Screening method and device for lithium battery thermal runaway escape gas sensor

Abstract

The invention relates to a screening method and device of a lithium battery thermal runaway escape gas sensor. The method comprises the following steps: acquiring response recovery data obtained by performing a gas sensitive experiment on a plurality of sensors to be evaluated; according to the response recovery data, calculating to obtain a performance index; based on each performance index, executing a plurality of evaluation algorithms on each to-be-evaluated sensor to obtain a plurality of evaluation scores; performing fusion processing on the plurality of evaluation scores of each to-be-evaluated sensor to obtain a comprehensive evaluation score of each to-be-evaluated sensor; and selecting the to-be-evaluated sensor meeting a preset condition as a target sensor according to each comprehensive evaluation score. According to the method, multiple evaluation algorithms are adopted to evaluate a large number of sensor samples, and complex performance data are converted into visual quantitative scores, so that rapid comparison and screening among different types and batches of sensors are facilitated, and a sensor material or product with optimal performance can be found more accurately.

Description

Technical Field

[0001] This application relates to the field of detection technology, and in particular to a screening method and apparatus for a sensor of thermal runaway gas from a lithium battery. Background Technology

[0002] Lithium-ion batteries, with their advantages of high energy density, long cycle life, low self-discharge rate, and environmental friendliness, have become the main energy source for new energy vehicles and the electrochemical energy storage industry. Thermal runaway is one of the common failures that occur during the use of lithium-ion batteries. Lithium-ion thermal runaway causes the battery to release various flammable gases and generate enormous heat, potentially leading to battery fires and explosions, which limits the widespread application of lithium batteries in many fields. During thermal runaway, several characteristic gases are released, including carbon monoxide (CO), hydrogen (H2), methane (CH4), and ethylene (C2H4). These gases are key early warning signals for thermal runaway. Therefore, it is necessary to detect these flammable gases to ensure the proper operation of lithium batteries. In recent years, gas sensors based on metal oxide semiconductors have become a research hotspot due to their high sensitivity and stability, enabling early warning of lithium battery thermal runaway and providing an efficient solution for lithium battery safety.

[0003] Currently, semiconductor gas sensors used for lithium battery thermal runaway detection face numerous technical challenges. First, the compatibility of multiple sensors is ambiguous. For the same target gas, although different models and batches of sensors all have detection capabilities, they perform differently in terms of response speed, recovery ability, and long-term stability. In addition, previous studies have only used response values ​​as the evaluation criterion when exploring the optimal sensor for different gases. This approach is one-sided and makes it difficult to make a comprehensive judgment to select the optimal gas sensor. Summary of the Invention

[0004] Therefore, it is necessary to provide a screening method and device for lithium battery thermal runaway gas sensors to address the above-mentioned technical problems, so as to overcome the shortcomings of traditional technologies that make it difficult to make a comprehensive judgment and thus screen out the optimal gas sensor.

[0005] In a first aspect, this application provides a method for screening lithium battery thermal runaway gas sensors, the method comprising:

[0006] Acquire response recovery data obtained from gas-sensing experiments on multiple sensors to be evaluated;

[0007] Based on the response recovery data of each of the sensors to be evaluated, the performance indicators are calculated.

[0008] Based on the performance indicators, multiple evaluation algorithms are executed for each sensor to be evaluated to obtain multiple evaluation scores;

[0009] A fusion process is performed on the multiple evaluation scores of each sensor to be evaluated to obtain a comprehensive score for each sensor to be evaluated;

[0010] Based on the comprehensive scores described above, the sensor to be evaluated that meets the preset conditions is selected as the target sensor.

[0011] In one embodiment, the response recovery data includes the sensor's output signal and time parameters; the performance metrics include response time and recovery time.

[0012] The performance indicators calculated based on the response recovery data of each of the sensors to be evaluated include:

[0013] Based on the time parameter, the first time when the sensor to be evaluated comes into contact with the test gas and the second time when the output signal rises to a first preset threshold are obtained. Based on the difference between the second time and the first time, the response time is obtained.

[0014] Based on the time parameter, the third time when the sensor to be evaluated is removed from the test gas and the fourth time when the output signal drops to the second preset threshold are obtained. The recovery time is obtained based on the difference between the fourth time and the third time.

[0015] In one embodiment, the response recovery data includes the sensor's stable resistance; the performance metric includes the response value;

[0016] The performance indicators calculated based on the response recovery data of each of the sensors to be evaluated include:

[0017] Obtain the first stable resistance of the sensor to be evaluated when exposed to air;

[0018] Obtain the second stable resistance of the sensor to be evaluated when exposed to the test gas;

[0019] The response value is obtained based on the absolute value of the ratio of the first stabilizing resistor to the second stabilizing resistor.

[0020] In one embodiment, the performance metric also includes repeatability;

[0021] The step of calculating the performance index based on the response recovery data of each of the sensors to be evaluated also includes:

[0022] The standard deviation and average value of multiple response values ​​of the sensor to be evaluated are obtained, wherein the multiple response values ​​are obtained by performing multiple gas-sensitive experiments on the sensor to be evaluated;

[0023] The coefficient of variation is obtained based on the absolute value of the ratio of the standard deviation to the mean; the coefficient of variation is used to characterize the repeatability, and the smaller the coefficient of variation, the higher the repeatability.

[0024] In one embodiment, the performance metric further includes a resilience coefficient;

[0025] The step of calculating the performance index based on the response recovery data of each of the sensors to be evaluated also includes:

[0026] Obtain the third stable resistance of the sensor to be evaluated after it has been purged by air;

[0027] Obtain the first difference between the first stable resistor and the third stable resistor;

[0028] Obtain the second difference between the second stable resistor and the first stable resistor;

[0029] The recovery coefficient is obtained based on the absolute value of the ratio of the first difference to the second difference.

[0030] In one embodiment, the method further includes:

[0031] Preprocessing each of the performance indicators yields a performance variable matrix; wherein the preprocessing steps include:

[0032] Each of the aforementioned performance indicators is mapped to a corresponding performance variable, and a matrix is ​​constructed based on the sensor dimension and the indicator dimension.

[0033] For performance variables in the matrix whose evaluation direction is negative, perform monotonic direction correction to change the evaluation direction to positive.

[0034] The transformed performance variable matrix is ​​normalized to obtain the performance variable matrix.

[0035] In one embodiment, the evaluation algorithm includes the coefficient of variation method;

[0036] Based on the aforementioned performance indicators, multiple evaluation algorithms are executed on each of the sensors to be evaluated to obtain multiple evaluation scores, including:

[0037] The standard deviation and mean of the performance variable matrix are obtained, and the ratio of the standard deviation to the mean is calculated to obtain the coefficient of variation.

[0038] Calculate the sum of the coefficients of variation of all performance indicators, and calculate the ratio of the coefficient of variation of each performance indicator to the sum to obtain the weight of each performance indicator.

[0039] Based on the weights mentioned above, the performance variable matrix is ​​weighted and summed to obtain the evaluation score for each sensor to be evaluated.

[0040] In one embodiment, the evaluation algorithm further includes an ideal solution similarity ranking method;

[0041] The method of performing multiple evaluation algorithms on each of the aforementioned performance indicators to obtain multiple evaluation scores also includes:

[0042] Construct an information entropy probability matrix based on the performance variable matrix;

[0043] The information entropy of each performance index is calculated based on the information entropy probability matrix, and the entropy weight is determined based on the information entropy.

[0044] The performance variable matrix is ​​linearly weighted based on the entropy weight to obtain a weighted index matrix.

[0045] Calculate a first distance between each of the sensors to be evaluated and a first vector, and a second distance between each sensor and a second vector; wherein the first vector is constructed based on the maximum value of each performance index among all the sensors to be evaluated, and the second vector is constructed based on the minimum value;

[0046] The sum of the first distance and the second distance is calculated to obtain the third distance; the evaluation score of each sensor to be evaluated is obtained based on the second distance and the third distance.

[0047] In one embodiment, the evaluation algorithm further includes grey relational analysis.

[0048] The method of performing multiple evaluation algorithms on each of the aforementioned performance indicators to obtain multiple evaluation scores also includes:

[0049] For each performance index of each sensor to be evaluated, the absolute difference between the measured value of the performance index and the first vector is obtained; based on the global minimum value, global maximum value and preset resolution coefficient of the absolute difference, the corresponding grey relational coefficient is calculated.

[0050] The gray relation coefficients are linearly weighted based on the entropy weights to obtain the evaluation score for each sensor to be evaluated.

[0051] Secondly, this application also provides a screening device for a lithium battery thermal runaway gas sensor, the device comprising:

[0052] The acquisition module is used to acquire response recovery data obtained from gas-sensing experiments on multiple sensors to be evaluated.

[0053] The calculation module is used to calculate the performance index based on the response recovery data of each of the sensors to be evaluated;

[0054] The evaluation module is used to perform multiple evaluation algorithms on each of the aforementioned performance indicators to obtain multiple evaluation scores; and to perform fusion processing on the multiple evaluation scores of each of the aforementioned sensors to obtain a comprehensive score for each of the aforementioned sensors.

[0055] The selection module is used to select the sensor to be evaluated that meets the preset conditions as the target sensor based on the comprehensive score.

[0056] The aforementioned screening method and apparatus for lithium battery thermal runaway gas sensors involves acquiring response recovery data from gas-sensing experiments on multiple sensors to be evaluated; calculating performance indicators based on the response recovery data; executing multiple evaluation algorithms on each sensor to be evaluated based on each performance indicator to obtain multiple evaluation scores; performing fusion processing on the multiple evaluation scores of each sensor to be evaluated to obtain a comprehensive score for each sensor to be evaluated; and selecting the sensor to be evaluated that meets preset conditions as the target sensor based on the comprehensive scores. This application employs multiple evaluation algorithms to evaluate a large number of sensor samples, transforming complex performance data into intuitive quantitative scores, facilitating rapid comparison and screening between different types and batches of sensors. This allows for more accurate identification of the sensor material or product with optimal performance and can be used to practically screen the optimal sensor for detecting thermal runaway gas in lithium batteries. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 This is a structural block diagram of a gas-sensitive measurement system in one embodiment;

[0059] Figure 2 This is a schematic diagram of the sensor resistance measurement circuit in one embodiment;

[0060] Figure 3 This is a flowchart illustrating a screening method for a lithium battery thermal runaway gas sensor in one embodiment.

[0061] Figure 4 This is a schematic diagram showing the changes in response time and recovery time in one embodiment;

[0062] Figure 5 This is a schematic diagram of the response value change curve in one embodiment;

[0063] Figure 6This is a schematic diagram of repeatable variation curves in one embodiment;

[0064] Figure 7 This is a schematic diagram of the change curve of recovery capability in one embodiment;

[0065] Figure 8 This is a schematic diagram illustrating the comprehensive scoring of a sensor composed of multiple sensitive materials in one embodiment;

[0066] Figure 9 This is a structural block diagram of a screening device for a lithium battery thermal runaway gas sensor in one embodiment;

[0067] Figure 10 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0069] Please see Figure 1 In one exemplary embodiment, this application provides a gas-sensitive measurement system, including a gas supply module, a gas chamber, a data acquisition module, and a host computer.

[0070] The gas supply module, connected to the gas chamber, provides the test gas for gas-sensitive experiments. The module includes a gas tubing and a mass flow controller mounted on the tubing. One end of the tubing connects to the gas source, and the other end connects to the gas chamber. By adjusting the mass flow controller, the flow ratio between the background gas and the standard gas can be adjusted, thereby configuring a test gas of a specific concentration. The test gases include various characteristic gases such as carbon monoxide (CO), hydrogen (H2), methane (CH4), and ethylene (C2H4).

[0071] An air chamber is used to provide a sealed evaluation environment for the sensor.

[0072] The data acquisition module, connected to the sensor inside the gas chamber, is used to acquire the response recovery data generated by the sensor in gas-sensitive experiments. This response recovery data includes the sensor's output signal, time parameters, and the sensor's stable resistance in different gases.

[0073] Among them, the stable resistance is the resistance value at which the output resistance of the sensor reaches a stable state in the gas-sensitive experiment.

[0074] Please see Figure 2 , Figure 2 The circuit shown is for measuring the resistance of the sensor. Figure 2Vref is a stable reference voltage, used as the excitation source. The sensor and the variable resistor are connected in series between Vref and ground. Vsensor is the voltage drop across the sensor, Vsensor = Vref × (Rsensor / (Rpot + Rsensor)), where Rpot is the resistance of the variable resistor. By measuring Vsensor, the sensor resistance Rsensor can be deduced.

[0075] For example, the data acquisition module in this embodiment uses a high-speed data acquisition card (DAQ, NI USB-6218), which can simultaneously record 24 signals. Each channel is connected to a gas sensor via a DC voltage divider circuit. In the experiment, the total flow rate was set to 400 sccm, and the exposure time was set to 100 s.

[0076] The host computer, connected to the data acquisition module, is used to store and process the received response recovery data.

[0077] During the gas-sensitive experiment, the sensor to be evaluated is placed in the gas chamber and connected to the data acquisition module. The gas supply module introduces different test gases into the gas chamber to obtain the response recovery data of the sensor under different gas environments, which is then displayed and stored on the host computer software for subsequent analysis.

[0078] Please see Figure 3 In one exemplary embodiment, this application provides a method for screening lithium battery thermal runaway gas sensors. This screening method is based on the aforementioned gas-sensitive measurement system and includes the following steps:

[0079] Step 302: Obtain response recovery data from gas-sensing experiments on multiple sensors to be evaluated.

[0080] Specifically, conducting gas-sensitive experiments on the sensor to be evaluated involves exposing the sensor to different gas environments and obtaining a set of response recovery data through multiple experiments.

[0081] Step 304: Calculate the performance index based on the response recovery data of each sensor to be evaluated.

[0082] Specifically, performance metrics are used to characterize the working performance of a sensor. Traditional sensor evaluation methods often use a single evaluation dimension, typically considering only a single metric, making it difficult to comprehensively evaluate a sensor. This application's embodiments construct a multi-dimensional evaluation system that incorporates performance metrics including response value, response time, recovery time, recovery capability, and repeatability into a unified framework. Higher response values, faster response and recovery speeds, lower repeatability fluctuations, and better recovery capabilities collectively determine superior performance.

[0083] Step 306: Based on each performance index, perform multiple evaluation algorithms on each sensor to be evaluated to obtain multiple evaluation scores.

[0084] Specifically, the evaluation algorithm is used to quantitatively score and rank the performance of multiple sensors or the same sensor under different operating conditions to assist in selection, process optimization, or quality control. Exemplarily, the evaluation algorithm in this application includes at least one of the coefficient of variation method, ideal solution similarity ranking method, and grey relational analysis method.

[0085] The coefficient of variation method is a statistical method used to calculate the degree of change of each indicator in a system. It directly utilizes the information contained in each indicator to calculate the weight of the indicator.

[0086] The ideal solution similarity ranking method, also known as the superior-inferior solution distance method, is a commonly used comprehensive evaluation method. It primarily ranks the existing objectives based on their closeness to the idealized objective, thereby obtaining an evaluation of the relative superiority or inferiority of the existing objectives.

[0087] Grey relational analysis measures the relative strength of an item's influence by other factors within a grey system. The degree of correlation between factors in two systems, changing over time or between different objects, is called the degree of correlation. A high degree of synchronous change between two factors indicates a high degree of correlation; conversely, a low degree of correlation indicates a low degree of correlation. Therefore, grey relational analysis provides a quantitative measure of the degree of correlation between factors based on the similarity or dissimilarity of their development trends.

[0088] Step 308: Perform fusion processing on the multiple evaluation scores of each sensor to be evaluated to obtain the comprehensive score of each sensor to be evaluated; select the sensor to be evaluated that meets the preset conditions as the target sensor based on the comprehensive score.

[0089] Specifically, fusion processing refers to normalizing and summarizing the evaluation results obtained from different evaluation algorithms to obtain a unified score that reflects the overall performance of the sensor.

[0090] Optionally, the fusion process can employ various statistical methods, such as any one of the following: weighted average, arithmetic average, maximum value strategy, or voting method.

[0091] For example, the preset conditions can be set by methods such as scoring thresholds, ranking of the top few, and matching of performance index combinations. By comparing the comprehensive scores of each sensor to be evaluated with the preset conditions, one or more target sensors that meet the screening criteria are determined. These target sensors can be used in subsequent product finalization, mass production, or performance optimization applications.

[0092] The aforementioned screening method for lithium battery thermal runaway gas sensors constructs a multi-dimensional evaluation system and employs various evaluation algorithms to evaluate a large number of sensor samples. This transforms complex performance data into intuitive quantitative scores, facilitating rapid comparison and screening between different types and batches of sensors. It enables objective quantitative analysis and optimization of sensor performance, improves evaluation efficiency and accuracy, and can more accurately identify the sensor materials or products with the best performance. This method can be used to practically screen the optimal sensors for detecting lithium battery thermal runaway gas.

[0093] Optionally, when the response recovery data includes the sensor's output signal and time parameters, and the performance metrics include response time and recovery time, the performance metrics are calculated based on the response recovery data of each sensor to be evaluated, including:

[0094] Based on the time parameter, the first time when the sensor to be evaluated comes into contact with the test gas and the second time when the output signal rises to the first preset threshold are obtained. The response time is obtained based on the difference between the second time and the first time.

[0095] Based on the time parameter, the third time when the sensor to be evaluated is removed from the test gas and the fourth time when the output signal drops to the second preset threshold are obtained. The recovery time is obtained based on the difference between the fourth time and the third time.

[0096] Specifically, the time parameter is used to characterize the time process by which the sensor reaches a specific state during the response process. This application obtains the response time and recovery time based on the time parameter.

[0097] Please see Figure 4 The response time (tres) refers to the time interval from when the sensor comes into contact with the target gas until its output signal rises to a first preset threshold. For example, the first preset threshold is 90% of the stable value of the output signal. When the sensor comes into contact with a gas, its output electrical signal (such as changes in resistance or current) will gradually rise to a stable value. The time from contact with the gas until the signal reaches 90% of the stable value is the response time.

[0098] Recovery time (trec) refers to the time interval from when the sensor leaves the target gas environment until the output signal drops to a second preset threshold. For example, the second preset threshold is 10% of the initial value of the output signal. When the sensor leaves the gas environment, the output signal will gradually decrease to the initial state. The time from leaving the gas until the signal drops to 10% of the baseline value is the recovery time.

[0099] Optionally, if the response recovery data also includes the sensor's stabilizing resistance, and the performance metrics also include the response value, the performance metrics are calculated based on the response recovery data of each sensor to be evaluated, including:

[0100] Obtain the first stable resistance of the sensor to be evaluated when exposed to air.

[0101] Obtain the second stable resistance of the sensor to be evaluated when exposed to the test gas.

[0102] The response value is obtained based on the absolute value of the ratio of the first stable resistance to the second stable resistance.

[0103] Please see Figure 5 When a semiconductor sensor responds to a characteristic gas, the surface potential barrier changes, leading to a change in macroscopic resistance. Generally, the sensor's response value (S) can be defined as:

[0104]

[0105] Among them, R a R represents the first stable resistance. g This indicates the second stable resistor.

[0106] When the sensor is overly sensitive to the test gas, even exhibiting resistance changes on the order of magnitude, the definition of the above formula would cause S to approach 1, making it difficult to represent the response differences at different concentrations. To address this, this application uses the following formula to describe the response value (S):

[0107]

[0108] Optionally, if the performance metrics also include repeatability, the performance metrics calculated based on the response recovery data of each sensor to be evaluated may also include:

[0109] Obtain the standard deviation and average value of multiple response values ​​of the sensor to be evaluated, where multiple response values ​​are obtained by performing multiple gas-sensitive experiments on the sensor to be evaluated.

[0110] The coefficient of variation is obtained by taking the absolute value of the ratio of the standard deviation to the mean. The coefficient of variation is used to characterize repeatability, and the smaller the coefficient of variation, the higher the repeatability.

[0111] Please see Figure 6 Repeatability is another important indicator for gas sensors. For repeated tests, the coefficient of variation (CV) is introduced to evaluate the fluctuation of the sensor's response value in repeated experiments. The lower the CV value, the smaller the fluctuation of several response values, and it is described by the following formula:

[0112]

[0113] Among them, SD S M represents the standard deviation of multiple response values. S This represents the average of multiple response values.

[0114] Optionally, if the performance indicators also include the recovery capability coefficient, the performance indicators calculated based on the response recovery data of each sensor to be evaluated may also include:

[0115] Obtain the third stable resistance of the sensor to be evaluated after it has been purged by air.

[0116] Obtain the first difference between the first stable resistance and the third stable resistance.

[0117] Obtain the second difference between the second stable resistance and the first stable resistance.

[0118] The recovery coefficient is obtained based on the absolute value of the ratio of the first difference to the second difference.

[0119] Please see Figure 7 The recovery capability represents the baseline drift of the gas sensor and determines its ability to operate stably over a long period of time. It is defined as follows:

[0120]

[0121] Among them, R b R represents the third stabilizing resistance of the sensor after air purging. a R represents the first stable resistance. g This represents the second stable resistance. If R... a and R b If the difference is too large, it means that the sensor baseline has drifted and the recovery ability is poor.

[0122] Using the above scheme, the performance indicators of each sensor to be evaluated are calculated based on the response recovery data.

[0123] Optionally, the screening method for lithium battery thermal runaway gas sensors further includes:

[0124] Preprocessing each performance index yields a performance variable matrix.

[0125] The preprocessing steps include:

[0126] Each performance metric is mapped to a corresponding performance variable, and a matrix is ​​constructed based on the sensor dimension and the metric dimension.

[0127] For performance variables in the matrix whose evaluation direction is negative, perform monotonic direction correction to change the evaluation direction to positive.

[0128] The transformed performance variable matrix is ​​normalized to obtain the performance variable matrix.

[0129] Specifically, assuming there are n sensors to be evaluated, and each sensor has m performance metrics, the original matrix of performance metrics can be represented as follows:

[0130]

[0131] Where, x ij Let j be the performance variable of the i-th sample; i = 1, 2, ..., n; j = 1, 2, ..., m.

[0132] Furthermore, for ease of comparison, during the data preprocessing stage, all negative indicators are converted to positive indicators. All negative indicators are converted to positive indicators using the following formula:

[0133] X j =(-1) C ×X j

[0134] Where j = 1, 2, ..., m. C = 1 indicates a negative index, and C = 0 indicates a positive index.

[0135] Furthermore, considering that scale differences in performance variables may affect the screening results, this application also normalizes each performance variable to obtain the same expansion space. The normalized performance variable matrix can be represented as:

[0136]

[0137] Using the above scheme, the original matrix is ​​used to uniformly represent the performance of each sensor to be evaluated across multiple performance dimensions. Considering that the evaluation direction of some performance variables is negative, with smaller values ​​indicating better performance, a monotonic direction correction is performed on these variables to ensure that the evaluation direction of all performance variables is positive, i.e., larger values ​​indicate better performance, thus guaranteeing consistency in the evaluation direction. Finally, the matrix is ​​normalized to eliminate differences in units, dimensions, and numerical ranges among different performance variables, improving the comparability between performance variables and preventing unfair influences from individual indicators due to differences in units during subsequent scoring. This preprocessing significantly improves the adaptability and accuracy of subsequent evaluation algorithms, ensuring objective evaluation and precise ranking of the overall sensor performance, ultimately achieving reasonable selection of the target sensor.

[0138] Optionally, when the evaluation algorithm includes the coefficient of variation method, multiple evaluation algorithms are executed for each sensor to be evaluated based on various performance indicators to obtain multiple evaluation scores, including:

[0139] Obtain the standard deviation and mean of the performance variable matrix, calculate the ratio of the standard deviation to the mean, and obtain the coefficient of variation.

[0140] Calculate the sum of the coefficients of variation of all performance indicators, and calculate the ratio of the coefficient of variation of each performance indicator to the sum to obtain the weight of each performance indicator.

[0141] Based on each weight, the performance variable matrix is ​​weighted and summed to obtain the evaluation score for each sensor to be evaluated.

[0142] Specifically, the coefficient of variation method uses the coefficient of variation to determine the weights, reflecting the relative importance of each performance indicator. The coefficient of variation v j and weight ω j It can be calculated using the following formula:

[0143]

[0144] Where j = 1, 2, ..., m represents the index of each performance index; σ represents the mean; j It represents the standard deviation.

[0145] Furthermore, the evaluation results calculated based on the coefficient of variation method represent the performance of each sensor in detecting thermal runaway gases from lithium batteries, and are defined as follows:

[0146]

[0147] Where i = 1, 2, ..., n; j = 1, 2, ..., m.

[0148] Optionally, if the evaluation algorithm also includes an ideal solution similarity ranking method, multiple evaluation algorithms are executed for each sensor to be evaluated based on various performance indicators to obtain multiple evaluation scores, including:

[0149] Construct an information entropy probability matrix based on the performance variable matrix;

[0150] Calculate the information entropy of each performance index based on the information entropy probability matrix, and determine the entropy weights based on the information entropy;

[0151] The performance variable matrix is ​​linearly weighted based on entropy weighting to obtain a weighted index matrix.

[0152] Calculate the first distance between each sensor to be evaluated and the first vector, and the second distance between each sensor and the second vector; wherein the first vector is constructed based on the maximum value of each performance index among all sensors to be evaluated, and the second vector is constructed based on the minimum value.

[0153] The sum of the first distance and the second distance is calculated to obtain the third distance; the second distance and the third distance constitute the evaluation score for each sensor to be evaluated.

[0154] Specifically, first, select the optimal value for each performance metric. and worst value To construct the optimal ideal vector Z + and the worst ideal vector Z -The expression is:

[0155]

[0156] Where i = 1, 2, ..., n; j = 1, 2, ..., m.

[0157] Calculate the information entropy probability matrix p of the j-th performance variable for the i-th sample. ij and the information entropy e of the j-th performance variable j The expression is as follows:

[0158]

[0159] Furthermore, the weight ω based on the entropy weight method is obtained. j Its expression is:

[0160]

[0161] Furthermore, the proximity of each material to the optimal vector (Di+) and the worst vector (Di-) is calculated as follows:

[0162]

[0163] Furthermore, by estimating the distances to both positive and negative ideal materials, the evaluation results of the ideal solution similarity ranking method are obtained:

[0164]

[0165] Optionally, if the evaluation algorithm also includes grey relational analysis, multiple evaluation algorithms are executed for each sensor to be evaluated based on various performance indicators to obtain multiple evaluation scores, including:

[0166] For each performance index of each sensor to be evaluated, the absolute difference between the measured value of the performance index and the first vector is obtained; based on the global minimum and global maximum values ​​of the absolute difference and the preset resolution coefficient, the corresponding grey relational coefficient is calculated.

[0167] The grey relational coefficients are linearly weighted based on entropy weights to obtain the evaluation score for each sensor to be evaluated.

[0168] Specifically, the grey relational coefficient is defined as follows:

[0169]

[0170] Where ρ is 0.5; i = 1, 2, ..., n; j = 1, 2, ..., m; Z + This is the optimal ideal vector.

[0171] Furthermore, according to the information entropy e in the above embodimentsj The weights calculated through grey relational analysis can be derived from the following formula:

[0172] F GRA (i)=ω×ζ

[0173] Using the above scheme, three evaluation algorithms are employed to assess the performance of the sensor under evaluation, resulting in three evaluation results. These three results are then fused to quantitatively select the optimal sensor for detecting thermal runaway gases from lithium batteries. For example, the average score of the above evaluation results is calculated to obtain the final comprehensive score of the sensor, expressed as:

[0174] F = mean(F) VCM ,F TOPSIS ,F GRA )×100%

[0175] Among them, F VCM The score is based on the coefficient of variation method; F TOPSIS The score for the similarity ranking method for ideal solutions; F GRA The score is calculated using grey relational analysis. A higher overall score F for each sensor indicates better performance in detecting gases released during thermal runaway from lithium batteries.

[0176] Furthermore, using the above screening method, this application can conduct experiments on gas sensors made of materials with different concentrations of doping. Since gas sensors made of materials with different doping concentrations have different performance conversion scores, the gas-sensing experiments on the sensor to be evaluated can also intuitively demonstrate the differences in material performance under different doping combinations, providing an effective basis for screening high-performance sensor materials. Tables 1-7 show the research results of mixed oxide materials doped with different concentrations. The study found that when SnCl2 is doped into In2O3, a 10% SnCl2 doping amount can achieve a material performance quantification score of 100 points. Further doping with other oxides, such as 10% WO3, 5% NiO, 5% SnO2, and 10% TiO2, can all achieve a material performance quantification score of 100 points, while partial doping ratios can lead to a sharp drop in score or even 0 points. This indicates that a suitable doping concentration is crucial for improving sensor material performance, demonstrating that the evaluation method proposed in this application can intuitively demonstrate the differences in material performance under different doping combinations, providing an effective basis for screening high-performance sensor materials.

[0177] Table 1. Performance Quantitative Fractions of Sensors Fabricated with Different Concentrations of SnCl2 Doped with In2O3 Material

[0178] Material Performance Score In2O3 64 <![CDATA[In2O3+10%SnCl2]]> 100 <![CDATA[In2O3+20%SnCl2]]> 0

[0179] Table 2. Quantitative performance fractions of sensors made from 10% SnCl2-In2O3 materials doped with different concentrations of WO3.

[0180] Material Performance Score In₂O₃ + 10% SnCl₂ 31.26 In₂O₃ + 10% SnCl₂ + 5% WO₃ 77.09 In₂O₃ + 10% SnCl₂ + 10% WO₃ 100 In₂O₃ + 10% SnCl₂ + 20% WO₃ 0

[0181] Table 3. Quantitative performance fractions of sensors made from SnCl2-In2O3 materials doped with different concentrations of CuO.

[0182] Material Performance Score In₂O₃ + 10% SnCl₂ 100 In₂O₃ + 10% SnCl₂ + 5% CuO 0 In₂O₃ + 10% SnCl₂ + 10% CuO 23.05 In₂O₃ + 10% SnCl₂ + 20% CuO -

[0183] Table 4. Quantitative performance fractions of sensors made from SnCl2-In2O3 materials doped with different concentrations of NiO.

[0184] Material Performance Score In₂O₃ + 10% SnCl₂ 99.78 In₂O₃ + 10% SnCl₂ + 5% NiO 100 In₂O₃ + 10% SnCl₂ + 10% NiO 42.31 In₂O₃ + 10% SnCl₂ + 20% NiO 0

[0185] Table 5. Quantitative fractions of performance of sensors made from SnCl2-In2O3 materials doped with different concentrations of TiO2.

[0186] Material Performance Score In₂O₃ + 10% SnCl₂ 49.15 In₂O₃ + 10% SnCl₂ + 5% TiO₂ 0 In₂O₃ + 10% SnCl₂ + 10% TiO₂ 100 In₂O₃ + 10% SnCl₂ + 20% TiO₂ 94.45

[0187] Table 6. Quantitative fractions of performance of sensors made from SnCl2-In2O3 materials doped with different concentrations of SnO2.

[0188] Material Performance Score In₂O₃ + 10% SnCl₂ 34.76 In₂O₃ + 10% SnCl₂ + 5% SnO₂ 100 In₂O₃ + 10% SnCl₂ + 10% SnO₂ 48.37 In₂O₃ + 10% SnCl₂ + 20% SnO₂ 0

[0189] Table 7. Quantitative fractions of performance of sensors made from SnCl2-In2O3 materials doped with different concentrations of ZnO.

[0190] Material Performance Score In₂O₃ + 10% SnCl₂ 100 In₂O₃ + 10% SnCl₂ + 5% ZnO 99.39 In₂O₃ + 10% SnCl₂ + 10% ZnO 59.69 In₂O₃ + 10% SnCl₂ + 20% ZnO 0

[0191] To further verify the broad applicability of the proposed comprehensive evaluation method, this application further conducted comparative experiments on sensors made of different materials. Table 8 lists the comparison between the four prepared metal oxide composite materials, four commercial sensors, and four nanomaterial sensors in existing literature. These sensors can all be used to detect gases released during thermal runaway from lithium batteries. From a practical point of view, the four composite materials exhibit suitable resistivity similar to that of commercial gas sensors, indicating that they are easy to test and use. Analysis through the proposed comprehensive evaluation process revealed that 10% WO3-10% SnCl2-In2O3 achieved the highest comprehensive score.

[0192] Please see Figure 8 , Figure 8 This is a comprehensive score estimate for each sensor calculated based on Table 8. Figure 8In the diagram, a higher score indicates better sensor performance, and the scores for all comparison sensors are listed. The yellow area represents the ideal range for baseline resistance, but excessively low resistivity will lead to high power consumption. From Figure 8 As can be seen, the gas sensor made of 10% WO3-10% SnCl2-In2O3 material is the most effective at detecting gases released during thermal runaway of lithium batteries.

[0193] As can be seen, by using the screening method for lithium battery thermal runaway gas sensors of this application, the evaluation scores of semiconductor gas sensors with different concentrations can be obtained, and the material properties of the optimal doping combination can be selected. The gas sensor made in this way has high recognition accuracy and is not easily affected by other gases during operation, thus avoiding false alarms.

[0194] Table 8. Various sensitive materials for thermal runaway gases in lithium batteries and their ranking.

[0195]

[0196]

[0197] The aforementioned screening method for lithium battery thermal runaway gas sensors constructs a multi-dimensional evaluation system and employs various evaluation algorithms to evaluate a large number of sensor samples. This transforms complex performance data into intuitive quantitative scores, facilitating rapid comparison and screening between different types and batches of sensors. It enables objective quantitative analysis and optimization of sensor performance, improves evaluation efficiency and accuracy, and can more accurately identify the sensor materials or products with the best performance. This method can be used to practically screen the optimal sensors for detecting lithium battery thermal runaway gas.

[0198] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0199] Based on the same inventive concept, this application also provides a screening device for a lithium battery thermal runaway gas sensor, which implements the screening method for lithium battery thermal runaway gas sensors described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more device embodiments provided below can be found in the limitations of the method described above, and will not be repeated here.

[0200] Please see Figure 9 In one exemplary embodiment, this application also provides a screening device for a lithium battery thermal runaway gas sensor, including: an acquisition module, a calculation module, an evaluation module, and a selection module.

[0201] The acquisition module is used to acquire response recovery data obtained from gas-sensing experiments on multiple sensors to be evaluated.

[0202] The calculation module is used to calculate the performance indicators based on the response recovery data of each sensor to be evaluated.

[0203] The evaluation module is used to perform multiple evaluation algorithms on each sensor to be evaluated based on various performance indicators, and obtain multiple evaluation scores; and to perform fusion processing on the multiple evaluation scores of each sensor to be evaluated to obtain a comprehensive score of each sensor to be evaluated.

[0204] The selection module is used to select the sensor to be evaluated that meets the preset conditions as the target sensor based on the comprehensive scores.

[0205] Optionally, when the response recovery data includes the sensor's output signal and time parameters, and the performance indicators include response time and recovery time, the calculation module calculates the performance indicators based on the response recovery data of each sensor to be evaluated, including: based on the time parameters, obtaining the first time when the sensor to be evaluated comes into contact with the test gas, and the second time when the output signal rises to a first preset threshold, and obtaining the response time based on the difference between the second time and the first time; based on the time parameters, obtaining the third time when the sensor to be evaluated leaves the test gas, and the fourth time when the output signal drops to a second preset threshold, and obtaining the recovery time based on the difference between the fourth time and the third time.

[0206] Optionally, if the response recovery data also includes the sensor's stabilization resistance and the performance index also includes the response value, the calculation module calculates the performance index based on the response recovery data of each sensor to be evaluated, including: obtaining the first stabilization resistance of the sensor to be evaluated exposed to air; obtaining the second stabilization resistance of the sensor to be evaluated exposed to the test gas; and obtaining the response value based on the absolute value of the ratio of the first stabilization resistance and the second stabilization resistance.

[0207] Optionally, if the performance index also includes repeatability, the calculation module calculates the performance index based on the response recovery data of each sensor to be evaluated, and further includes: obtaining the standard deviation and average value of multiple response values ​​of the sensor to be evaluated, wherein the multiple response values ​​are obtained by performing multiple gas-sensitive experiments on the sensor to be evaluated; obtaining the coefficient of variation based on the absolute value of the ratio of the standard deviation and the average value; the coefficient of variation is used to characterize repeatability, and the smaller the coefficient of variation, the higher the repeatability.

[0208] Optionally, if the performance index also includes a recovery capability coefficient, the calculation module calculates the performance index based on the response recovery data of each sensor to be evaluated, and further includes: obtaining the third stabilizing resistance of the sensor to be evaluated after being blown by air; obtaining the first difference between the first stabilizing resistance and the third stabilizing resistance; obtaining the second difference between the second stabilizing resistance and the first stabilizing resistance; and obtaining the recovery capability coefficient based on the absolute value of the ratio of the first difference and the second difference.

[0209] Optionally, the screening device for the lithium battery thermal runaway gas sensor further includes a preprocessing module.

[0210] The preprocessing module is used to preprocess each performance index to obtain a performance variable matrix. The preprocessing steps include: mapping each performance index to its corresponding performance variable and constructing a matrix based on the sensor and index dimensions; performing monotonic direction correction on performance variables with negative evaluation directions in the matrix to convert the evaluation direction to positive; and normalizing the transformed performance variable matrix to obtain the final performance variable matrix.

[0211] Optionally, when the evaluation algorithm includes the coefficient of variation method, the evaluation module performs multiple evaluation algorithms on each sensor to be evaluated based on each performance index to obtain multiple evaluation scores, including: obtaining the standard deviation and mean of the performance variable matrix, calculating the ratio of the standard deviation to the mean to obtain the coefficient of variation; calculating the sum of the coefficients of variation of all performance indices, and calculating the ratio of the coefficient of variation of each performance index to the sum to obtain the weight of each performance index; and performing a weighted summation of the performance variable matrix according to each weight to obtain the evaluation score of each sensor to be evaluated.

[0212] Optionally, if the evaluation algorithm also includes the ideal solution similarity ranking method, the evaluation module performs multiple evaluation algorithms on each sensor to be evaluated based on each performance index to obtain multiple evaluation scores. This also includes: constructing an information entropy probability matrix based on the performance variable matrix; calculating the information entropy of each performance index based on the information entropy probability matrix and determining the entropy weight based on the information entropy; linearly weighting the performance variable matrix based on the entropy weight to obtain a weighted index matrix; calculating a first distance between each sensor to be evaluated and a first vector, and a second distance between each sensor and a second vector; wherein the first vector is constructed based on the maximum value of each performance index among all sensors to be evaluated, and the second vector is constructed based on the minimum value; calculating the sum of the first and second distances to obtain a third distance; the second and third distances constitute the evaluation score for each sensor to be evaluated.

[0213] Optionally, when the evaluation algorithm also includes grey relational analysis, the evaluation module performs multiple evaluation algorithms on each sensor to be evaluated based on each performance index to obtain multiple evaluation scores. This also includes: for each performance index of each sensor to be evaluated, obtaining the absolute difference between the measured value of the performance index and the first vector; calculating the corresponding grey relational coefficient based on the global minimum and global maximum values ​​of the absolute difference and the preset resolution coefficient; and linearly weighting the grey relational coefficient based on entropy weight to obtain the evaluation score of each sensor to be evaluated.

[0214] The aforementioned screening device for lithium battery thermal runaway gas sensors constructs a multi-dimensional evaluation system and employs various evaluation algorithms to evaluate a large number of sensor samples. It transforms complex performance data into intuitive quantitative scores, facilitating rapid comparison and screening between different types and batches of sensors. This enables objective quantitative analysis and optimization of sensor performance, improves evaluation efficiency and accuracy, and allows for more precise identification of the best-performing sensor materials or products. It can be used to practically screen the optimal sensors for detecting lithium battery thermal runaway gas.

[0215] Each module in the screening device for the lithium battery thermal runaway gas sensor can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0216] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 10As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When executed by the processor, the computer program implements a screening method for a lithium battery thermal runaway gas sensor. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0217] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0218] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the screening method for lithium battery thermal runaway gas sensors described above.

[0219] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the screening method for lithium battery thermal runaway gas sensors described above.

[0220] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the screening method for lithium battery thermal runaway gas sensors described above.

[0221] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0222] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0223] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for screening sensors for gases emitted during thermal runaway from lithium batteries, characterized in that, The method includes: Acquire response recovery data obtained from gas-sensing experiments on multiple sensors to be evaluated; Based on the response recovery data of each of the sensors to be evaluated, the performance indicators are calculated. Based on the performance indicators, multiple evaluation algorithms are executed for each sensor to be evaluated to obtain multiple evaluation scores; A fusion process is performed on the multiple evaluation scores of each sensor to be evaluated to obtain a comprehensive score for each sensor to be evaluated; Based on the comprehensive scores described above, the sensor to be evaluated that meets the preset conditions is selected as the target sensor.

2. The method according to claim 1, characterized in that, The response recovery data includes the sensor's output signal and time parameters; the performance indicators include response time and recovery time. The performance indicators calculated based on the response recovery data of each of the sensors to be evaluated include: Based on the time parameter, the first time when the sensor to be evaluated comes into contact with the test gas and the second time when the output signal rises to a first preset threshold are obtained. Based on the difference between the second time and the first time, the response time is obtained. Based on the time parameter, the third time when the sensor to be evaluated is removed from the test gas and the fourth time when the output signal drops to the second preset threshold are obtained. The recovery time is obtained based on the difference between the fourth time and the third time.

3. The method according to claim 1, characterized in that, The response recovery data includes the sensor's stable resistance; The performance metrics include response values; The performance indicators calculated based on the response recovery data of each of the sensors to be evaluated include: Obtain the first stable resistance of the sensor to be evaluated when exposed to air; Obtain the second stable resistance of the sensor to be evaluated when exposed to the test gas; The response value is obtained based on the absolute value of the ratio of the first stabilizing resistor to the second stabilizing resistor.

4. The method according to claim 2, characterized in that, The performance indicators also include repeatability; The step of calculating the performance index based on the response recovery data of each of the sensors to be evaluated also includes: The standard deviation and average value of multiple response values ​​of the sensor to be evaluated are obtained, wherein the multiple response values ​​are obtained by performing multiple gas-sensitive experiments on the sensor to be evaluated; The coefficient of variation is obtained based on the absolute value of the ratio of the standard deviation to the mean; the coefficient of variation is used to characterize the repeatability, and the smaller the coefficient of variation, the higher the repeatability.

5. The method according to claim 2, characterized in that, The performance indicators also include the recovery coefficient; The step of calculating the performance index based on the response recovery data of each of the sensors to be evaluated also includes: Obtain the third stable resistance of the sensor to be evaluated after it has been purged by air; Obtain the first difference between the first stable resistor and the third stable resistor; Obtain the second difference between the second stable resistor and the first stable resistor; The recovery coefficient is obtained based on the absolute value of the ratio of the first difference to the second difference.

6. The method according to claim 1, characterized in that, The method further includes: Preprocessing each of the performance indicators yields a performance variable matrix; wherein the preprocessing steps include: Each of the aforementioned performance indicators is mapped to a corresponding performance variable, and a matrix is ​​constructed based on the sensor dimension and the indicator dimension. For performance variables in the matrix whose evaluation direction is negative, perform monotonic direction correction to change the evaluation direction to positive. The transformed performance variable matrix is ​​normalized to obtain the performance variable matrix.

7. The method according to claim 6, characterized in that, The evaluation algorithm includes the coefficient of variation method; Based on the aforementioned performance indicators, multiple evaluation algorithms are executed on each of the sensors to be evaluated to obtain multiple evaluation scores, including: The standard deviation and mean of the performance variable matrix are obtained, and the ratio of the standard deviation to the mean is calculated to obtain the coefficient of variation. Calculate the sum of the coefficients of variation of all performance indicators, and calculate the ratio of the coefficient of variation of each performance indicator to the sum to obtain the weight of each performance indicator. Based on the weights mentioned above, the performance variable matrix is ​​weighted and summed to obtain the evaluation score for each sensor to be evaluated.

8. The method according to claim 7, characterized in that, The evaluation algorithm also includes an ideal solution similarity ranking method; The method of performing multiple evaluation algorithms on each of the aforementioned performance indicators to obtain multiple evaluation scores also includes: Construct an information entropy probability matrix based on the performance variable matrix; The information entropy of each performance index is calculated based on the information entropy probability matrix, and the entropy weight is determined based on the information entropy. The performance variable matrix is ​​linearly weighted based on the entropy weight to obtain a weighted index matrix. Calculate a first distance between each of the sensors to be evaluated and a first vector, and a second distance between each sensor and a second vector; wherein the first vector is constructed based on the maximum value of each performance index among all the sensors to be evaluated, and the second vector is constructed based on the minimum value; The sum of the first distance and the second distance is calculated to obtain the third distance; the evaluation score of each sensor to be evaluated is obtained based on the second distance and the third distance.

9. The method according to claim 8, characterized in that, The evaluation algorithm also includes grey relational analysis. The method of performing multiple evaluation algorithms on each of the aforementioned performance indicators to obtain multiple evaluation scores also includes: For each performance index of each sensor to be evaluated, the absolute difference between the measured value of the performance index and the first vector is obtained; based on the global minimum value, global maximum value and preset resolution coefficient of the absolute difference, the corresponding grey relational coefficient is calculated. The gray relation coefficients are linearly weighted based on the entropy weights to obtain the evaluation score for each sensor to be evaluated.

10. A screening device for a lithium battery thermal runaway gas sensor, characterized in that, The device includes: The acquisition module is used to acquire response recovery data obtained from gas-sensing experiments on multiple sensors to be evaluated. The calculation module is used to calculate the performance index based on the response recovery data of each of the sensors to be evaluated; The evaluation module is used to perform multiple evaluation algorithms on each of the aforementioned performance indicators to obtain multiple evaluation scores; and to perform fusion processing on the multiple evaluation scores of each of the aforementioned sensors to obtain a comprehensive score for each of the aforementioned sensors. The selection module is used to select the sensor to be evaluated that meets the preset conditions as the target sensor based on the comprehensive score.

Images