Thermal Disaster Alarm System and Electronic Equipment for Lithium-Ion Battery Energy Storage Power Station

Abstract

The present invention discloses a thermal disaster alarm system and an electronic device for a lithium-ion battery energy storage power station, which relates to the technical field of thermal disaster monitoring. The system includes: a power station division module for dividing the lithium-ion energy storage power station into M thermal disaster risk control monitoring areas; a node configuration module for configuring M edge thermal runaway prediction nodes; a thermal runaway suppression decision module for outputting a real-time thermal runaway prevention and control strategy; a decision generation module for generating a real-time fire extinguishing risk control decision according to the evolution result; a risk level mapping module for obtaining the real-time risk level; and an alarm signal sending module for sending a thermal disaster alarm signal to the power station operation and maintenance center. The present invention solves the technical problem of slow fire risk early warning and prevention and control response speed of the lithium-ion battery energy storage power station, and achieves the technical effects of accurate thermal runaway risk prediction and timely prevention and control based on real-time prediction of edge nodes and cloud decision optimization, and improving the fire early warning and emergency response efficiency of the power station.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal disaster monitoring, and particularly to a thermal disaster alarm system and an electronic device for a lithium-ion battery energy storage power station. Background Art

[0002] With the transformation of the global energy structure, energy storage technology has become one of the key technologies to support the development of renewable energy. In particular, lithium-ion battery energy storage systems have shown broad application prospects in regulating grid load, improving power quality, and providing backup power. Lithium-ion battery energy storage power stations have become an important part of modern power systems due to their high efficiency and high-density power storage capabilities. However, with the continuous increase in the capacity and power density of energy storage power stations, the safety issues of lithium-ion batteries have gradually emerged. When lithium-ion batteries encounter extreme operating conditions, equipment failures, or environmental changes, thermal runaway may occur, which is usually accompanied by the release of a large amount of heat and toxic gases, and may even lead to fire or explosion accidents in severe cases, posing significant safety hazards. For example, the frequent occurrence of lithium-ion battery fires in recent years has attracted wide attention and exposed the weak links in the current fire prevention and control of energy storage power stations. Existing thermal disaster alarm systems often have problems such as untimely response, inability to accurately predict the trend of fire spread, and difficulty in making effective responses quickly. These systems fail to make full use of advanced edge computing and cloud decision-making technologies, resulting in the inability to conduct real-time and accurate risk assessment and emergency decision-making in the event of thermal runaway. Summary of the Invention

[0003] This application provides a thermal disaster alarm system and an electronic device for a lithium-ion battery energy storage power station, which are used to solve the technical problem of slow response speed in the early warning and prevention and control of fire risks in lithium-ion battery energy storage power stations.

[0004] In view of the above problems, this application provides a thermal disaster alarm system and an electronic device for a lithium-ion battery energy storage power station.

[0005] In the first aspect of this application, a thermal disaster alarm system for a lithium-ion battery energy storage power station is provided. The system includes:

[0006] A power station division module, configured to divide a lithium-ion energy storage power station into M thermal disaster risk control and monitoring areas through thermal runaway propagation evolution; a node configuration module, configured to perform risk control sensitivity analysis on the M thermal disaster risk control and monitoring areas, and configure M edge thermal runaway prediction nodes in the M thermal disaster risk control and monitoring areas according to the analysis results, wherein the M edge thermal runaway prediction nodes are communicatively connected to a thermal disaster alarm decision cloud; a thermal runaway suppression decision module, configured to receive, by the thermal disaster alarm decision cloud, and perform thermal runaway suppression decision according to M real-time thermal runaway risk characteristics transmitted back by the M edge thermal runaway prediction nodes, and output a real-time thermal runaway prevention and control strategy, wherein the real-time thermal runaway prevention and control strategy has a prevention and control time window identifier; a decision generation module, configured to perform thermal runaway prevention and control evolution based on the real-time thermal runaway prevention and control strategy, and generate a real-time fire extinguishing risk control decision according to the evolution result; a risk level mapping module, configured to traverse a risk level mapping table by using the prevention and control time window identifier and the real-time fire extinguishing risk control decision to obtain a real-time risk level; an alarm signal sending module, configured to package the real-time risk level, the real-time thermal runaway prevention and control strategy, and the real-time fire extinguishing risk control decision into a thermal disaster alarm signal and send it to a power station operation and maintenance center.

[0007] In a second aspect of the present application, there is provided an electronic device, including: a memory, configured to store executable instructions; a processor, configured to implement a thermal disaster alarm system for a lithium-ion battery energy storage power station when executing the executable instructions stored in the memory.

[0008] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0009] By performing thermal runaway propagation evolution, a lithium-ion energy storage power station is divided into M thermal disaster risk control and monitoring areas; risk control sensitivity analysis is performed on the M thermal disaster risk control and monitoring areas, and M edge thermal runaway prediction nodes are configured in the M thermal disaster risk control and monitoring areas according to the analysis results; the thermal disaster alarm decision cloud receives and performs thermal runaway suppression decision according to M real-time thermal runaway risk characteristics transmitted back by the M edge thermal runaway prediction nodes, and outputs a real-time thermal runaway prevention and control strategy; thermal runaway prevention and control evolution is performed based on the real-time thermal runaway prevention and control strategy, and a real-time fire extinguishing risk control decision is generated according to the evolution result; a real-time risk level is obtained by traversing a risk level mapping table by using the prevention and control time window identifier and the real-time fire extinguishing risk control decision; the real-time risk level, the real-time thermal runaway prevention and control strategy, and the real-time fire extinguishing risk control decision are packaged into a thermal disaster alarm signal and sent to a power station operation and maintenance center. The technical effect of accurate thermal runaway risk prediction and timely prevention and control based on real-time prediction of edge nodes and cloud decision optimization is achieved, and the fire warning and emergency response efficiency of the power station is improved. Description of the Drawings

[0010] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0011] Figure 1 It is a schematic flow chart of the thermal disaster alarm system for a lithium-ion battery energy storage power station provided by an embodiment of the present application.

[0012] Figure 2 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application.

[0013] Explanation of reference numerals: power station division module 11, node configuration module 12, thermal runaway suppression decision module 13, decision generation module 14, risk level mapping module 15, alarm signal sending module 16, processor 21, memory 22, input device 23, output device 24. Detailed implementation manners

[0014] The present application provides a thermal disaster alarm system for a lithium-ion battery energy storage power station and an electronic device, which are used to solve the technical problem of slow fire risk early warning and prevention and control response speed of the lithium-ion battery energy storage power station.

[0015] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts belong to the scope of protection of the present application.

[0016] Embodiment 1, as Figure 1 shown, the present application provides a thermal disaster alarm system for a lithium-ion battery energy storage power station, and the system includes:

[0017] A power station division module 11, configured to divide the lithium-ion energy storage power station into M thermal disaster risk control and monitoring areas through the propagation and evolution of thermal runaway.

[0018] Specifically, the power station division module 11 first divides the lithium-ion energy storage power station into M thermal disaster risk control monitoring areas according to the possible thermal runaway expansion of the battery in different areas through the thermal runaway propagation evolution method. Specifically, when the battery pack has thermal runaway, a large amount of heat will be released, and may trigger a chain reaction of the surrounding batteries. This process spreads gradually. Therefore, by analyzing the temperature gradient between battery packs, it is possible to predict which areas are most likely to have thermal runaway, and then divide different monitoring areas. To this end, the system terminal combines different historical energy storage scenarios and analyzes the temperature gradient between battery packs through a thermal runaway propagation model to identify and divide M thermal disaster risk control monitoring areas. The division of each area helps to conduct targeted risk assessment and monitoring according to the different working conditions of the battery and the environment, so as to timely discover potential thermal runaway risks and avoid disasters such as fire or explosion.

[0019] In a possible implementation, the lithium-ion energy storage power station is divided into M thermal disaster risk control monitoring areas through the evolution of thermal runaway propagation. The system includes:

[0020] The risk control factors of the lithium-ion energy storage power station are mined to obtain K thermal disaster risk control factors; the risk control monitoring points are located in the lithium-ion energy storage power station according to the K thermal disaster risk control factors, and a power station monitoring plan is generated, wherein the power station monitoring plan includes K groups of thermal disaster risk control monitoring points; multiple historical energy storage scenarios of the lithium-ion energy storage power station are called; the thermal runaway propagation evolution of the K groups of thermal disaster risk control monitoring points is performed using the multiple historical energy storage scenarios, and the lithium-ion energy storage power station is divided into the M thermal disaster risk control monitoring areas.

[0021] Specifically, in the power station division module 11, the system terminal first uses the thermal runaway accident data of similar lithium-ion battery energy storage power stations to mine the risk control factors that may affect battery safety, and obtains K types of thermal disaster risk control factors. These factors include, but are not limited to, battery temperature, current, voltage, environmental humidity, charge and discharge rate, and battery aging state. These factors are potential triggering factors for thermal disasters. Therefore, by analyzing these factors in detail, it is possible to provide basic data for subsequent monitoring point configuration and risk warning. Subsequently, according to the K types of thermal disaster risk control factors mined, the risk control monitoring points are located in the power station. That is, the points with the same monitoring equipment in the lithium-ion energy storage power station are counted, and K groups of thermal disaster risk control monitoring points are obtained. Each group of thermal disaster risk control monitoring points corresponds to a type of thermal disaster risk control factor. For example, if point a has a current sensor and point b also has a current sensor, then point a and point b will be stored in the same group of thermal disaster risk control monitoring points. After locating the risk control factors and monitoring points, the system terminal will call multiple historical energy storage scenarios. These scenarios are based on the records of the actual operation of the power station in the past or experimental data, which can help understand the working state of the battery pack and the power station and the possible thermal runaway situations under different conditions. For example, the historical energy storage scenarios may include situations where the battery temperature rises and the battery charge and discharge rate is relatively fast. These data help understand the occurrence mode of thermal runaway. After that, using these historical energy storage scenarios, the system terminal will analyze each monitoring point through the thermal runaway propagation model, simulating how factors such as temperature rise and internal battery reactions during thermal runaway spread between battery packs. Through this propagation evolution, the system terminal can understand the temperature change situation of each monitoring point and perform similarity merging according to the predicted temperature change, dividing M thermal disaster risk control monitoring areas. The monitoring points in each thermal disaster risk control monitoring area have similar temperature change situations, and each thermal disaster risk control monitoring area has different risk levels and monitoring requirements. Through this process, the system terminal can not only monitor various risk control factors of the power station in real time, but also understand the thermal runaway risk situation of each area through the evolutionary analysis of historical data, thus providing data support for the fire prevention and emergency response of the entire lithium-ion energy storage power station.

[0022] In a possible implementation manner, for a lithium-ion energy storage power station, risk control factors are mined to obtain K types of thermal disaster risk control factors. The system includes:

[0023] Using the power station building characteristics of the lithium-ion battery energy storage power station for feature comparison and mining, multiple sample energy storage power stations are obtained; multiple sample thermal runaway accident data sets of the multiple sample energy storage power stations are called through the network, where each sample thermal runaway accident data includes a sample thermal runaway risk level and multiple sample associated risk control factors; the risk control factors of the multiple sample thermal runaway accident data sets are aggregated to obtain H types of sample risk control factors; based on statistical analysis, risk control factor correlation analysis is carried out on the multiple sample thermal runaway accident data sets to obtain H thermal runaway risk coefficients of the H types of sample risk control factors; a thermal runaway risk threshold is preset, and by using the thermal runaway risk threshold to traverse the H thermal runaway risk coefficients, the K types of thermal disaster risk control factors are screened out from the H types of sample risk control factors.

[0024] Specifically, in the power station division module 11, the system terminal obtains the power station building characteristics of the lithium-ion battery energy storage power station. The power station building characteristics refer to the overall structural design, layout configuration, and installation methods of battery packs and related equipment of the energy storage power station, etc., which can be obtained from the design drawings of the lithium-ion battery energy storage power station, such as the layout of energy storage battery packs, heat dissipation paths, heat dissipation materials, battery spacing, channel width, cable layout, etc. Then, obtain the power station building characteristics of other energy storage power stations respectively. By calculating the similarity between the power station building characteristics of the lithium-ion battery energy storage power station and those of other energy storage power stations, the similarity between the lithium-ion battery energy storage power station and other energy storage power stations is obtained, and the energy storage power stations with a similarity greater than or equal to the similarity threshold are recorded to obtain multiple sample energy storage power stations. Among them, the similarity calculation is obtained through Euclidean distance calculation. For example, when calculating the similarity of the energy storage battery pack layout, the coordinates of each energy storage battery pack are obtained from the design drawings. Through these coordinates, the similarity of the energy storage battery pack layout between the lithium-ion battery energy storage power station and each energy storage power station is calculated using Euclidean distance. When calculating the similarity of the heat dissipation materials, the similarity of the heat dissipation materials between the lithium-ion battery energy storage power station and each energy storage power station is calculated using Euclidean distance based on the thermal conductivity of the heat dissipation materials. By calculating the average value of the calculated similarities, the similarity between the lithium-ion battery energy storage power station and each energy storage power station is obtained; Subsequently, the thermal runaway accident data sets of multiple sample energy storage power stations are called through the network. These data sets help to understand the occurrence and severity of thermal runaway events under the influence of different risk control factors by collecting detailed data when historical thermal runaway accidents occur. Each sample thermal runaway accident data includes a sample thermal runaway risk level and multiple sample associated risk control factors. Among them, the sample thermal runaway risk level represents the severity of the thermal runaway event, usually graded from low to high according to risk. The multiple sample associated risk control factors record the data with large fluctuations when the thermal runaway event occurs, such as factors like battery temperature, current, voltage, ambient humidity, charge and discharge rate, etc.; After obtaining multiple sample thermal runaway accident data, the system terminal aggregates these risk control factors, that is, traverses multiple sample thermal runaway accident data sets, stores the data of the same type of risk control factor into a set to obtain multiple risk control factor data sets, and then compares the data volume in each risk control factor data set with the critical data volume, and extracts the risk control factors greater than or equal to the critical data volume. These risk control factors are the factors with large fluctuations that are generally present when the thermal runaway event occurs. By storing these risk control factors separately, H sample risk control factors are obtained;After that, statistical analysis is performed on the obtained H types of sample risk control factors, and correlation analysis is carried out on these factors. The purpose of this step is to obtain the influence degree of each risk control factor on the thermal runaway event by analyzing the relationship between each risk control factor and the thermal runaway risk level. Specifically, the system terminal uses methods such as linear regression or polynomial regression, taking the risk control factors as independent variables and the thermal runaway risk level as the dependent variable, to establish a regression function. Taking linear regression as an example, a linear fitting function is constructed, and the specific form is; , where R is the risk level, , ,..., are risk control factors, , ,..., are thermal runaway risk coefficients, indicating the influence degree of each risk control factor on the risk level. During the fitting process, the corresponding data in multiple sample thermal runaway accident datasets are used as inputs, including multiple sample thermal runaway risk levels and corresponding multiple sample associated risk control factors. To ensure that the fitting function can accurately reflect the influence of each risk control factor on the thermal runaway risk level, the least squares method is used to calculate the thermal runaway risk coefficients of the linear fitting function , ,..., . The goal of the least squares method is to minimize the sum of the squared errors between the predicted values and the actual values. Then, the selected data is substituted into the linear fitting function for fitting, and the parameters in the function are adjusted to make the predicted values after fitting as close as possible to the actual data, so as to obtain the specific coefficients of the linear fitting function. To verify the fitting effect, the mean squared error (MSE) of the fitting function is calculated. If the error is small, the fitting function is valid. At this time, the coefficients of the fitted linear fitting function will be stored as H thermal runaway risk coefficients for H types of sample risk control factors. Otherwise, a polynomial fitting function will be replaced. After obtaining H thermal runaway risk coefficients, the system terminal will preset a thermal runaway risk threshold and compare these risk coefficients by traversing them with the set threshold. The purpose of this step is to screen out those risk control factors that have a significant impact on the thermal runaway risk, and finally obtain K types of thermal disaster risk control factors (all greater than or equal to the preset thermal runaway risk threshold). These screened K types of risk control factors are the most influential factors, and they will be used as the main monitoring parameters for real-time tracking and risk assessment in practical applications. Through this process, combined with historical data, risk control factor analysis, and regression modeling, the most critical risk control factors can be effectively identified, and real-time monitoring and control can be carried out in practical applications, providing accurate risk prediction and prevention and control strategies for lithium-ion energy storage power stations.

[0025] In a possible implementation manner, the multiple historical energy storage scenarios are used to perform thermal runaway propagation evolution on the K groups of thermal disaster risk control monitoring points, and the lithium-ion energy storage power station is divided into the M thermal disaster risk control monitoring areas. The system includes:

[0026] Construct a thermal runaway propagation model by using the power station design information of the lithium-ion battery energy storage power station; input the multiple historical energy storage scenarios into the thermal runaway propagation model for thermal runaway propagation evolution to obtain multiple thermal runaway evolution models; preset a model grid scale to discretize the first thermal runaway evolution model to obtain a first three-dimensional temperature field distribution map; calculate the temperature gradient of the first three-dimensional temperature field distribution map to obtain a first temperature change rate array; use the watershed algorithm to segment the first temperature change rate array to obtain a first sample area segmentation strategy, where the first sample area segmentation strategy includes multiple sample temperature change characteristics of multiple sample monitoring areas; and so on, segment the multiple thermal runaway evolution models to obtain multiple sample area segmentation strategies; perform similarity merging on the multiple sample area segmentation strategies based on the temperature change characteristics to obtain the M thermal disaster risk control monitoring areas.

[0027] Specifically, in the power station division module 11, in order to construct a thermal runaway propagation model for a lithium-ion battery energy storage power station, various types of information on power station design are required, and physical principles such as heat conduction, convection, and radiation are combined to describe the thermal runaway process. First, collect the power station design information of the lithium-ion battery energy storage power station, including battery pack layout, heat dissipation path, cooling system configuration, and the thermal characteristics of the battery itself (such as thermal conductivity and specific heat capacity). At the same time, external factors such as the temperature, humidity, and wind speed of the environment where the battery pack is located should also be considered. These design information provide the basis for constructing the thermal runaway propagation model; based on this information, a physical model of thermal runaway propagation can be constructed. Usually, the thermal runaway propagation process involves heat transfer inside and outside the battery. The heat conduction model is used to describe the heat propagation process inside the battery or between battery packs, and the corresponding mathematical form is the heat conduction equation. The heat exchange process between the battery surface and air or coolant requires the use of a heat convection model, while the radiative heat transfer between the high-temperature battery surface and the external environment is described by the heat radiation model; after selecting these models, initial conditions and boundary conditions need to be set. The initial condition is the temperature state of the battery pack at the start of operation, usually close to the ambient temperature. The boundary conditions describe the heat exchange process between the battery surface and the environment, including the heat convection coefficient and the external environment temperature, etc.; after setting the initial conditions and boundary conditions of the physical model of thermal runaway propagation, the system terminal can obtain a thermal runaway propagation model, which can describe the thermal runaway propagation process in the energy storage power station; once the thermal runaway propagation model is established, the system terminal will input multiple historical energy storage scenarios into the thermal runaway propagation model for simulation. These historical energy storage scenarios contain various factors such as battery states, environmental conditions, charge and discharge rates, etc., which will affect the occurrence and propagation of thermal runaway. By calculating the thermal runaway evolution process under different energy storage scenarios through the thermal runaway propagation model, multiple thermal runaway evolution models are obtained, and each model represents the propagation mode of thermal runaway in different environments; subsequently, a first thermal runaway evolution model is randomly extracted from multiple thermal runaway evolution models. The system terminal will select a numerical method to solve the thermal runaway propagation model. Due to the involvement of complex geometric shapes and multi-physics field coupling, common numerical solution methods include the finite difference method (FDM) and the finite element method (FEM). These methods discretize the model region. The battery pack region is divided into multiple small grids, and the division scale is the preset model grid scale (determined based on business requirements and expert decisions). Each grid represents a spatial unit, and the temperature change of each grid unit is calculated. The discretization in time is completed by selecting an appropriate time step; through numerical solution, the temperature distribution inside and outside the battery pack (the temperature values of each grid) is obtained, and these values are plotted into a three-dimensional temperature field distribution map, showing the temperature changes in each area of the energy storage power station, providing the basis for analyzing the thermal runaway propagation; after that, based on these temperature field distribution maps, the temperature gradient of each area can be further calculated. This temperature gradient is obtained through the gradient operator ( ), which is defined as the rate of change of temperature in each direction, and its mathematical expression is ; where T is the temperature, x, y, and z are spatial coordinates, are the rates of change of temperature in the x, y, and z directions respectively; for the rate of change in the x direction, it can be calculated by the forward difference method, and the forward difference method formula is: ; where T is the temperature, is the temperature at the grid point , is the temperature at the current grid point , $\Delta x$ is the spatial step size of the grid in the x direction, that is, the preset model grid scale. Similarly, the change rates in the y direction and z direction are calculated, and then the final temperature gradient is obtained. After obtaining the temperature gradients of all grids, the system terminal stores these temperature gradients to obtain the first temperature change rate array; further, the watershed algorithm is used to segment the first temperature change rate array. The watershed algorithm is an image processing technique that divides an image into multiple regions by finding the water flow dividing line in the image (i.e., the region where the temperature change is drastic). Specifically, the system terminal uses the temperature change rate array as the input image, and the temperature change rate value will be used as the intensity in the image. The high-gradient region corresponds to the bright area in the image, and the low-gradient region corresponds to the dark area. Then, some representative seed points are selected as water sources. These seed points are usually selected in the regions with larger gradients, that is, the regions where the temperature changes drastically, indicating potential thermal runaway regions; then, the water flow is simulated to expand from these seed points in all directions. Where the temperature gradient changes drastically, the water flow cannot continue to expand, forming a dividing line, and these dividing lines are the boundaries of different regions, thus dividing the entire temperature change rate array into multiple regions, and the temperature change characteristics within each region are similar; after obtaining these segmented regions, the system terminal will further identify and mark these regions as different monitoring regions according to the temperature change characteristics of each region, such as the average temperature change rate, etc. The temperature change characteristics of each monitoring region can provide potential thermal runaway risk information, which is convenient for subsequent risk assessment and prevention and control decision-making. Through the above process, the system terminal obtains the first sample region segmentation strategy, which includes the sample temperature change characteristics of multiple sample monitoring regions, providing an effective basis for subsequent thermal disaster risk control monitoring; after completing the segmentation of the first temperature change rate array, the system terminal uses the same method to perform model discretization, temperature gradient calculation, and array segmentation on the remaining thermal runaway evolution models, so as to obtain the sample region segmentation strategy of each thermal runaway evolution model; finally, the system terminal extracts the labels (average temperature change rate) of each sample monitoring region and calculates the absolute difference between the pairwise labels, and judges whether the calculation result is within the deviation threshold. If it is within the deviation threshold, the system terminal will merge these two sample monitoring regions (including merging the sample temperature change characteristics), and then recalculate the label of the merged sample monitoring region, repeating the above process until the difference between all labels is greater than the deviation threshold, thus obtaining M thermal disaster risk control monitoring regions. These regions represent different risk regions in the lithium-ion battery energy storage power station, and the thermal runaway risks within each region have certain commonalities, providing a basis for subsequent risk assessment and prevention and control strategies.

[0028] A node configuration module is used to perform risk control sensitivity analysis on the M thermal disaster risk control monitoring areas, and configure M edge thermal runaway prediction nodes in the M thermal disaster risk control monitoring areas according to the analysis results. Among them, the M edge thermal runaway prediction nodes are communicatively connected to the thermal disaster alarm decision cloud.

[0029] Specifically, for the M thermal disaster risk control monitoring areas obtained by division, the node configuration module 12 will first use these areas as grouping constraints, and divide each area based on the previously determined K groups of thermal disaster risk control monitoring points to form M thermal disaster risk control monitoring point sets; subsequently, perform risk control sensitivity analysis on these monitoring point sets, analyze the risk characteristics within each area, and then generate M sets of precursor risk characteristics. These sets of precursor risk characteristics are obtained by comprehensively analyzing the monitoring data within the area, and are the potential thermal runaway risk characteristics of each area; based on these analysis results, the system terminal will construct an edge thermal runaway prediction node in each thermal disaster risk control monitoring area. These prediction nodes are configured according to the risk characteristics within the area, aiming to provide the power station with more accurate thermal runaway prevention capabilities through continuous monitoring and early warning; finally, these edge thermal runaway prediction nodes will be communicatively connected to the thermal disaster alarm decision cloud, and will transmit the monitoring data and risk prediction information collected by each of them to the cloud in real time. Through the connection with the cloud, the real-time data of all nodes can be aggregated, comprehensively analyzed and decided, so as to timely identify potential thermal runaway risks and take corresponding prevention and control measures.

[0030] In a possible implementation manner, for performing risk control sensitivity analysis on the M thermal disaster risk control monitoring areas and configuring M edge thermal runaway prediction nodes in the M thermal disaster risk control monitoring areas according to the analysis results, the system includes:

[0031] Taking the M thermal disaster risk control monitoring areas as grouping constraints, dividing the K groups of thermal disaster risk control monitoring points into M thermal disaster risk control monitoring point sets; performing risk control sensitivity analysis on the M thermal disaster risk control monitoring point sets to obtain M sets of precursor risk characteristics; constructing M edge thermal runaway prediction nodes in the M thermal disaster risk control monitoring areas according to the M sets of precursor risk characteristics; pre-constructing a thermal disaster alarm decision cloud, and communicatively connecting the M edge thermal runaway prediction nodes to the thermal disaster alarm decision cloud.

[0032] Specifically, in the node configuration module 12, the system terminal takes M thermal disaster risk control monitoring areas as grouping constraints to divide K groups of thermal disaster risk control monitoring points, that is, the thermal disaster risk control monitoring points belonging to the same area are divided into one set, thus forming M thermal disaster risk control monitoring point sets. This step is to ensure that the risk control monitoring points within each area can be centralized and effectively managed. After division, each thermal disaster risk control monitoring point set represents all the monitoring points within an area; Subsequently, the system terminal conducts risk control sensitivity analysis on these M thermal disaster risk control monitoring point sets. The purpose of this analysis is to evaluate the risk control sensitivity of each monitoring point set and understand the impact of different risk control factors (such as temperature, humidity, charge and discharge rate, etc.) within each area on the thermal runaway risk. The analysis results will generate M precursor risk feature sets, and each feature set contains the key risk features within the area. These features reflect the potential thermal runaway risk within the area; After that, based on the M precursor risk feature sets obtained from the analysis, the system terminal will configure an edge thermal runaway prediction node within each thermal disaster risk control monitoring area. The main function of these edge prediction nodes is to monitor the real-time data within the area according to the risk features obtained previously, and predict and warn of potential thermal runaway risks through the edge thermal runaway prediction channel. Each prediction node will be responsible for the real-time monitoring and data feedback of multiple important parameters such as temperature and charging status in its area; Finally, the system terminal will communicate and connect the M edge thermal runaway prediction nodes with the pre-constructed thermal disaster alarm decision cloud (multiple thermal runaway suppression decision models are pre-constructed and connected in parallel to the thermal disaster alarm decision cloud). Through the cloud, all prediction nodes can aggregate the risk data they collect and conduct comprehensive analysis, so as to make an overall thermal disaster decision. The thermal disaster alarm decision cloud can also send timely warnings to the operation and maintenance center of the power station according to the real-time feedback of the nodes to ensure quick response and take necessary preventive measures.

[0033] In a possible implementation manner, the system includes:

[0034] Using the first thermal disaster risk control monitoring area as a retrieval condition, call multiple first regional thermal runaway data from historical thermal disaster alarm records; extract W groups of thermal runaway monitoring time series data of W thermal disaster risk control monitoring points from the multiple first regional thermal runaway data; perform thermal runaway feature aggregation on the W groups of thermal runaway monitoring time series data to obtain a first precursor risk feature set, where the first precursor risk feature set includes W thermal runaway precursor risk features, and the thermal runaway precursor risk features include the monitoring index change range and the monitoring index change rate; use the W thermal runaway precursor risk features as risk judgment benchmarks to construct W marginal thermal runaway prediction channels; connect the W marginal thermal runaway prediction channels in parallel, and communicatively connect the input ends of the W marginal thermal runaway prediction channels with the W risk control monitoring sensors of the W thermal disaster risk control monitoring points to complete the configuration of the first marginal thermal runaway prediction node; and so on, perform risk control sensitivity analysis on the M thermal disaster risk control monitoring point sets to obtain the M precursor risk feature sets; and so on, construct the M marginal thermal runaway prediction nodes in the M thermal disaster risk control monitoring areas according to the M precursor risk feature sets.

[0035] Specifically, in the node configuration module 12, the system terminal randomly extracts one heat disaster risk control monitoring area from M heat disaster risk control monitoring areas as the first heat disaster risk control monitoring area, and uses this area as the retrieval condition to call the thermal runaway data related to this area from the historical heat disaster alarm records to form multiple first area thermal runaway data. These data contain detailed records of thermal runaway events that occurred in this area in the past, such as parameters like temperature, humidity, battery charge and discharge status, as well as other factors related to thermal runaway. By calling these historical data, the system terminal can understand the past thermal runaway trend in this area, which serves as the basis for subsequent analysis. Subsequently, the system terminal extracts the first heat disaster risk control monitoring point set from the multiple first area thermal runaway data, and extracts W heat disaster risk control monitoring points from the first heat disaster risk control monitoring point set. For these monitoring points, the thermal runaway monitoring time series data of each monitoring point will be extracted to form W groups of thermal runaway monitoring time series data. These thermal runaway monitoring time series data include information such as temperature changes, charging status, and battery health at different time points. These time series data provide the dynamic characteristics of each monitoring point over time, helping the system terminal identify potential thermal runaway risks. After that, the thermal runaway feature aggregation is performed on the W groups of thermal runaway monitoring time series data to obtain the first precursor risk feature set. In this process, the system terminal aggregates each group of thermal runaway monitoring time series data according to K types of heat disaster risk control factors, that is, aggregates the data of the same heat disaster risk control factor in each group of thermal runaway monitoring time series data into a set. For each group of aggregated features, the system terminal extracts the maximum value and the minimum value of this group of aggregated features to obtain the monitoring index change range of this group of aggregated features. Then, the ratio of the difference between the feature value of each time point in each group of aggregated features and the feature value of the previous time point to the time difference is calculated to obtain the feature change rate of each time point. Then, the mean value of the feature change rates of all time points is calculated to obtain the monitoring index change rate of this group of aggregated features. By summarizing the calculated monitoring index change range and monitoring index change rate of each group of aggregated features, the thermal runaway precursor risk feature corresponding to this group of thermal runaway monitoring time series data is obtained. Then, the thermal runaway precursor risk features of the W groups of thermal runaway monitoring time series data are summarized to obtain the first precursor risk feature set. This first precursor risk feature set is the evaluation basis for the potential thermal runaway risk of the first heat disaster risk control monitoring area. Then, the system terminal uses the W thermal runaway precursor risk features as the risk judgment benchmark to construct W edge thermal runaway prediction channels. The function of each prediction channel is to predict the possibility of thermal runaway by analyzing the relationship between the real-time monitoring data and the precursor risk features. Each channel corresponds to an edge prediction node, which specifically monitors the thermal runaway risk of this area and provides early warnings.To ensure that the system terminal can process multiple monitoring points in parallel, the system terminal connects W edge thermal runaway prediction channels in parallel, enabling each monitoring point to independently perform prediction and early warning. At the same time, the input end of each prediction channel is communicatively connected to the W risk control monitoring sensors of the corresponding W thermal disaster risk control monitoring points to obtain the status data of the monitoring points in real time and perform corresponding risk analysis. After this process is completed, the first edge thermal runaway prediction node is configured and has the functions of real-time data collection, analysis, and early warning. Finally, the system terminal continues to perform similar processing on other thermal disaster risk control monitoring point sets, including risk control sensitivity analysis, generating corresponding precursor risk feature sets based on the analysis results, and configuring corresponding edge thermal runaway prediction nodes for each monitoring area based on these precursor risk feature sets. Each node will be responsible for monitoring data such as temperature, humidity, and battery status in its corresponding area and performing real-time thermal runaway risk assessment. Through this series of steps, the system terminal realizes intelligent early warning based on historical data and real-time monitoring data, can take timely measures before the occurrence of thermal runaway risk, thereby improving the safety of the lithium-ion battery energy storage power station.;

[0036] The thermal runaway suppression decision module 13 is used for the thermal disaster alarm decision cloud to receive and make a thermal runaway suppression decision based on the M real-time thermal runaway risk features transmitted back by the M edge thermal runaway prediction nodes, and output a real-time thermal runaway prevention and control strategy, where the real-time thermal runaway prevention and control strategy has a prevention and control time window identifier.

[0037] Specifically, the main task of the thermal disaster alarm decision cloud is to receive M real-time thermal runaway risk characteristics transmitted back from M edge thermal runaway prediction nodes. These characteristics include important parameters such as the temperature change rate, charge and discharge rate, and battery health status. This data provides real-time and accurate risk assessment information for the decision-making of the cloud. When the M edge thermal runaway prediction nodes detect abnormal data, the M edge thermal runaway prediction nodes will transmit the monitored monitoring index parameters and index change rates as real-time thermal runaway risk characteristics back to the thermal disaster alarm decision cloud. Based on the real-time thermal runaway risk characteristics received by the thermal disaster alarm decision cloud, the thermal runaway suppression decision module 13 analyzes and processes the thermal runaway risks in each region through the thermal runaway suppression decision mechanism. The core of this process is to judge the current risk level through the integrated M thermal runaway suppression decision models and generate real-time thermal runaway prevention and control strategies. For example, if the temperature in some regions is too high or the charge and discharge rate is abnormal, the generated real-time thermal runaway prevention and control strategies may be to start the cooling system, reduce the charge and discharge rate, etc. Once a decision is made, the thermal disaster alarm decision cloud will output this real-time thermal runaway prevention and control strategy. This strategy not only includes specific control measures but also comes with a prevention and control time window identifier, which represents the effective time range of the prevention and control measures, ensuring that thermal runaway events do not occur. The design of the prevention and control time window enables the system terminal to dynamically adjust the strategy according to the actual situation and provides accurate guidance for the operation and maintenance personnel to ensure the safety and stability of the lithium-ion battery energy storage power station.

[0038] In a possible implementation manner, the thermal disaster alarm decision cloud receives and makes a thermal runaway suppression decision based on the M real-time thermal runaway risk characteristics transmitted back from the M edge thermal runaway prediction nodes, and outputs a real-time thermal runaway prevention and control strategy. The system includes:

[0039] Using the first thermal disaster risk control monitoring area as the retrieval condition, call multiple first historical area prevention and control strategies of the multiple first area thermal runaway data from the historical thermal disaster alarm records; construct a first thermal runaway suppression decision model using a CNN network; use the multiple first area thermal runaway data and multiple first historical area prevention and control strategies as training data to optimize the parameters of the first thermal runaway suppression decision model; and so on, construct M thermal runaway suppression decision models for the M thermal disaster risk control monitoring areas, and connect the M thermal runaway suppression decision models in parallel to the thermal disaster alarm decision cloud to complete the decision function configuration of the thermal disaster alarm decision cloud; the thermal disaster alarm decision cloud receives and maps the M real-time thermal runaway risk characteristics transmitted back from the M edge thermal runaway prediction nodes into the M thermal runaway suppression decision models for thermal runaway suppression decision, and obtains the real-time thermal runaway prevention and control strategy, where the real-time thermal runaway prevention and control strategy includes M real-time prevention and control sub-strategies.

[0040] Specifically, in the thermal runaway suppression decision module 13, when pre-building the thermal disaster alarm decision cloud, the system terminal uses the first thermal disaster risk control monitoring area as the retrieval condition, and calls the thermal runaway data related to this area from the historical thermal disaster alarm records. These data include the information of thermal runaway events that have occurred in this area in history, as well as the corresponding first historical area prevention and control strategies. These prevention and control strategies include the measures taken to deal with thermal runaway events in the past and the prevention and control time window. By calling these historical data, the system terminal can understand the past thermal runaway situation in this area, which serves as the basis for formulating new prevention and control strategies. Subsequently, the system terminal uses a convolutional neural network (CNN) to build the first thermal runaway suppression decision model. At this time, the system terminal uses the thermal runaway data of multiple first areas and multiple first historical area prevention and control strategies as training data to train the CNN model. Specifically, the system terminal uses CNN to build an initial thermal runaway suppression decision model structure, including an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer. Then, it uses the random initialization method to set the initial weights for each layer of the model, and inputs the training data into this initialized model for forward propagation. The data is passed layer by layer through the input layer, convolutional layer, pooling layer, fully connected layer, and output layer to calculate the prediction result of the thermal runaway risk suppression strategy. After that, the cross-entropy loss function and the mean squared error loss function are used to calculate the loss between the prediction result and the actual result, and the gradient of the loss with respect to the weights of each layer is calculated layer by layer through the backpropagation algorithm. Then, the Adam optimizer is used to optimize the model parameters and adjust the weights of each layer to minimize the value of the loss function. This process will continue until the maximum number of iterations is reached. After training, a validation set (data not used for training) is used to test the performance of the model and evaluate the accuracy of the model in the thermal runaway risk suppression decision task. If the accuracy rate reaches the expected accuracy rate, the currently trained thermal runaway suppression decision model is output as the final model. Otherwise, hyperparameters such as the learning rate and batch size are adjusted to further improve the prediction ability of the model. And so on, M thermal runaway suppression decision models are built for M thermal disaster risk control monitoring areas respectively. Each model is trained and optimized according to the thermal runaway historical data and prevention and control strategies of its respective area, and then all the models are connected in parallel and integrated into the thermal disaster alarm decision cloud to complete the decision function configuration of the thermal disaster alarm decision cloud. When the M edge thermal runaway prediction nodes send back the M real-time thermal runaway risk characteristics to the cloud, the thermal disaster alarm decision cloud will receive these data and input the thermal runaway risk characteristics of each area into the M thermal runaway suppression decision models. Each model will make predictions and decisions based on the input real-time data, thereby generating corresponding real-time thermal runaway prevention and control strategies. This strategy includes M real-time prevention and control sub-strategies, and each sub-strategy corresponds to specific prevention and control measures within different monitoring areas.Finally, these real-time prevention and control sub-strategies will help the power station adopt precise risk response measures, such as adjusting the charge and discharge rate, starting the cooling system, increasing the heat dissipation channels, etc., to ensure that the battery energy storage power station can adopt timely and effective response strategies when facing the risk of thermal runaway, thereby improving the safety and stability of the power station.

[0041] The decision-making generation module 14 is used to perform thermal runaway prevention and control evolution based on the real-time thermal runaway prevention and control strategy, and generate a real-time fire extinguishing risk control decision according to the evolution result.

[0042] Specifically, the decision-making generation module 14 first performs thermal runaway prevention and control evolution on the thermal runaway risk in the battery energy storage power station based on the real-time thermal runaway prevention and control strategy. This process mainly simulates the effects of different prevention and control measures in the future period of time through the thermal runaway propagation model according to the current prevention and control strategy, and adjusts the prevention and control strategy in real time. The goal of the prevention and control evolution is to evaluate in real time whether the prevention and control measures are effective and whether further strengthening of the prevention and control measures is needed based on the monitoring data such as the temperature and charge and discharge rate in the battery energy storage power station. For example, when abnormal temperature is detected in a certain area, the system terminal will first start the basic prevention and control measures (i.e., the real-time thermal runaway prevention and control strategy). If the prevention and control measures do not relieve the risk in time, the system terminal will decide whether to take stronger prevention and control measures according to the evolution result. If the real-time thermal runaway prevention and control strategy can relieve the risk within the prevention and control time window label, the real-time thermal runaway prevention and control strategy will be used as the real-time fire extinguishing risk control decision. Otherwise, the real-time thermal runaway prevention and control strategy and the activation of the explosion suppression and fire extinguishing device will be used as the real-time fire extinguishing risk control decision and marked with the prevention and control time window label to effectively contain the spread of the fire and ensure the safety of the battery energy storage power station. Among them, the explosion suppression and fire extinguishing device has the functions of spraying nitrogen and liquid nitrogen. The response time of the explosion suppression and fire extinguishing device is not higher than 3s. The open fire is extinguished within 5s after the liquid nitrogen is sprayed. Within 10min after the fire is extinguished, the temperature at other positions except the centers on both sides of the battery where thermal runaway occurs is ≤70°C. Within 24h after the fire is extinguished, all batteries no longer experience thermal runaway or combustion. This entire process is dynamic, and the prevention and control evolution is continuously adjusted based on real-time data to ensure that timely and effective response decisions can be made regardless of how complex the thermal runaway risk is, maximizing the safety of equipment and personnel.

[0043] The risk level mapping module 15 is used to traverse the risk level mapping table by using the prevention and control time window label and the real-time fire extinguishing risk control decision to obtain the real-time risk level.

[0044] Specifically, the risk level mapping module 15 first uses the prevention and control time window identifier and the real-time fire extinguishing risk control decision, in combination with the risk level mapping table, to evaluate the current thermal runaway risk level. Specifically, the prevention and control time window identifier represents the effective time range of the current prevention and control strategy, which indicates how long the prevention and control measures should remain effective. The real-time fire extinguishing risk control decision includes the specific prevention and control measures currently taken, such as activating the explosion suppression and fire extinguishing device, adjusting the charge and discharge rate, activating the cooling system, increasing the heat dissipation channels, etc. These decisions directly affect the thermal runaway risk of the current power station. The system terminal traverses the risk level mapping table and inputs the current prevention and control time window identifier and the real-time fire extinguishing risk control decision into the table. The risk level mapping table is a predefined table that shows the relationship between different risk levels and the corresponding prevention and control measures and time windows. It helps the system terminal to obtain the real-time risk level of the power station based on the current prevention and control measures and time window. Finally, the real-time risk level will reflect the current thermal runaway risk level of the power station, ensuring the safety of the lithium-ion battery energy storage power station.

[0045] The alarm signal sending module 16 is used to package the real-time risk level, the real-time thermal runaway prevention and control strategy, and the real-time fire extinguishing risk control decision into a thermal disaster alarm signal and send it to the power station operation and maintenance center.

[0046] Specifically, the alarm signal sending module 16 first integrates the current real-time risk level, the real-time thermal runaway prevention and control strategy, and the real-time fire extinguishing risk control decision together to form a complete thermal disaster alarm signal. The real-time risk level represents the current thermal runaway risk level of the power station, the real-time thermal runaway prevention and control strategy, the specific prevention and control measures based on the current risk assessment, and the real-time fire extinguishing risk control decision are the measures actually taken for prevention and control. After packaging these key information into an alarm signal, the system terminal will send it to the power station operation and maintenance center. After receiving this alarm signal, the operation and maintenance personnel at the power station operation and maintenance center can quickly understand the current safety status of the power station, the prevention and control measures taken, and whether further emergency responses are needed. Through this mechanism, the safety management of the power station can obtain timely feedback and response, ensuring the safe and stable operation of the battery energy storage power station.

[0047] Embodiment 2 Figure 2 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention, showing a block diagram of an exemplary electronic device suitable for implementing the embodiments of the present invention. Figure 2 The displayed electronic device is only an example and should not bring any limitations to the functions and usage scopes of the embodiments of the present invention. As Figure 2 shown, the electronic device includes a processor 21, a memory 22, an input device 23, and an output device 24; the number of processors 21 in the electronic device can be one or more. Figure 2Taking a processor 21 as an example, the processor 21, the memory 22, the input device 23, and the output device 24 in the electronic device can be connected through a bus or other means. Figure 2 Taking the connection through the bus as an example.

[0048] It should be noted that the above-mentioned order of the embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above describes specific embodiments of this specification. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0049] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the protection scope of the present application.

[0050] This specification and the drawings are only exemplary descriptions of the present application and are considered to have covered any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. Thermal disaster alarm system for lithium-ion battery energy storage power station, characterized in that, The system includes: A power station division module, which is used to divide a lithium-ion energy storage power station into M thermal disaster risk control and monitoring areas through the evolution of thermal runaway propagation; A node configuration module, which is used to perform risk control sensitivity analysis on the M thermal disaster risk control and monitoring areas, and configure M edge thermal runaway prediction nodes in the M thermal disaster risk control and monitoring areas according to the analysis results, wherein the M edge thermal runaway prediction nodes are communicatively connected to a thermal disaster alarm decision cloud; A thermal runaway suppression decision module, which is used to receive, by the thermal disaster alarm decision cloud, and perform thermal runaway suppression decision according to M real-time thermal runaway risk characteristics transmitted back by the M edge thermal runaway prediction nodes, and output a real-time thermal runaway prevention and control strategy, wherein the real-time thermal runaway prevention and control strategy has a prevention and control time window identifier; A decision generation module, which is used to perform thermal runaway prevention and control evolution based on the real-time thermal runaway prevention and control strategy, and generate a real-time fire extinguishing risk control decision according to the evolution result; A risk level mapping module, which is used to traverse a risk level mapping table by using the prevention and control time window identifier and the real-time fire extinguishing risk control decision to obtain a real-time risk level; An alarm signal sending module, which is used to package the real-time risk level, the real-time thermal runaway prevention and control strategy, and the real-time fire extinguishing risk control decision into a thermal disaster alarm signal and send it to a power station operation and maintenance center; Through the evolution of thermal runaway propagation, a lithium-ion energy storage power station is divided into M thermal disaster risk control and monitoring areas. The system includes: Excavate risk control factors of a lithium-ion energy storage power station to obtain K types of thermal disaster risk control factors; Locate risk control monitoring points in the lithium-ion energy storage power station according to the K types of thermal disaster risk control factors to generate a power station monitoring plan, wherein the power station monitoring plan includes K groups of thermal disaster risk control monitoring points; Call multiple historical energy storage scenarios of the lithium-ion energy storage power station; Use the multiple historical energy storage scenarios to perform thermal runaway propagation evolution on the K groups of thermal disaster risk control monitoring points, and divide the lithium-ion energy storage power station into the M thermal disaster risk control and monitoring areas; Perform risk control sensitivity analysis on the M thermal disaster risk control and monitoring areas, and configure M edge thermal runaway prediction nodes in the M thermal disaster risk control and monitoring areas according to the analysis results. The system includes: Taking the M thermal disaster risk control and monitoring areas as grouping constraints, divide the K groups of thermal disaster risk control monitoring points into M thermal disaster risk control monitoring point sets; Perform risk control sensitivity analysis on the M thermal disaster risk control monitoring point sets to obtain M sets of precursor risk characteristics; Construct M edge thermal runaway prediction nodes in the M thermal disaster risk control and monitoring areas according to the M sets of precursor risk characteristics; Pre-construct a thermal disaster alarm decision cloud, and communicatively connect the M edge thermal runaway prediction nodes to the thermal disaster alarm decision cloud.

2. The thermal disaster alarm system for a lithium-ion battery energy storage power station according to claim 1, characterized in that, Excavate risk control factors of a lithium-ion energy storage power station to obtain K types of thermal disaster risk control factors. The system includes: Use the power station building characteristics of the lithium-ion battery energy storage power station for feature comparison and excavation to obtain multiple sample energy storage power stations; Network and call multiple sample thermal runaway accident data sets of the multiple sample energy storage power stations, wherein each sample thermal runaway accident data includes a sample thermal runaway risk level and multiple sample associated risk control factors; Aggregate risk control factors for the multiple sample thermal runaway accident datasets to obtain H sample risk control factors; Conduct risk control factor correlation analysis on the multiple sample thermal runaway accident datasets based on statistical analysis to obtain H thermal runaway risk coefficients for the H sample risk control factors; Preset a thermal runaway risk threshold, and traverse the H thermal runaway risk coefficients by using the thermal runaway risk threshold to screen out the K thermal disaster risk control factors from the H sample risk control factors.

3. The thermal disaster alarm system for a lithium-ion battery energy storage power station according to claim 1, characterized in that, Use the multiple historical energy storage scenarios to perform thermal runaway propagation and evolution on the K groups of thermal disaster risk control monitoring points, and divide the lithium-ion energy storage power station into the M thermal disaster risk control monitoring areas. The system includes: Construct a thermal runaway propagation model by using the power station design information of the lithium-ion battery energy storage power station; Input the multiple historical energy storage scenarios into the thermal runaway propagation model for thermal runaway propagation and evolution to obtain multiple thermal runaway evolution models; Preset a grid scale discretization of the first thermal runaway evolution model to obtain a first three-dimensional temperature field distribution map; Calculate the temperature gradient of the first three-dimensional temperature field distribution map to obtain a first temperature change rate array; Use the watershed algorithm to segment the first temperature change rate array to obtain a first sample area segmentation strategy, where the first sample area segmentation strategy includes multiple sample temperature change characteristics of multiple sample monitoring areas; And so on, segment the multiple thermal runaway evolution models to obtain multiple sample area segmentation strategies; Based on the temperature change characteristics, perform similarity merging on the multiple sample area segmentation strategies to obtain the M thermal disaster risk control monitoring areas.

4. The thermal disaster alarm system for a lithium-ion battery energy storage power station according to claim 3, characterized in that, The system includes: Use the first thermal disaster risk control monitoring area as the retrieval condition to call multiple first area thermal runaway data from the historical thermal disaster alarm records; Extract W groups of thermal runaway monitoring time series data of W thermal disaster risk control monitoring points in the first thermal disaster risk control monitoring point set from the multiple first area thermal runaway data; Perform thermal runaway feature aggregation on the W groups of thermal runaway monitoring time series data to obtain a first precursor risk feature set, where the first precursor risk feature set includes W thermal runaway precursor risk features, and the thermal runaway precursor risk features include the monitoring index change range and the monitoring index change rate; Use the W thermal runaway precursor risk features as the risk judgment benchmark to construct W marginal thermal runaway prediction channels; Connect the W marginal thermal runaway prediction channels in parallel, and communicatively connect the input ends of the W marginal thermal runaway prediction channels with the W risk control monitoring sensors of the W thermal disaster risk control monitoring points to complete the configuration of the first marginal thermal runaway prediction node; And so on, perform risk control sensitivity analysis on the M thermal disaster risk control monitoring point sets to obtain the M precursor risk feature sets; And so on, construct the M marginal thermal runaway prediction nodes in the M thermal disaster risk control monitoring areas according to the M precursor risk feature sets.

5. The thermal disaster alarm system for a lithium-ion battery energy storage power station according to claim 4, characterized in that, The thermal disaster alarm decision cloud receives and makes a thermal runaway suppression decision according to the M real-time thermal runaway risk features transmitted back by the M marginal thermal runaway prediction nodes, and outputs a real-time thermal runaway prevention and control strategy. The system includes: Using the first thermal disaster risk control monitoring area as a retrieval condition, retrieve multiple first historical area prevention and control strategies of the multiple first area thermal runaway data from the historical thermal disaster alarm records; Construct a first thermal runaway suppression decision model using a CNN network; Use the multiple first area thermal runaway data and multiple first historical area prevention and control strategies as training data to adjust and optimize the parameters of the first thermal runaway suppression decision model; And so on, construct M thermal runaway suppression decision models for the M thermal disaster risk control monitoring areas, and connect the M thermal runaway suppression decision models in parallel to the thermal disaster alarm decision cloud to complete the decision function configuration of the thermal disaster alarm decision cloud; The thermal disaster alarm decision cloud receives and inputs the M real-time thermal runaway risk feature maps transmitted back by the M edge thermal runaway prediction nodes into the M thermal runaway suppression decision models for thermal runaway suppression decision, and obtains the real-time thermal runaway prevention and control strategy, where the real-time thermal runaway prevention and control strategy includes M real-time prevention and control sub-strategies.

6. An electronic device, characterized in that, The electronic device includes: A memory for storing executable instructions; A processor for deploying the thermal disaster alarm system of the lithium-ion battery energy storage power station according to any one of claims 1 to 5 when executing the executable instructions stored in the memory.

Images