A method and system for managing energy storage power supply BMS based on gas concentration detection

Abstract

The present application provides a method and system for managing a BMS energy storage power supply based on gas concentration detection, including: receiving sampling data from a gas concentration sensor, establishing a gas concentration threshold judgment model, setting a leakage warning concentration threshold, and triggering a leakage warning signal when the concentration of the gas detected by the gas concentration sensor exceeds the leakage warning concentration threshold; if the leakage warning signal is received, increasing the sampling frequency of the gas concentration sensor within a preset range around the leakage point, and simultaneously starting the temperature sensor and the pressure sensor within the range to obtain multi-dimensional battery pack status data; performing a fusion operation on the multi-dimensional battery pack status data, and if the gas concentration continues to exceed the warning threshold and the temperature rise rate or the pressure rise rate exceeds the preset rate threshold, issuing a leakage confirmation signal and starting an emergency disposal mechanism.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method and system for managing an energy storage power supply BMS based on gas concentration detection. Background Art

[0002] Background of the problem: When setting up a gas concentration sensor inside an energy storage power source to detect battery leakage, a series of problems are faced, including sensor selection, layout location, sampling frequency, measurement range, sensitivity, stability, and the ability to withstand the harsh environment of the battery pack. Gas concentration itself is a process of dynamic change in time and space, and it is difficult for the sensor to accurately and comprehensively reflect the real-time concentration level of each area inside the battery pack. The concentration of gas leakage is extremely low at the beginning, which is easily disturbed by environmental factors, making it difficult to detect signs of problems in time. When the concentration exceeds the standard, it means that the leakage has spread to a larger area and the risk is higher. How to balance the cost of sensor layout and the comprehensiveness of detection, and balance sensitivity and false alarm rate, will be the key to determining the feasibility of this technical solution. Summary of the invention

[0003] The present invention provides a BMS management method for energy storage power supply based on gas concentration detection, which mainly includes:

[0004] The internal structural characteristics of the battery pack are obtained, and a battery pack gas diffusion model is established using computational fluid dynamics. The battery pack gas diffusion model includes gas diffusion coefficient, fluid density, pressure distribution and temperature field distribution parameters, and the gas concentration distribution data and flow trajectory inside the battery pack are obtained; based on the gas concentration distribution data, the gas concentration gradient change rate inside the battery pack is calculated. If the gas concentration gradient change rate is greater than a preset gradient threshold, the area is determined to be a gas concentration change sensitive area, and the area is used as a gas concentration sensor data layout point; a high-stability gas concentration sensor is installed at the sensor layout point, and the sampling frequency of the gas concentration sensor is dynamically adjusted according to the gas concentration change value. When the gas concentration is lower than the preset concentration value under normal operating conditions, the sampling frequency of the gas concentration sensor is adjusted. The sampling frequency is reduced within a limited time, and the sampling frequency is increased when the gas concentration is higher than the concentration value under normal operating conditions; the sampling data of the gas concentration sensor is received, a gas concentration threshold judgment model is established, and a leakage warning concentration threshold is set. When the concentration of the gas detected by the gas concentration sensor exceeds the leakage warning concentration threshold, a leakage warning signal is triggered; if the leakage warning signal is received, the sampling frequency of the gas concentration sensor within a preset range around the leakage point is increased, and the temperature sensor and pressure sensor within the range are started at the same time to obtain multi-dimensional battery pack status data; the multi-dimensional battery pack status data is fused and calculated, and if the gas concentration continues to exceed the warning threshold and the temperature rise rate or the pressure rise rate exceeds the preset rate threshold, a leakage confirmation signal is issued and the emergency disposal mechanism is started.

[0005] The present invention provides a BMS management system for energy storage power supply based on gas concentration detection, which mainly includes: a battery pack gas diffusion model module, which is used to obtain gas concentration distribution data and flow trajectory inside the battery pack; a gas concentration sensor layout module, which is used to determine the sensitive area of ​​gas concentration change as the sensor layout point; a gas concentration sensor sampling module, which is used to dynamically adjust the sampling frequency of the gas concentration sensor; a gas concentration threshold judgment module, which is used to set the leakage warning concentration threshold to trigger the leakage warning signal; a multi-dimensional battery pack status monitoring module, which is used to obtain multi-dimensional data such as gas concentration, temperature, pressure, etc. around the leakage point; a leakage confirmation and emergency disposal module, which is used to confirm the occurrence of leakage and start the emergency disposal mechanism.

[0006] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0007] The present invention discloses a method for monitoring gas leakage of a battery pack. In view of the problem that gas leakage inside the battery pack is difficult to detect in time and accurately locate, a gas diffusion model of the battery pack is established through computational fluid dynamics, the gas concentration distribution data and flow trajectory are analyzed, and the sensitive areas of gas concentration changes are determined as sensor layout points. High-stability gas concentration sensors are installed at these points, and the sampling frequency is dynamically adjusted according to the change in gas concentration. When it is detected that the gas concentration exceeds the warning threshold, the sampling frequency of the surrounding sensors is increased and the temperature and pressure sensors are started, and the multi-dimensional state data is fused and analyzed to achieve accurate judgment of leakage and timely warning. This method improves the accuracy and real-time performance of battery pack gas leakage monitoring through gas diffusion modeling and multi-sensor collaborative monitoring, and effectively reduces the risk of battery pack safety accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 The present invention is a flow chart of a method for managing an energy storage power supply BMS based on gas concentration detection.

[0009] Figure 2 It is a schematic diagram of a BMS management method and system for energy storage power supply based on gas concentration detection of the present invention.

[0010] Figure 3 This is another schematic diagram of a method and system for managing an energy storage power supply BMS based on gas concentration detection according to the present invention.

[0011] Figure 4 It is a structural schematic diagram of an energy storage power supply BMS management system based on gas concentration detection of the present invention. DETAILED DESCRIPTION

[0012] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0013] like Figure 1-4 In this embodiment, a BMS management system for energy storage power supply based on gas concentration detection may specifically include:

[0014] Step S101, obtaining the internal structural characteristics of the battery pack, and using computational fluid dynamics to establish a battery pack gas diffusion model, the battery pack gas diffusion model includes gas diffusion coefficient, fluid density, pressure distribution and temperature field distribution parameters, to obtain the gas concentration distribution data and flow trajectory inside the battery pack.

[0015] Acquire the internal structural characteristic data of the battery pack, the internal structural characteristic data of the battery pack including internal material properties and geometric dimension parameters; establish a battery pack gas diffusion model in computational fluid dynamics software according to the internal structural characteristic data of the battery pack, and set the boundary conditions and initial conditions of the battery pack gas diffusion model; in the battery pack gas diffusion model, set the gas diffusion coefficient and fluid density as key physical parameters, and combine the temperature field and pressure distribution inside the battery pack; use the finite volume method to solve the control equation of the battery pack gas diffusion model to obtain gas concentration distribution data and velocity field data at different positions inside the battery pack; use the streamline tracing algorithm to simulate the flow trajectory of the gas inside the battery pack according to the velocity field data; judge whether there is a risk of gas accumulation inside the battery pack according to the gas concentration distribution data and the gas flow trajectory; if it is found that the gas concentration exceeds the preset explosion limit threshold or the gas flow is blocked at the key position, it is determined that the internal structural design of the battery pack needs to be optimized; optimize the internal structural design of the battery pack to obtain an optimized internal structural design scheme of the battery pack, including increasing the ventilation space and adjusting the arrangement position of the tabs and the partitions.

[0016] Specifically, obtaining the data of the internal structural characteristics of the battery pack is to obtain the material properties and geometric dimensional parameters inside the battery pack. Material properties include the density, thermal conductivity, specific heat capacity, etc. of the various materials that make up the battery pack. For example, the shell material of the battery cell may be aluminum alloy, and its density, thermal conductivity, specific heat capacity and other parameters need to be known. The geometric dimensional parameters include the length, width and height of the battery pack, the arrangement of the battery cells, the size and position of the cooling plate, the size and shape of the ventilation channel, etc. For example, the precise geometric model of the battery pack can be obtained through a 3D scanner, and then the required dimensional parameters can be extracted from it. These data are the basis for establishing a gas diffusion model for the battery pack. Based on the obtained data of the internal structural characteristics of the battery pack, a gas diffusion model for the battery pack can be established in the computational fluid dynamics (CFD) software. Commonly used CFD software include Fluent, Star-CCM+, COMSOL, etc. The process of establishing the model includes importing the geometric model, dividing the mesh, defining the material properties, etc. For example, the 3D model of the battery pack can be imported into the Fluent software, and then the model can be meshed to discretize the model into many small control volumes. Setting the boundary conditions and initial conditions of the battery pack gas diffusion model is a crucial step. Boundary conditions refer to the way the battery pack interacts with the external environment. For example, the outside of the battery pack can be set to an insulating boundary condition, indicating that there is no heat transfer. Initial conditions refer to the state of the battery pack at the beginning of the simulation. For example, the initial temperature inside the battery pack can be set to 25°C, the initial pressure can be set to standard atmospheric pressure, and the initial gas concentration can be set to zero. In the battery pack gas diffusion model, some key physical parameters need to be set, such as gas diffusion coefficient and fluid density. The gas diffusion coefficient indicates how fast the gas diffuses inside the battery pack, which is related to factors such as temperature and pressure. Fluid density represents the ratio of the mass to volume of the gas, which is also related to factors such as temperature and pressure. At the same time, it is necessary to combine the temperature field and pressure distribution inside the battery pack for simulation. For example, during the operation of the battery, heat is generated, causing the temperature inside the battery pack to rise, which in turn affects the diffusion coefficient and density of the gas. The finite volume method can be used to solve the control equations of the battery pack gas diffusion model. The finite volume method is a commonly used numerical calculation method that discretizes the internal space of the battery pack into many small control volumes and then applies conservation equations to each control volume. By solving these equations, we can obtain the gas concentration distribution data and velocity field data at different positions inside the battery pack. Based on the obtained velocity field data, the streamline tracing algorithm can be used to simulate the flow trajectory of the gas inside the battery pack. The streamline tracing algorithm is a visualization method that can show how the gas flows inside the battery pack. By observing the flow trajectory of the gas, it can be determined whether there is a risk of gas accumulation inside the battery pack. Based on the obtained gas concentration distribution data and gas flow trajectory, it can be determined whether there is a risk of gas accumulation inside the battery pack.For example, if the gas concentration at a certain location exceeds the preset explosion limit threshold, or the flow of gas is obstructed, it means that there is a risk of gas accumulation, and the internal structure design of the battery pack needs to be optimized. If the gas concentration is found to exceed the preset explosion limit threshold or the gas flow is obstructed at a key location, the internal structure design of the battery pack needs to be optimized. The optimization directions include increasing the ventilation space and adjusting the layout of the tabs and partitions. For example, some ventilation channels can be added inside the battery pack to promote the circulation of gas, or the layout of the tabs and partitions can be adjusted to avoid gas accumulation. Assuming that the simulation results show that the gas concentration in a corner at the bottom of the battery pack is high, a vent can be added at that location and the vent can be connected to the outside of the battery pack to discharge the accumulated gas. In addition, if it is found that the gas flow is obstructed near the tab, the shape or position of the tab can be adjusted, or a guide structure can be added near the tab to improve the flow of gas.

[0017] Step S102, based on the gas concentration distribution data, calculate the gas concentration gradient change rate inside the battery pack. If the gas concentration gradient change rate is greater than a preset gradient threshold, determine that the area is a gas concentration change sensitive area, and use the area as a gas concentration sensor data layout point.

[0018] The gas concentration sensor data of different areas inside the battery pack are obtained, and the gas concentration sensor data are preprocessed to extract the statistical characteristics of the gas concentration in each area; the gas concentration gradient change rate of each area is calculated, and the threshold of the gas concentration gradient change rate is determined through experiments, and the area exceeding the threshold is marked as a gas concentration change sensitive area; according to the location distribution of the gas concentration change sensitive area, the K-means clustering algorithm is used to determine the layout points of the gas concentration sensors, and the gas concentration sensors are arranged near the cluster center; the gas concentration data collected by the gas concentration sensors at the layout points are obtained to construct a training data set for the support vector machine model; the optimal hyperparameters of the support vector machine model are determined by grid search and cross-validation to obtain a trained support vector machine classification model; the support vector machine classification model is used to judge the safety threshold of the newly collected gas concentration data. If it is predicted that the safety concentration threshold is exceeded, an audible and visual alarm is triggered, and the battery management system is controlled to perform cooling processing; a sliding window is used to detect the change trend of the gas concentration in each area inside the battery pack. If the concentration change trend in a certain area exceeds the preset threshold, the layout density of the gas concentration sensors in the area is dynamically adjusted.

[0019] Specifically, various gases are generated inside the battery pack. In order to monitor the concentration of these gases and ensure safety, a reliable monitoring system needs to be established. First, gas concentration sensors need to be arranged in different areas inside the battery pack. For example, sensors can be placed near battery cells, around cooling plates, and in ventilation channels. These sensors can monitor the gas concentration in different areas inside the battery pack in real time and transmit the data to the central processing unit. The raw data collected by the sensor may be affected by various noises, such as changes in ambient temperature, electromagnetic interference, etc. Therefore, these data need to be preprocessed to improve the reliability of the data. Preprocessing methods include removing outliers, smoothing filtering, etc. For example, a sliding average filter can be used to smooth the data and eliminate short-term fluctuations. In addition, the statistical characteristics of the gas concentration in each area, such as the mean value, standard deviation, etc., can also be calculated to better understand the distribution of gas concentrations. Assume that the gas concentration data collected in the top area of ​​the battery pack is: [10, 12, 11, 13, 15, 14, 16, 17, 18, 15], and the data obtained after preprocessing is: [11, 12, 12, 13, 14, 15, 16, 17, 17, 16]. In order to determine which areas are more sensitive to gas concentration changes, it is necessary to calculate the gas concentration gradient change rate of each area. For example, the difference in gas concentration between two adjacent time points can be calculated, and then divided by the time interval to obtain the gas concentration gradient change rate. The threshold of the gas concentration gradient change rate is determined by experiment. For example, the threshold can be set to 5ppm / s. The area exceeding the threshold is marked as a gas concentration change sensitive area. Assuming that the gas concentration in area A rises from 10ppm to 16ppm within 1 second, the gas concentration gradient change rate in this area is (16-10) / 1=6ppm / s, which exceeds the threshold of 5ppm / s, so area A is marked as a gas concentration change sensitive area. Next, it is necessary to determine the optimal layout points of the gas concentration sensors based on the location distribution of the gas concentration change sensitive areas. The K-means clustering algorithm can be used to divide these sensitive areas into several categories, and then the gas concentration sensors can be arranged near the center of each class. This can effectively cover all sensitive areas and reduce the number of sensors. Assuming that the K-means clustering algorithm divides the sensitive areas into three categories, one sensor can be arranged near the center of each class, for a total of three sensors. The gas concentration data collected by the gas concentration sensors at the layout points are obtained and divided into a training set and a test set. The training set is used to train the support vector machine model, and the test set is used to evaluate the performance of the model. For example, 80% of the data can be used as a training set and 20% of the data can be used as a test set. The performance of the support vector machine model is affected by hyperparameters, such as the kernel function type, penalty coefficient, etc. Grid search and cross-validation can be used to determine the optimal hyperparameters of the support vector machine model.Grid search refers to traversing all possible hyperparameter combinations within a certain range and selecting the combination with the best performance. Cross validation refers to dividing the training set into several parts, using one of them as the validation set each time, and the rest as the training set, and finally selecting the hyperparameter combination with the best average performance. The trained support vector machine classification model can be used to determine whether the newly collected gas concentration data exceeds the safe concentration threshold. If the prediction result is that the safe concentration threshold is exceeded, the sound and light alarm is triggered, and the battery management system is controlled to cool down. For example, the speed of the cooling fan can be controlled, or the coolant circulation system can be started to reduce the temperature of the battery pack to prevent thermal runaway. In order to adapt to the dynamic changes in the gas concentration distribution inside the battery pack, it is necessary to regularly monitor the changing trend of the gas concentration in each area. A sliding window can be used to detect the changing trend of the gas concentration in each area inside the battery pack. For example, a sliding window with a length of 10 can be used to calculate the average value of the gas concentration in the window, and compare whether the average value change of two adjacent windows exceeds the preset threshold. If the concentration change trend in a certain area exceeds the preset threshold, the density of the gas concentration sensor layout in the area is dynamically adjusted. For example, the number of sensors in the area can be increased, or the sensor can be moved closer to the location where the gas concentration changes drastically. Assuming that the gas concentration in area B has continued to rise over a period of time and exceeds a preset threshold, the number of sensors in area B can be increased to more accurately monitor the gas concentration in the area.

[0020] Step S103, install a high-stability gas concentration sensor at the sensor layout point, dynamically adjust the sampling frequency of the gas concentration sensor according to the gas concentration change value, reduce the sampling frequency when the gas concentration is lower than the preset lower limit of the normal operating concentration value, and increase the sampling frequency when the gas concentration is higher than the normal operating concentration value.

[0021] The historical gas concentration data of the gas concentration sensor is obtained, and the gas concentration value range under normal working conditions is determined by the K-means clustering algorithm; for each gas concentration sensor, a preset lower limit and a preset upper limit of the gas concentration are set as thresholds for triggering the sampling frequency adjustment; based on the historical gas concentration data, a dynamic adjustment model of the sampling frequency of the gas concentration sensor is established; the gas concentration data collected by each gas concentration sensor is obtained in real time to determine whether the gas concentration exceeds the normal working condition range; if the gas concentration is lower than the preset lower limit, the sampling frequency of the gas concentration sensor is reduced; if the gas concentration is higher than the preset upper limit, the sampling frequency of the gas concentration sensor is increased; according to the trend of gas concentration changes, the preset lower limit and the preset upper limit of the gas concentration are dynamically adjusted by the ARIMA time series prediction algorithm; the adjusted sampling frequency setting of the gas concentration sensor is synchronized to each gas concentration sensor node, and the working state of the gas concentration sensor is controlled by the intelligent gateway.

[0022] Specifically, the K-means clustering algorithm can determine the range of gas concentration values ​​under normal operating conditions based on historical gas concentration data. For example, the gas concentration data of a certain location inside the battery pack within one month is collected, and the K-means algorithm is used to divide these data into two categories: normal concentration and abnormal concentration. The data range in the normal concentration category can be used as the normal operating range of the gas concentration at that location. The advantage of this is that the normal range can be adaptively determined based on historical data, avoiding errors caused by manually setting thresholds. Setting a preset lower limit and a preset upper limit for each gas concentration sensor can be used as a threshold for triggering sampling frequency adjustment. For example, according to the normal operating range determined by the K-means clustering algorithm, the lower quartile of the range is set to the preset lower limit, and the upper quartile is set to the preset upper limit. Assuming that the normal concentration range of a sensor is 10ppm to 20ppm, the preset lower limit is 15ppm and the preset upper limit is 15ppm. This setting can capture abnormal changes in gas concentration more sensitively. A dynamic adjustment model for the sampling frequency of a gas concentration sensor is established, and the sampling frequency can be dynamically adjusted according to real-time gas concentration data. For example, when the gas concentration is lower than the preset lower limit, the sampling frequency is reduced; when the gas concentration is higher than the preset upper limit, the sampling frequency is increased. Assuming that the current sampling frequency is 1 time / second, when the concentration is lower than 15ppm, the sampling frequency is reduced to 5 times / second; when the concentration is higher than 15ppm, the sampling frequency is increased to 2 times / second. This can save system resources and improve monitoring accuracy under abnormal conditions. The gas concentration data collected by each sensor is obtained in real time, and it is determined whether the gas concentration exceeds the normal operating range. For example, if the current concentration collected by sensor A is 22ppm, which exceeds the preset upper limit of 15ppm, it is considered that the data of the sensor exceeds the normal range and the sampling frequency needs to be increased. This can timely detect and respond to abnormal changes in gas concentration. According to the trend of gas concentration changes, the preset lower limit and preset upper limit are dynamically adjusted using the ARIMA time series prediction algorithm. For example, if the gas concentration in a certain area continues to rise, and the ARIMA model predicts that the concentration will rise further in the future, the preset upper limit can be increased in advance, for example, from 15ppm to 19ppm, and the sampling frequency can be adjusted accordingly. This can predict the trend of gas concentration changes and take preventive measures in advance. The adjusted sampling frequency settings are synchronized to each sensor node, and the working status of the sensor is controlled by the smart gateway. For example, the sampling frequency of sensor A needs to be adjusted from 1 time / second to 2 times / second. The smart gateway will send the new sampling frequency setting to sensor A. After receiving the instruction, sensor A will sample according to the new frequency. This can achieve centralized control and management of all sensors. For example: Assume that there are three gas concentration sensors A, B, and C inside the battery pack, which are located in different positions of the battery pack.By analyzing historical data using the K-means clustering algorithm, the normal operating ranges of the three sensors were determined: A is 10-20ppm, B is 5-15ppm, and C is 15-25ppm. The preset lower and upper limits are set respectively, for example, A is 15ppm and 15ppm, B is 5ppm and 15ppm, and C is 15ppm and 25ppm. The initial sampling frequency is 1 time / second. At a certain moment, sensor A collects a concentration of 22ppm, which exceeds the upper limit of 15ppm, so the sampling frequency of A is increased to 2 times / second. At the same time, the ARIMA model predicts that the concentration in area A will continue to rise, so the preset upper limit of A is adjusted to 19ppm. Sensor B collects a concentration of 6ppm, which is lower than the lower limit of 5ppm, so the sampling frequency of B is reduced to 5 times / second. The concentration collected by sensor C is within the normal range, and the sampling frequency remains unchanged. The intelligent gateway synchronizes these adjusted parameters to each sensor node to control the working status of the sensor.

[0023] Step S104, receiving sampling data from the gas concentration sensor, establishing a gas concentration threshold judgment model, setting a leakage warning concentration threshold, and triggering a leakage warning signal when the gas concentration detected by the gas concentration sensor exceeds the leakage warning concentration threshold.

[0024] The gas concentration sampling data collected by the gas concentration sensor is obtained, and the gas concentration sampling data is transmitted to the data processing module; the data processing module preprocesses the received gas concentration sampling data, and adopts a moving average filtering algorithm to remove outliers and noise interference to obtain a smoothed gas concentration data sequence; according to the preprocessed gas concentration data sequence, a gas concentration threshold judgment model is established by adopting a support vector machine algorithm, and the gas concentration data sequence is divided into a training set and a test set; the statistical features of the gas concentration data in the training set are extracted, and the statistical features include mean, variance and kurtosis, and the extracted statistical features are used as input feature vectors of the support vector machine algorithm; for each sample in the training set, a corresponding gas concentration state label is set, and the gas concentration state label includes two states: normal and leakage; the support vector machine algorithm is trained by adopting the input feature vector and the corresponding gas concentration state label to obtain the gas concentration threshold judgment model; the kernel function type, penalty coefficient and kernel function of the support vector machine algorithm are optimized by using a grid search and cross-validation method. The gas concentration threshold judgment model is deployed to the gas leakage warning system, and the gas leakage warning concentration threshold is preset; the gas leakage warning system includes a data receiving module, a data processing module, a threshold judgment module and a warning signal sending module; the data receiving module receives the gas concentration data detected by the gas concentration sensor, and transmits the received gas concentration data to the data processing module; the data processing module pre-processes the received gas concentration data, extracts the gas concentration data features, and inputs the extracted gas concentration data features to the threshold judgment module; the threshold judgment module uses the deployed gas concentration threshold judgment model to classify and judge the input gas concentration data features, and if the judgment result is that the gas concentration exceeds the preset gas leakage warning concentration threshold, the gas leakage warning sign is output; if the threshold judgment module outputs the gas leakage warning sign, the warning signal sending module triggers the gas leakage warning signal, and notifies the relevant personnel to handle it by SMS or email.

[0025] Specifically, the moving average filtering algorithm can effectively remove outliers and noise interference in gas concentration sampling data. For example, a sensor collects a set of concentration data: 10ppm, 12ppm, 11ppm, 50ppm, 13ppm, 14ppm. Among them, 50ppm is likely to be an outlier. Using the moving average filtering algorithm with a window size of 3, the first filter value is (10+12+11) / 3=11ppm, the second filter value is (12+11+50) / 3=23ppm, the third filter value is (11+50+13) / 3=27ppm, and the fourth filter value is (50+13+14) / 3=27ppm, and so on. The filtered data sequence is smoother and can better reflect the true concentration change trend. This can improve the accuracy of subsequent models. The support vector machine algorithm is a commonly used classification algorithm that can judge the gas concentration state based on the statistical characteristics of gas concentration data. For example, the collected gas concentration data sequence is divided into a training set and a test set. Assume that the training set contains two types of data: normal state and leakage state. The mean, variance and kurtosis of each group of gas concentration data in the training set are extracted as feature vectors. For example, the feature vector of a normal state sample is (15ppm, 2ppm², 5), and the feature vector of a leakage state sample is (50ppm, 10ppm², 3). Use these feature vectors and the corresponding state labels to train the support vector machine model so that it can distinguish between the normal state and the leakage state. In this way, an effective concentration threshold judgment model can be established. Grid search and cross validation can be used to optimize the parameters of the support vector machine model and improve the judgment accuracy of the model. For example, when selecting the kernel function type, you can try linear kernel, polynomial kernel and radial basis function kernel. For each kernel function, you can use the grid search method to try different penalty coefficients and kernel function parameter combinations. For example, the penalty coefficient can be set to 1, 10, 100, and the kernel function parameters can be set to 1, 1, 10. Under each parameter combination, the cross validation method is used to evaluate the performance of the model, and finally the parameter combination with the best performance is selected. In this way, the most suitable model parameters can be found and the generalization ability of the model can be improved. The setting of the gas leakage warning concentration threshold needs to be determined according to the specific application scenario and safety requirements. For example, for some flammable and explosive gases, the warning threshold needs to be set lower so that the warning can be issued earlier. Assuming that the upper limit of the safe concentration of a certain gas is 20ppm, the warning threshold can be set to 15ppm. When the gas concentration exceeds 15ppm, the system will issue a warning signal to remind relevant personnel to take measures. This can effectively prevent the occurrence of safety accidents. Assume that a factory uses a gas leak warning system to monitor the gas concentration in the production workshop. The data receiving module receives concentration data from the sensor, such as 12ppm, 14ppm, 17ppm, 22ppm.The data processing module pre-processes the data, for example, using a moving average filter algorithm to remove noise, and then extracts feature vectors, such as (15ppm, 3ppm², 8). The threshold judgment module uses the trained support vector machine model to classify and judge the feature vectors. Assuming that the preset warning threshold is 15ppm, when the model determines that the concentration exceeds 15ppm (for example, 22ppm), an early warning sign is output. After receiving the early warning sign, the early warning signal sending module notifies relevant personnel via SMS or email, such as "The gas concentration in the workshop exceeds the standard, please deal with it immediately." This can detect and deal with gas leaks in a timely manner to ensure production safety.

[0026] Step S105, if the leakage warning signal is received, the sampling frequency of the gas concentration sensor within a preset range around the leakage point is increased, and the temperature sensor and the pressure sensor within the range are started at the same time to obtain multi-dimensional battery pack status data.

[0027] According to the pre-established gas concentration threshold, temperature threshold and pressure threshold under the normal working state of the battery pack, the gas concentration data, temperature data and pressure data of the battery pack are acquired in real time; if the gas concentration data exceeds the gas concentration threshold, and the temperature data and pressure data also exceed the temperature threshold and pressure threshold, it is judged that the battery pack is leaking, and a leakage warning signal is output; if a leakage warning signal is received, the sampling frequency of the gas concentration sensor within the preset range around the leakage point is increased to twice the normal frequency, and the temperature sensor and pressure sensor within the preset range are started at the same time to acquire the gas concentration state data, temperature state data and pressure state data of the battery pack; the Kalman filter algorithm is used to perform numerical analysis on the acquired gas concentration state data, temperature state data and pressure state data. According to the fusion, the optimal gas concentration, temperature and pressure of the battery pack are obtained; the optimal gas concentration, temperature and pressure are input into a pre-trained support vector machine (SVM) classification model to determine whether the battery pack is in a leakage state; if the battery pack is in a leakage state, an alarm is triggered; based on the optimal gas concentration, temperature and pressure, the particle swarm optimization algorithm (PSO) is used to estimate the location coordinates and leakage rate of the leakage point; the estimated leakage source location coordinates and leakage rate are input into a pre-established leakage diffusion numerical model to predict the diffusion range and concentration distribution of the leaked gas; based on the predicted diffusion range and concentration distribution of the leaked gas, new gas concentration sensors, temperature sensors and pressure sensors are deployed in areas where the concentration is higher than a preset threshold.

[0028] Specifically, the safety of the battery pack is of vital importance, and an effective leak detection and early warning system can minimize safety risks. The system monitors the gas concentration, temperature and pressure of the battery pack in real time, combined with advanced algorithms and models, to achieve accurate leak detection, positioning and diffusion prediction. First, the system pre-sets the gas concentration threshold, temperature threshold and pressure threshold of the battery pack under normal working conditions. The determination of these thresholds needs to be based on a large amount of experimental data and safety specifications. For example, for a certain model of battery pack, the gas concentration during normal operation is usually below 5ppm, the temperature is between 25℃ and 35℃, and the pressure is around 1 standard atmosphere. Therefore, the gas concentration threshold can be set to 10ppm, the temperature threshold to 40℃, and the pressure threshold to 2 standard atmospheres. These thresholds are not fixed and need to be adjusted according to actual conditions. The system obtains the gas concentration data, temperature data and pressure data of the battery pack in real time. These data can be collected through a sensor network deployed around the battery pack. When the gas concentration data exceeds the set gas concentration threshold, and the temperature data and pressure data also exceed the corresponding temperature threshold and pressure threshold respectively, the system preliminarily determines that the battery pack may have leaked and outputs a leak warning signal. This multi-parameter joint judgment method can effectively avoid misjudgment caused by fluctuations in a single parameter. For example, the temperature of the battery pack may rise during fast charging, but this does not necessarily mean that a leak has occurred. Only when the gas concentration, temperature and pressure are abnormal at the same time, a leak is more likely to occur. Once the system receives a leak warning signal, it will immediately increase the sampling frequency of the gas concentration sensor within the preset range around the leak point. For example, the sampling frequency is increased from once per minute to twice per second. At the same time, the temperature sensor and pressure sensor within the preset range are started to obtain more intensive and comprehensive data. This can capture the dynamic change process of the leak more quickly and accurately. For the large amount of gas concentration state data, temperature state data and pressure state data obtained, the system uses the Kalman filter algorithm for data fusion. The Kalman filter algorithm can effectively filter out noise and errors in the sensor data and estimate the optimal gas concentration, temperature and pressure values ​​of the battery pack. For example, the gas concentration data collected by multiple sensors may be slightly different. The Kalman filter algorithm can comprehensively consider these data and combine the system's state equation and observation equation to estimate the gas concentration closest to the true value. The obtained optimal values ​​of gas concentration, temperature and pressure are input into the pre-trained support vector machine (SVM) classification model. The model can determine whether the battery pack is in a leaking state based on these parameters. For example, if the model determines that the battery pack is in a leaking state, an alarm will be triggered and subsequent emergency procedures will be initiated. The system uses the particle swarm optimization algorithm (PSO) to estimate the location coordinates and leakage rate of the leak point.The particle swarm optimization algorithm is a heuristic search algorithm that can infer the location and leakage rate of the leakage source based on existing data, such as gas concentration values ​​at different locations. For example, if the gas concentration in a certain area is the highest and the concentration gradually decreases toward the surrounding area, it can be inferred that the leakage source is located in the center of the area, and the leakage rate is estimated based on the concentration gradient. The estimated leakage source location coordinates and leakage rate are input into the pre-established leakage diffusion numerical model. The model can simulate the diffusion process of the leaked gas in space and predict the diffusion range and concentration distribution of the leaked gas. For example, if the leak occurs indoors, the model can consider factors such as the geometry of the room and ventilation conditions to predict the concentration changes of the gas at different locations. Finally, according to the predicted diffusion range and concentration distribution of the leaked gas, new gas concentration sensors, temperature sensors, and pressure sensors are deployed in areas where the concentration is higher than the preset threshold. For example, if it is predicted that the gas concentration in a certain area is about to exceed the safety threshold, it is necessary to add sensors in the area to more accurately monitor the changes in gas concentration and take corresponding safety measures in time. This strategy of dynamically adjusting sensor deployment can improve the monitoring accuracy and response speed of the system.

[0029] Step S106, performing a fusion operation on the multi-dimensional battery pack status data, if the gas concentration continues to exceed the warning threshold and the temperature rise rate or the pressure rise rate exceeds the preset rate threshold, a leakage confirmation signal is issued and an emergency response mechanism is initiated.

[0030] Acquire multi-dimensional battery pack status data of the battery pack, pre-process the multi-dimensional battery pack status data to obtain pre-processed multi-dimensional battery pack status data, wherein the pre-processing includes denoising, outlier processing and data normalization; use the Kalman filter algorithm to fuse the pre-processed multi-dimensional battery pack status data to obtain a comprehensive status evaluation value of the battery pack; based on the comprehensive status evaluation value, use a support vector machine to build a battery pack safety status evaluation model; use the battery pack safety status evaluation model to make real-time judgments on the safety status of the battery pack; obtain the gas concentration parameters of the battery pack, and pre-set the normal value range and warning threshold for the gas concentration parameters; determine whether the gas concentration parameters exceed the warning threshold for multiple consecutive sampling cycles, and if so, trigger a gas concentration warning; obtain the temperature change rate and pressure change rate of the battery pack, and set the temperature change rate and pressure change rate respectively normal range; determine whether it is in the gas concentration warning period, if so, further determine whether the temperature change rate or pressure change rate exceeds the corresponding normal range, if so, determine it is a high-risk state of battery leakage; if it is determined to be a high-risk state of battery leakage, execute a predefined emergency disposal process, the emergency disposal process includes: cutting off the power supply of the faulty battery; isolating the faulty battery from other batteries; starting the exhaust system to strengthen ventilation; during the emergency disposal process, continuously collect the battery pack status data to obtain key parameters of gas concentration, temperature and pressure; determine whether each key parameter has returned to the normal range and remains stable within a preset time, if so, determine that the battery pack has returned to a safe state and generate a disposal completion signal; record and store the process data of the entire leakage event, the process data includes the triggering conditions when the leakage occurs, the duration of the leakage state, the emergency disposal measures and their effects.

[0031] Specifically, the core of the battery pack safety status assessment and leakage emergency disposal process is to ensure the safe operation of the battery system and prevent accidents caused by problems such as electrolyte leakage. This process realizes real-time monitoring and effective management of the battery pack safety status through multi-dimensional data collection, preprocessing, fusion and evaluation, combined with early warning mechanism and emergency disposal process. First of all, preprocessing the collected multi-dimensional battery pack status data is the key to ensuring data quality. The preprocessing process includes denoising, outlier processing and data normalization. For example, the temperature sensor reading of the battery pack may be disturbed by environmental noise, resulting in fluctuations in readings. Denoising techniques, such as Kalman filtering, can effectively remove these noise interferences. Outlier processing can identify and eliminate abnormal data caused by sensor failure or other reasons. For example, if a temperature sensor suddenly reads much higher than other sensors, it can be identified as an outlier and eliminated. Data normalization can convert data of different dimensions into the same dimension, which is convenient for subsequent data fusion and model training. For example, convert temperature data from Celsius to standard fractions for comparison with other parameters. The Kalman filter algorithm can fuse the pre-processed multi-dimensional data to obtain the comprehensive state evaluation value of the battery pack. The Kalman filter algorithm can combine the state equation and observation equation of the system to perform weighted averaging of the readings of multiple sensors to obtain a more accurate state estimate. For example, if the temperature and pressure of the battery pack increase at the same time, the Kalman filter algorithm can determine whether the battery pack is in a dangerous state based on the pre-set model parameters and output the corresponding comprehensive state evaluation value. Based on the comprehensive state evaluation value, a battery pack safety state evaluation model is constructed using a support vector machine. A support vector machine is a powerful classification algorithm that can learn the boundaries between different states based on training data. For example, a large amount of data on battery packs in normal and leaking states can be collected and used to train a support vector machine model. The model can determine whether the battery pack is in a leaking state based on the comprehensive state evaluation value of the battery pack. The battery pack safety state evaluation model can make real-time judgments on the safety state of the battery pack. For example, if the model determines that the battery pack is in a leaking state, a corresponding warning signal will be triggered. The gas concentration parameter is an important indicator for determining whether the battery pack is leaking. The normal value range and warning threshold are pre-set. For example, the normal value range of gas concentration is set to 0-5ppm, and the warning threshold is set to 10ppm. If the gas concentration exceeds the warning threshold for multiple consecutive sampling cycles, the gas concentration warning is triggered. For example, if the gas concentration exceeds 10ppm for three consecutive sampling cycles, the gas concentration warning is triggered. The temperature change rate and pressure change rate are also important indicators for judging whether the battery pack is leaking. Set the normal range of temperature change rate and pressure change rate. For example, set the normal range of temperature change rate to -1℃ / min~1℃ / min, and the normal range of pressure change rate to -1atm / min~1atm / min.During the gas concentration warning period, if the temperature change rate or pressure change rate exceeds the corresponding normal range, it is judged as a high-risk state of battery leakage. For example, if the temperature change rate reaches 2°C / min during the gas concentration warning period, it is judged as a high-risk state of battery leakage. If it is judged as a high-risk state of battery leakage, the predefined emergency disposal process is executed. The emergency disposal process includes: cutting off the power supply of the faulty battery to prevent further chemical reactions; isolating the faulty battery from other batteries to avoid chain reactions; starting the exhaust system to strengthen ventilation and reduce the concentration of leaked gas. During the emergency disposal process, the battery pack status data is continuously collected to obtain the key parameters of gas concentration, temperature and pressure. If each key parameter returns to the normal range and remains stable within a preset time, such as 30 minutes, it is judged that the battery pack has returned to a safe state and a disposal completion signal is generated. Finally, the process data of the entire leakage event is recorded and stored, including the triggering conditions when the leakage occurs, the duration of the leakage state, the emergency disposal measures and their effects. These data can be used for subsequent fault analysis and system optimization. For example, analyzing the triggering conditions when the leakage occurs can help improve the design and manufacturing process of the battery pack, thereby improving the safety of the battery pack.

[0032] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for managing energy storage power supply BMS based on gas concentration detection, characterized in that: The following steps are involved: S101, obtaining the internal structural characteristics of the battery pack, and establishing a battery pack gas diffusion model using computational fluid dynamics, wherein the battery pack gas diffusion model includes gas diffusion coefficient, fluid density, pressure distribution and temperature field distribution parameters, and obtaining gas concentration distribution data and flow trajectory inside the battery pack; S102, calculating the gas concentration gradient change rate inside the battery pack according to the gas concentration distribution data, and if the gas concentration gradient change rate is greater than a preset gradient threshold, determining that the area is a gas concentration change sensitive area, and using the area as a gas concentration sensor data layout point; S103, installing a high-stability gas concentration sensor at the sensor arrangement point, dynamically adjusting the sampling frequency of the gas concentration sensor according to the gas concentration change value, reducing the sampling frequency when the gas concentration is lower than the preset lower limit of the normal working condition concentration value, and increasing the sampling frequency when the gas concentration is higher than the normal working condition concentration value; S104, receiving sampling data from a gas concentration sensor, establishing a gas concentration threshold judgment model, setting a leakage warning concentration threshold, and triggering a leakage warning signal when the concentration of the gas detected by the gas concentration sensor exceeds the leakage warning concentration threshold; S105: if the leakage warning signal is received, the sampling frequency of the gas concentration sensor within a preset range around the leakage point is increased, and the temperature sensor and the pressure sensor within the range are started at the same time to obtain multi-dimensional battery pack status data; S106: Perform a fusion operation on the multi-dimensional battery pack status data. If the gas concentration continues to exceed the warning threshold and the temperature rise rate or the pressure rise rate exceeds the preset rate threshold, a leakage confirmation signal is issued and an emergency response mechanism is initiated.

2. A method for managing energy storage power supply BMS based on gas concentration detection according to claim 1, characterized in that: The step S101 comprises: Acquiring internal structural characteristic data of a battery pack, wherein the internal structural characteristic data of the battery pack includes internal material properties and geometric dimension parameters; According to the internal structural characteristic data of the battery pack, a battery pack gas diffusion model is established in computational fluid dynamics software, and boundary conditions and initial conditions of the battery pack gas diffusion model are set; In the battery pack gas diffusion model, the gas diffusion coefficient and fluid density are set as key physical property parameters, and the temperature field and pressure distribution inside the battery pack are combined; The finite volume method is used to solve the control equation of the battery pack gas diffusion model to obtain gas concentration distribution data and velocity field data at different positions inside the battery pack; According to the velocity field data, a streamline tracking algorithm is used to simulate the flow trajectory of the gas inside the battery pack; Determining whether there is a risk of gas accumulation inside the battery pack according to the gas concentration distribution data and the gas flow trajectory; If the gas concentration is found to exceed the preset explosion limit threshold or the gas flow is obstructed at the critical location, it is determined that the internal structure design of the battery pack needs to be optimized; The internal structure design of the battery pack is optimized to obtain an optimized internal structure design of the battery pack, including increasing the ventilation space and adjusting the layout of the tabs and partitions.

3. The energy storage power supply BMS management method based on gas concentration detection according to claim 1 is characterized in that: The step S102 comprises: Acquire gas concentration sensor data of different areas inside the battery pack, pre-process the gas concentration sensor data, and extract statistical characteristics of gas concentration in each area; Calculate the gas concentration gradient change rate in each area, determine the threshold of the gas concentration gradient change rate through experiments, and mark the area exceeding the threshold as a gas concentration change sensitive area; According to the location distribution of the gas concentration change sensitive area, the K-means clustering algorithm is used to determine the layout points of the gas concentration sensors, and the gas concentration sensors are arranged near the cluster center; Obtain gas concentration data collected by gas concentration sensors at deployment points and construct a training data set for a support vector machine model; Grid search and cross validation are used to determine the optimal hyperparameters of the support vector machine model and obtain a trained support vector machine classification model; The support vector machine classification model is used to judge the safety threshold of the newly collected gas concentration data. If it is predicted that the safety concentration threshold is exceeded, an audible and visual alarm is triggered, and the battery management system is controlled to perform a cooling process; A sliding window is used to detect the changing trend of gas concentration in each area inside the battery pack. If the concentration change trend in a certain area exceeds the preset threshold, the density of gas concentration sensors in that area is dynamically adjusted.

4. The energy storage power supply BMS management method based on gas concentration detection according to claim 1 is characterized in that: The step S103 comprises: Obtain historical gas concentration data from the gas concentration sensor and determine the gas concentration value range under normal working conditions using the K-means clustering algorithm; For each gas concentration sensor, a preset lower limit and a preset upper limit of the gas concentration are set as thresholds for triggering sampling frequency adjustment; According to the historical gas concentration data, a gas concentration sensor sampling frequency dynamic adjustment model is established; Obtain the gas concentration data collected by each gas concentration sensor in real time to determine whether the gas concentration exceeds the normal operating range; If the gas concentration is lower than the preset lower limit, the sampling frequency of the gas concentration sensor is reduced; If the gas concentration is higher than the preset upper limit, the sampling frequency of the gas concentration sensor is increased; According to the trend of gas concentration changes, the preset lower limit and upper limit of gas concentration are dynamically adjusted through the ARIMA time series prediction algorithm; The adjusted sampling frequency setting of the gas concentration sensor is synchronized to each gas concentration sensor node, and the working state of the gas concentration sensor is controlled by the intelligent gateway.

5. The energy storage power supply BMS management method based on gas concentration detection according to claim 1 is characterized in that: The step S104 comprises: Acquire gas concentration sampling data collected by the gas concentration sensor, and transmit the gas concentration sampling data to the data processing module; The data processing module pre-processes the received gas concentration sampling data, and uses a moving average filtering algorithm to remove abnormal values ​​and noise interference to obtain a smooth gas concentration data sequence; According to the preprocessed gas concentration data sequence, a gas concentration threshold judgment model is established using a support vector machine algorithm, and the gas concentration data sequence is divided into a training set and a test set; Extracting statistical features of the gas concentration data in the training set, the statistical features including mean, variance and kurtosis, and using the extracted statistical features as input feature vectors of the support vector machine algorithm; For each sample in the training set, a corresponding gas concentration state label is set, and the gas concentration state label includes two states: normal and leakage; The input feature vector and the corresponding gas concentration state label are used to train the support vector machine algorithm to obtain the gas concentration threshold judgment model; The grid search and cross-validation methods are used to optimize the kernel function type, penalty coefficient and kernel function parameters of the support vector machine algorithm to improve the judgment accuracy of the gas concentration threshold judgment model; Deploy the established gas concentration threshold judgment model into the gas leakage warning system, and preset the gas leakage warning concentration threshold; The gas leakage warning system comprises a data receiving module, a data processing module, a threshold judgment module and a warning signal sending module; The data receiving module receives the gas concentration data detected by the gas concentration sensor, and transmits the received gas concentration data to the data processing module; The data processing module pre-processes the received gas concentration data, extracts the gas concentration data features, and inputs the extracted gas concentration data features into the threshold judgment module; The threshold judgment module uses the deployed gas concentration threshold judgment model to classify and judge the input gas concentration data features. If the judgment result is that the gas concentration exceeds the preset gas leakage warning concentration threshold, a gas leakage warning sign is output; If the threshold judgment module outputs a gas leakage warning sign, the warning signal sending module triggers a gas leakage warning signal and notifies relevant personnel to handle the situation via SMS or email.

6. The energy storage power supply BMS management method based on gas concentration detection according to claim 1 is characterized in that: The step S105 comprises: According to the pre-established gas concentration threshold, temperature threshold and pressure threshold under the normal working state of the battery pack, the gas concentration data, temperature data and pressure data of the battery pack are obtained in real time; If the gas concentration data exceeds the gas concentration threshold, and the temperature data and the pressure data also exceed the temperature threshold and the pressure threshold, it is determined that the battery pack has leaked, and a leakage warning signal is output; If a leakage warning signal is received, the sampling frequency of the gas concentration sensor within the preset range around the leakage point is increased to twice the normal frequency, and the temperature sensor and pressure sensor within the preset range are started at the same time to obtain the gas concentration status data, temperature status data and pressure status data of the battery pack; Based on the acquired gas concentration state data, temperature state data and pressure state data, the Kalman filter algorithm is used for data fusion to obtain the optimal gas concentration value, temperature value and pressure value of the battery pack; Inputting the optimal value of gas concentration, the optimal value of temperature and the optimal value of pressure into a pre-trained support vector machine (SVM) classification model to determine whether the battery pack is in a leakage state; If the battery pack is in a leaking state, an alarm is triggered; According to the optimal values ​​of gas concentration, temperature and pressure, the particle swarm optimization algorithm PSO is used to estimate the location coordinates and leakage rate of the leakage point; The estimated leakage source location coordinates and leakage rate are input into the pre-established leakage diffusion numerical model to predict the diffusion range and concentration distribution of the leaked gas; Based on the predicted diffusion range and concentration distribution of the leaked gas, new gas concentration sensors, temperature sensors and pressure sensors are deployed in areas where the concentration is higher than the preset threshold.

7. The energy storage power supply BMS management method based on gas concentration detection according to claim 1 is characterized in that: The step S106 comprises: Acquire multi-dimensional battery pack status data of the battery pack, and pre-process the multi-dimensional battery pack status data to obtain pre-processed multi-dimensional battery pack status data, wherein the pre-processing includes denoising, outlier processing, and data normalization; Using a Kalman filter algorithm to fuse the pre-processed multi-dimensional battery pack status data to obtain a comprehensive status evaluation value of the battery pack; Based on the comprehensive state evaluation value, a battery pack safety state evaluation model is constructed using a support vector machine; Using the battery pack safety status assessment model to make a real-time judgment on the safety status of the battery pack; Obtaining gas concentration parameters of the battery pack, and presetting a normal value range and a warning threshold for the gas concentration parameters; Determine whether the gas concentration parameter exceeds the warning threshold for multiple consecutive sampling periods, and if so, trigger a gas concentration warning; Obtain the temperature change rate and pressure change rate of the battery pack, and set the normal ranges of the temperature change rate and pressure change rate respectively; Determine whether it is in the gas concentration warning period, and if so, further determine whether the temperature change rate or the pressure change rate exceeds the corresponding normal range, and if so, determine that it is a high-risk state of battery leakage; If it is determined to be a high-risk state of battery leakage, a predefined emergency handling process is executed, and the emergency handling process includes: cutting off the power supply of the faulty battery; Isolate the faulty battery from other batteries; Start the exhaust system to enhance ventilation; During the emergency response process, the battery pack status data is continuously collected to obtain key parameters of gas concentration, temperature and pressure; Determine whether each key parameter has returned to a normal range and remains stable within a preset time. If so, determine that the battery pack has returned to a safe state and generate a disposal completion signal; The process data of the entire leakage event is recorded and stored, and the process data includes the triggering conditions when the leakage occurs, the duration of the leakage state, and the emergency disposal measures and their effects.

8. A BMS management system for energy storage power supply based on gas concentration detection, characterized in that: The system comprises: Battery pack gas diffusion model module, used to obtain gas concentration distribution data and flow trajectory inside the battery pack; A gas concentration sensor layout module is used to determine the gas concentration change sensitive area as the sensor layout point; Gas concentration sensor sampling module, used to dynamically adjust the sampling frequency of the gas concentration sensor; A gas concentration threshold judgment module is used to set a leakage warning concentration threshold to trigger a leakage warning signal; Multi-dimensional battery pack status monitoring module, used to obtain multi-dimensional data such as gas concentration, temperature, and pressure around the leakage point; The leak confirmation and emergency response module is used to confirm the occurrence of a leak and initiate the emergency response mechanism.

Images