an alternative embodiment, the safety assessment method for the target battery based on the battery health indicator includes: Classify the health level of the target battery according to the battery health indicator, where the health level is low health, medium health, and high health.

[0070] Specifically, after calculating the battery health indicator ( ), considering reasons such as the continuous growth of the SEI film and the start of lithium dendrite formation inside the battery, its internal impedance has the characteristics of steady growth - accelerated growth - rapid growth. Therefore, according to the value of the battery health indicator, the battery health indicator HI is classified, and the specific classification criteria are as follows: ; When the target battery has low health, the safety assessment method is determined as the abnormal fluctuation analysis method based on information entropy.

[0071] Specifically, when the state of health (SOH) of the battery drops to the low health range (HI < 0.3), irreversible physical and chemical changes usually occur inside, including but not limited to lithium metal precipitation, solid electrolyte interface (SEI) film rupture, active material loss, etc. At this time, the information entropy analysis method can effectively capture the abnormal fluctuation characteristics of voltage / current signals and achieve accurate assessment of the safety state. This method is based on the principle of Shannon information theory and establishes a safety evaluation system by quantifying the complexity characteristics of operation data.

[0072] When the target battery is in medium health, the safety assessment method is determined as the multi-parameter fuzzy logic fusion analysis method.

[0073] Specifically, when the state of health (SOH) of the battery is in the medium range (0.3 ≤ HI < 0.7), its aging process exhibits multi-physical field coupling characteristics. Single-parameter analysis methods are difficult to accurately characterize the synergistic effects of multiple failure mechanisms such as SEI film growth and lithium ion diffusion limitation. Therefore, the multi-parameter fusion evaluation method based on fuzzy logic constructs a non-linear safety evaluation model by dynamically weighting and fusing voltage, current, SOC, and temperature parameters.

[0074] When the target battery is in high health, the safety assessment method is determined as the long short-term memory network for time series prediction analysis method.

[0075] Specifically, the battery parameters change smoothly at this stage, which is suitable for time series modeling. This method uses a two-layer LSTM network architecture to achieve safety state prediction through time series feature extraction and dynamic weight adjustment.

[0076] In an alternative embodiment, see Figure 3 as shown Figure 3 shows a flowchart of a method for determining a risk assessment level provided in the first embodiment of the present invention. Among them, obtaining the risk assessment level of the target battery based on the safety assessment method includes steps S301~S302: S301: When the safety assessment method is the abnormal fluctuation analysis method based on information entropy, discretize the voltage signal into multiple intervals and calculate the probability distribution of each interval.

[0077] Specifically, perform probability distribution modeling, discretize the voltage into k intervals, and calculate the probability distribution of each interval :

[0078] where is the number of samples in the j-th interval, is the total number of samples, is the number of discretization intervals (recommended 10 - 15).

[0079] S302: Calculate the normalized information entropy value of the voltage signal according to the probability distribution.

[0080] Specifically, first calculate the information entropy of the voltage signal according to the probability distribution : ; where is the probability that the voltage appears in the j-th interval.

[0081] Then, normalize the information entropy through the following expression to obtain the normalized information entropy value :

[0082] where = , which is the theoretical maximum entropy value.

[0083] S303: Determine the risk assessment level according to the normalized information entropy value.

[0084] Specifically, determine the risk assessment level according to the following expression : ; where represents the high-risk assessment level; represents the medium-risk assessment level; represents the low-risk assessment level.

[0085] In an alternative embodiment, as shown in Figure 4 shown, Figure 4 shows the flowchart of the second risk assessment level determination method provided in the first embodiment of the present invention, where the risk assessment level of the target battery is obtained by evaluating the target battery based on the security assessment method, including steps S401 to S403: S401: When the security assessment method is the multi-parameter fuzzy logic fusion analysis method, extract the voltage fluctuation coefficient, current peak-to-average ratio, SOC change rate, and temperature gradient as input parameters.

[0086] Specifically, determine several key features according to the historical operation data, including the voltage fluctuation coefficient , the current peak-to-average ratio , the SOC change rate and the temperature gradient :

[0087] where is the standard deviation of the voltage; is the mean value of the voltage; is the maximum current, is the root mean square value of the current, and the calculation method is , where n is the number of samples of the current value, is the th current value; is the difference between the SOC value at the current sampling point and the SOC value at the previous sampling point; is the maximum temperature, the minimum temperature, is the data time span, is the remaining battery capacity at the i-th.

[0088] S402: Fuzzify each parameter through a predefined membership function and perform fuzzy inference based on an expert rule base.

[0089] Specifically, first, define the membership functions of each parameter, including: Membership degree of voltage fluctuation coefficient and (universe of discourse: 0 - 0.2): ; ; When the membership degree of voltage fluctuation coefficient > 0.7, the state is high, that is ; when the membership degree of voltage fluctuation coefficient ≤0.7, the state is low, that is .

[0090] Peak-to-average ratio of current (universe of discourse: 1 - 5): ; When the membership degree of peak-to-average ratio < 0.3, the state is abnormal, that is ; when the membership degree of peak-to-average ratio ≥0.3, the state is normal, that is .

[0091] Membership degree of SOC change rate (universe of discourse: -0.1~0.1% / s): ; When the SOC change rate membership degree When <0.4, the state is , which is ; SOC change rate membership degree ≥0.4, the state is , which is .

[0092] Temperature gradient (universe of discourse: 0 - 5℃ / min): ; When the membership degree of temperature gradient <0.2, the state is , which is ; When the membership degree of temperature gradient ≥0.2, the state is , which is .

[0093] Then, construct a rule base containing 9 expert rules:

[0094] Among them, XOR represents the exclusive - or logical operation; the rule output value determines the weight through the Analytic Hierarchy Process (AHP).

[0095] S403: Use the area - centroid method to defuzzify the inference result to obtain the risk prediction value; Specifically, calculate the risk prediction value using the area - centroid method :

[0096] Among them, is the number of rules, is the activation strength of the i - th rule; is the risk benchmark value corresponding to the rule output.

[0097] S404: Determine the risk assessment level according to the risk prediction value.

[0098] Specifically, by establishing an asymmetric membership relationship between the risk prediction value and the safety level, smooth transition is achieved, avoiding level jumps, and solving the boundary jump problem caused by traditional hard - threshold division.

[0099] Specifically, let the risk prediction value of the battery system be ∈[0,1], and its safety - level set is L = {high, medium, low}. The membership functions corresponding to each level are defined as follows: Membership degree of high safety level : ; Membership degree of medium safety level

[0100] ; Membership degree of low safety level : ; Width of transition interval: A linear transition zone of 0.2 is set between adjacent levels (e.g., 0.3 → 0.5). Asymmetric design: The high → medium transition zone (0.3 - 0.5) is narrower than the medium → low transition zone (0.7 - 0.9), reflecting the normal distribution shift characteristics of the battery failure probability.

[0101] The risk assessment level is determined by the principle of maximum membership degree: ; Among them, when there are multiple maximum membership degrees, the decision is made according to the priority high > medium > low, and this design conforms to the conservative principle of battery safety assessment.

[0102] In an alternative embodiment, as shown in Figure 5 shown, Figure 5 The flowchart of the third method for determining the risk assessment level provided in the first embodiment of the present invention is shown, wherein the risk assessment level of the target battery is obtained by evaluating the target battery based on the safety assessment method, including steps S501 - S503: S501: When the safety assessment method is determined to be the time series prediction analysis method using a long short - term memory network, time series features including voltage fluctuation, power peak - to - average ratio, and temperature rise rate are constructed based on a sliding window for input.

[0103] Specifically, first, composite features are constructed according to historical operation data, including: Voltage fluctuation coefficient : ; Among them, is the data volume, that is, the number of sampling points; is the voltage value (V) at the i - th sampling point; is the average voltage within the time window; Power peak - to - average ratio :

[0104] Among them, is the maximum instantaneous power (kW), is the power mean value; Temperature rise rate :

[0105] where, t is the time window length (s).

[0106] Then, perform a sliding window process. Let the time window length be , the sliding step size be s, and construct a three-dimensional input tensor:

[0107] where, represents the set of real numbers, that is, each element in X is a real number, and the scale of the tensor is , m represents the total number of samples, d represents the feature dimension, that is, there are m samples, and the length of each sample in the time window dimension is , the length in the feature dimension is d, and N is the length of the original data.

[0108] S502: Extract the time series dependence relationship through the gating mechanism of the LSTM network to predict the risk prediction value within the future time window.

[0109] Specifically, first perform the LSTM gating calculation.

[0110] The cell state update formula is:

[0111] is the Sigmoid activation function, is the hyperbolic tangent activation function; , , , are all trainable weight matrices, are all bias vectors; is the hidden state at the previous moment (time t - 1), is the input at the current moment (time t), represents the output of the output gate at the current moment (time t); represents the updated cell state at the current moment (time t), which integrates the information of the forget gate, input gate, and candidate state, and preserves the long-term memory at the current moment, represents the cell state at the previous moment (time t - 1), which preserves the long-term memory information of the previous sequence; is the Hadamard product.

[0112] Then perform the safety factor mapping to predict the risk prediction value through the fully connected layer: ; Among them, is the weight of the fully connected layer, represents the hidden state of the LSTM network at the current time t, which contains the information of the previous sequence and is one of the inputs of the fully connected layer. is the bias vector of the fully connected layer, which is used to add an offset to the linear transformation result of the fully connected layer to improve the model fitting ability.

[0113] S503: Determine the risk assessment level according to the risk prediction value.

[0114] Specifically, according to the risk prediction value output by the LSTM ∈[0,1], an adaptive threshold method is used for three-level classification.

[0115]

[0116] Among them, = 0.4 is the basic high-risk threshold; = 0.7 is the basic low-safety threshold; is the standard deviation of the risk value within the current time window; k = 1.5 is the sigma coefficient, optimized based on the 3 principle.

[0117] Adopt a threshold dynamic adjustment mechanism. To avoid sudden noise interference, an exponential smoothing correction is introduced:

[0118] Among them, is the risk threshold after smoothing correction, is the basic high-risk threshold at the previous moment, = 0.85 is the smoothing factor (determined by grid search); is the moving average of the recent 100 risk prediction values; is the moving standard deviation of the recent risk prediction values.

[0119] This application adopts a dynamic threshold grading method, which combines statistical process control (SPC) and exponential smoothing prediction. Compared with the traditional fixed threshold method, it significantly enhances the anti-interference ability while ensuring the steady-state accuracy.

[0120] Embodiment 2 See Figure 6 as shown, Figure 6 shows a schematic structural diagram of an adaptive evaluation device for the operating safety state of an electrochemical energy storage battery provided by Embodiment 2 of the present invention. Among them, the device includes: An operating data acquisition module 601, configured to acquire historical operating data of a target battery in different working stages; A battery health indicator determination module 602, configured to determine a battery health indicator of the target battery based on the historical operating data in each working stage; An evaluation method determination module 603, configured to determine a safety evaluation method for the target battery based on the battery health indicator; An evaluation result determination module 604, configured to evaluate the target battery based on the safety evaluation method to obtain a risk evaluation level of the target battery.

[0121] In an optional implementation, the determining the battery health indicator of the target battery based on the historical operating data in each working stage includes: Determining health characteristics of the target battery in each working stage based on the historical operating data in each working stage; Performing standardization processing on the health characteristics in each working stage respectively, and assigning weights to the standardized health characteristics; Determining the battery health indicator based on the standardized health characteristics in each working stage and their weights.

[0122] In an optional implementation, the working stages include a charging state, a discharging state, and a static state; the health characteristics in the charging state include a slope of a capacity increment curve and a retention degree of a constant current charging duration; the health state in the discharging state includes a dynamic internal resistance and a retention degree of a discharging platform; the health characteristics in the static state include a voltage relaxation rate; The determining the health characteristics of the target battery in each working stage based on the historical operating data in each working stage includes: Determining the slope of the capacity increment curve according to a ratio of a voltage change amount to a capacity change amount; Determining the retention degree of the constant current charging duration according to a ratio of a constant current charging duration of an aged battery to that of an initial battery; Determining the dynamic internal resistance according to a ratio of a voltage drop during a discharging pulse to a discharging current; Determining the retention degree of the discharging platform according to a ratio of an integral of a voltage within a preset platform voltage range to a duration; Determining the voltage relaxation rate according to an exponential fitting parameter of a voltage relaxation process.

[0123] In an optional implementation, the determining the safety evaluation method for the target battery based on the battery health indicator includes: Classifying the health level of the target battery according to the battery health indicator, where the health levels are low health, medium health, and high health; When the target battery has a low health status, determine the safety assessment method as the abnormal fluctuation analysis method based on information entropy; When the target battery has a medium health status, determine the safety assessment method as the multi-parameter fuzzy logic fusion analysis method; When the target battery has a high health status, determine the safety assessment method as the time series prediction analysis method using a long short-term memory network.

[0124] In an optional implementation, evaluating the target battery based on the safety assessment method to obtain the risk assessment level of the target battery includes: When the safety assessment method is the abnormal fluctuation analysis method based on information entropy, discretize the voltage signal into multiple intervals and calculate the probability distribution of each interval; Calculate the normalized information entropy value of the voltage signal according to the probability distribution; Determine the risk assessment level according to the normalized information entropy value.

[0125] In an optional implementation, evaluating the target battery based on the safety assessment method to obtain the risk assessment level of the target battery includes: When the safety assessment method is the multi-parameter fuzzy logic fusion analysis method, extract the voltage fluctuation coefficient, current peak-to-average ratio, SOC change rate, and temperature gradient as input parameters; Perform fuzzy processing on each parameter through a predefined membership function and perform fuzzy inference based on an expert rule base; Use the area centroid method to defuzzify the inference result to obtain a risk prediction value; Determine the risk assessment level according to the risk prediction value.

[0126] In an optional implementation, evaluating the target battery based on the safety assessment method to obtain the risk assessment level of the target battery includes: When the safety assessment method is determined as the time series prediction analysis method using a long short-term memory network, construct a time series feature input including voltage fluctuation, power peak-to-average ratio, and temperature rise rate based on a sliding window; Extract the time series dependence relationship through the gating mechanism of the LSTM network and predict the risk prediction value within a future time window; Determine the risk assessment level according to the risk prediction value.

[0127] Embodiment III Based on the same application concept, refer to Figure 7 as shown Figure 7The figure shows a schematic structural diagram of a computer device provided in Embodiment 3 of the present invention. Among them, as Figure 7 shown, a computer device 700 provided in Embodiment 3 of the present application includes: A processor 701, a memory 702, and a bus 703. The memory 702 stores machine-readable instructions executable by the processor 701. When the computer device 700 runs, the processor 701 communicates with the memory 702 through the bus 703. When the machine-readable instructions are run by the processor 701, they execute the steps of the adaptive evaluation method for the operating safety state of the electrochemical energy storage battery shown in Embodiment 1 above.

[0128] Embodiment 4 Based on the same inventive concept, the present application embodiment also provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is run by a processor, it executes the steps of the adaptive evaluation method for the operating safety state of the electrochemical energy storage battery described in any one of the above embodiments.

[0129] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems and devices can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0130] The computer program product for adaptively evaluating the operating safety state of an electrochemical energy storage battery provided by the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, refer to the method embodiments and will not be repeated here.

[0131] The adaptive evaluation device for the operating safety state of the electrochemical energy storage battery provided by the embodiments of the present invention can be specific hardware on the device or software or firmware installed on the device, etc. For the device provided by the embodiments of the present invention, its implementation principle and the technical effects produced are the same as those of the foregoing method embodiments. For the sake of brief description, for the parts not mentioned in the device embodiments, reference can be made to the corresponding content in the foregoing method embodiments. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the foregoing-described systems, devices, and units can all refer to the corresponding processes in the above method embodiments, and will not be repeated here.

[0132] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For another example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some communication interfaces. The indirect coupling or communication connection of the devices or units can be in electrical, mechanical or other forms.

[0133] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0134] In addition, each functional unit in the embodiments provided by the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0135] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0136] It should be noted that similar reference numerals and letters indicate similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In addition, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0137] Finally, it should be noted that the above-described embodiments are only specific embodiments of the present invention, which are used to illustrate the technical solutions of the present invention, rather than limiting it. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that any person skilled in the technical field of the present invention can still modify the technical solutions described in the foregoing embodiments, or can easily conceive of changes, or perform equivalent replacements for some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. All should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. An adaptive evaluation method for the operating safety state of an electrochemical energy storage battery, characterized in that, The method includes: Obtaining historical operation data of the target battery in different working stages; Determining a battery health index of the target battery based on the historical operation data in each working stage; Determining a safety assessment method for the target battery based on the battery health index; Evaluating the target battery based on the safety assessment method to obtain a risk assessment level of the target battery.

2. The method according to claim 1, wherein The determining the battery health index of the target battery based on the historical operation data in each working stage includes: Determining health characteristics of the target battery in each working stage based on the historical operation data in each working stage; Performing standardization processing on the health characteristics in each working stage respectively, and assigning weights to the standardized health characteristics in each working stage; Determining the battery health index based on the standardized health characteristics in each working stage and their weights.

3. The method according to claim 2, wherein The working stages include a charging state, a discharging state, and a stationary state; the health characteristics in the charging state include a slope of a capacity increment curve and a retention degree of a constant current charging duration; the health state in the discharging state includes a dynamic internal resistance and a retention degree of a discharging platform; the health characteristics in the stationary state include a voltage relaxation rate. The determining the health characteristics of the target battery in each working stage based on the historical operation data in each working stage includes: Determining the slope of the capacity increment curve according to a ratio of a voltage change amount to a capacity change amount; Determining the retention degree of the constant current charging duration according to a ratio of a constant current charging duration of an aged battery to that of an initial battery; Determining the dynamic internal resistance according to a ratio of a voltage drop during a discharging pulse to a discharging current; Determining the retention degree of the discharging platform according to a ratio of an integral of a voltage within a preset platform voltage range to a duration; Determining the voltage relaxation rate according to an exponential fitting parameter of a voltage relaxation process.

4. The method according to claim 1, characterized in that, The determining the safety assessment method for the target battery based on the battery health index includes: Performing a health level classification on the target battery according to the battery health index, where the health levels are low health degree, medium health degree, and high health degree; When the target battery has a low health degree, determining the safety assessment method as an abnormal fluctuation analysis method based on information entropy; When the target battery has a medium health degree, determining the safety assessment method as a multi-parameter fuzzy logic fusion analysis method; When the target battery has a high health degree, determining the safety assessment method as a time series prediction analysis method using a long short-term memory network.

5. The method according to claim 4, characterized in that, The evaluating the target battery based on the safety assessment method to obtain a risk assessment level of the target battery includes: When the safety assessment method is an abnormal fluctuation analysis method based on information entropy, discretizing a voltage signal into multiple intervals and calculating a probability distribution of each interval; Calculating a normalized information entropy value of the voltage signal according to the probability distribution; Determining the risk assessment level according to the normalized information entropy value.

6. The method according to claim 4, characterized in that The evaluating the target battery based on the safety assessment method to obtain a risk assessment level of the target battery includes: When the safety assessment method is a multi-parameter fuzzy logic fusion analysis method, voltage fluctuation coefficient, current peak-to-average ratio, SOC change rate, and temperature gradient are extracted as input parameters; Each parameter is fuzzified through a predefined membership function, and fuzzy inference is performed based on an expert rule base; The area centroid method is used to defuzzify the inference result to obtain a risk prediction value; The risk assessment level is determined according to the risk prediction value.

7. The method according to claim 4, wherein The risk assessment level of the target battery is obtained by evaluating the target battery based on the safety assessment method, including: When the safety assessment method is determined to be a long short-term memory network for time series prediction analysis method, time series feature inputs including voltage fluctuation, power peak-to-average ratio, and temperature rise rate are constructed based on a sliding window; The time series dependence relationship is extracted through the gating mechanism of the LSTM network to predict the risk prediction value within a future time window; The risk assessment level is determined according to the risk prediction value.

8. An adaptive evaluation device for the operating safety state of an electrochemical energy storage battery, characterized in that, The device includes: An operating data acquisition module for acquiring historical operating data of the target battery at different operating stages; A health index determination module for determining the battery health index of the target battery based on the historical operating data at each operating stage; An assessment method determination module for determining the safety assessment method of the target battery based on the battery health index; An assessment result determination module for evaluating the target battery based on the safety assessment method to obtain the risk assessment level of the target battery.

9. A computer device, characterized in that, Including: A processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are executed by the processor, the steps of the adaptive assessment method for the operating safety state of an electrochemical energy storage battery as described in any one of claims 1 to 7 are executed.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is run by the processor, the steps of the adaptive assessment method for the operating safety state of an electrochemical energy storage battery as described in any one of claims 1 to 7 are executed.

Images