Standard modular energy storage thermal runaway early warning method and system based on digital twinning

Abstract

The invention discloses a standard modular energy storage thermal runaway early warning method and system based on digital twinning, and relates to the technical field of energy storage system safety monitoring. The method comprises the following steps: deploying multiple types of sensors in an energy storage unit to form a network, and collecting multi-dimensional data such as temperature, voltage and gas concentration; transmitting the data to the digital twin platform, and performing format conversion on the data for thermal runaway prediction; the platform is combined with the LSTM-GBDT hybrid model through physical modeling, and the risk probability is output; and comparing the risk threshold to trigger early warning. The system comprises a sensing module, a data acquisition and transmission module, a digital twinborn prediction module and an early warning execution module. According to the method, physical simulation and historical data training are fused through the digital twin platform, the heat transfer and out-of-control propagation process can be accurately simulated, virtual simulation optimization is supported, the early warning accuracy and reliability are improved, and the operation and maintenance burden and accident potential are reduced.

Description

Technical Field

[0001] This invention relates to the field of energy storage system safety monitoring technology, specifically to a standard modular energy storage thermal runaway early warning method and system based on digital twins. Background Technology

[0002] With the rapid development of the new energy industry, energy storage systems, as core equipment for mitigating fluctuations in new energy power generation and ensuring stable grid operation, are seeing continuous expansion in application scale. Among them, standard modular energy storage systems have become one of the mainstream technical solutions in the field of large-scale energy storage due to their advantages such as high degree of structural standardization, flexible expansion, and convenient operation and maintenance.

[0003] However, current early warning methods for thermal runaway in energy storage typically rely on data collected by a single type of sensor, reflecting only a local or one-dimensional operational status of the energy storage unit. This makes it difficult to comprehensively capture the multi-dimensional precursory characteristics before thermal runaway occurs. Furthermore, traditional early warning models are built upon empirical thresholds or simple physical laws, lacking accurate simulations of the complex heat transfer, electrochemical reactions, and thermal runaway propagation processes within the energy storage unit. This hinders early identification and quantitative assessment of thermal runaway risks, leading to decreased early warning accuracy. This not only increases the monitoring burden on maintenance personnel and reduces system safety but also risks potential accidents due to delayed or false alarms, affecting the stable operation of the energy storage system. Summary of the Invention

[0004] This invention addresses the technical problems of traditional energy storage thermal runaway early warning methods that rely on isolated parameters collected by a single type of sensor, making it difficult to comprehensively capture the multi-dimensional precursor features of thermal runaway. Furthermore, the early warning models lack accurate simulation of the complex heat transfer, electrochemical reactions, and thermal runaway propagation processes within the energy storage unit, resulting in the inability to achieve early identification and quantitative assessment of thermal runaway risks. Consequently, these methods lead to low early warning accuracy, increased maintenance burden, and significant safety hazards. This invention provides a standard modular energy storage thermal runaway early warning method and system based on digital twins.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: In a first aspect, the present invention provides a standard modular energy storage thermal runaway early warning method based on digital twins, including: Multiple types of sensors are deployed in standard modular energy storage units to form a sensor network; Based on the sensor network, the standard modular energy storage unit is used to collect operational data to obtain multi-dimensional sensing data; The multidimensional sensor data is transmitted to the digital twin platform to predict thermal runaway of the standard modular energy storage unit and obtain the probability of thermal runaway risk. Based on the aforementioned probability of thermal runaway risk, an early warning system for thermal runaway in energy storage is developed.

[0006] Secondly, this invention provides a standard modular energy storage thermal runaway early warning system based on digital twins, comprising: The sensing module is used to deploy multiple types of sensors in a standard modular energy storage unit to form a sensor network; The data acquisition and transmission module is used to acquire operational data of the standard modular energy storage unit based on the sensor network, obtain multi-dimensional sensing data, and transmit the multi-dimensional sensing data to the digital twin platform. The digital twin prediction module is used to predict thermal runaway of the standard modular energy storage unit based on the multidimensional sensing data, and obtain the probability of thermal runaway risk. The early warning execution module is used to provide early warning of thermal runaway in energy storage based on the probability of thermal runaway risk.

[0007] The beneficial effects of this invention are: Compared to existing technologies, this application firstly collects multi-dimensional sensing data by deploying multiple types of sensors, breaking through the limitations of single sensors and achieving comprehensive monitoring of the system's operating status; secondly, it relies on a digital twin platform to construct a real-time synchronized virtual model, accurately simulating heat transfer, electrochemical reactions, and thermal runaway propagation processes, providing a precise virtual foundation for risk assessment; thirdly, it uses a hybrid prediction method combining machine learning and physical models to achieve quantitative assessment of thermal runaway risk and early feature identification, overcoming the shortcomings of traditional empirical threshold methods; and finally, it supports virtual simulation optimization and emergency response formulation, improving the accuracy and reliability of early warnings, reducing the impact of accidents, and enhancing the safety of energy storage systems. Attached Figure Description

[0008] Figure 1 A flowchart illustrating the standard modular energy storage thermal runaway early warning method based on digital twins provided by this invention; Figure 2 A schematic diagram of the structure of the standard modular energy storage thermal runaway early warning system based on digital twin provided by the present invention.

[0009] In the attached diagram, the components represented by each number are as follows: The system includes a sensing module 11, a data acquisition and transmission module 12, a digital twin prediction module 13, and an early warning execution module 14. Detailed Implementation

[0010] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0011] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0012] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0013] Example 1, as Figure 1 As shown, this embodiment of the invention provides a standard modular energy storage thermal runaway early warning method based on digital twins, including: S10: Deploy multiple types of sensors in standard modular energy storage units to form a sensor network; Specifically, multiple types of sensors are deployed in standard modular energy storage units to form a sensor network, including: Obtain the unit model of the standard modular energy storage unit, and perform sensor matching in the sensor configuration library based on the unit model to determine multiple types of sensors; Sensors are configured in the standard modular energy storage unit according to multiple types of sensors to form a sensor network.

[0014] Standard modular energy storage units refer to basic functional units of energy storage that adopt standardized design and have unified interfaces and structural specifications. They are typically composed of a certain number of battery cells, heat dissipation structures, electrical connectors, etc., and can be combined and expanded to form large-scale energy storage systems. First, the unit model of the standard modular energy storage unit is obtained by reading the nameplate of the energy storage unit or retrieving it from the energy storage system management platform. The unit model is the core identifier that distinguishes the specifications of the energy storage unit, and includes key parameters such as capacity, battery type, and structural dimensions. It directly determines its thermal runaway risk characteristics. For example, high-capacity modules have a higher risk of thermal accumulation, so sensors need to be matched based on the model.

[0015] Secondly, the acquired unit model is input into the sensor configuration library retrieval interface. The system automatically matches the corresponding list of multiple sensor types for that model, including: temperature data collected by temperature sensors, including battery surface temperature, internal temperature, and ambient temperature; real-time current data collected by current sensors; real-time voltage data collected by voltage sensors; gas concentration data collected by gas concentration sensors, including hydrogen concentration, carbon monoxide concentration, and organic electrolyte gas concentration; and ambient humidity data collected by humidity sensors. The sensor configuration library is based on historical experimental data and failure cases, summarizing the thermal runaway risk points of different models of energy storage units, such as weak heat dissipation areas of large-capacity modules and high-incidence areas of electrical faults in high-rate charge and discharge modules. These risk points are bound to sensor types, forming a pre-established database of standardized configuration rules that stores the matching relationships between different models of standard modular energy storage units and their corresponding sensors.

[0016] Finally, prepare the sensors according to the matching list, deploy them at the installation locations specified in the configuration library, and connect the sensors to the data acquisition terminal via wires or wireless modules to complete the network setup, forming a sensor network for centralized data acquisition and transmission. Installation locations must cover risk-sensitive areas of the energy storage unit, such as the core heat-generating area of ​​the battery, weak electrical connections, and areas prone to gas accumulation, to ensure data representativeness. Deploying multiple types of sensor networks in standard modular energy storage units enables comprehensive real-time monitoring of key parameters such as temperature, voltage, current, and gas concentration, providing a rich data foundation for thermal runaway early warning.

[0017] S20: Based on the sensor network, collect operational data of the standard modular energy storage unit to obtain multi-dimensional sensing data; Specifically, based on the sensor network, operational data of the standard modular energy storage unit is collected to obtain multi-dimensional sensing data, including: Data is collected from the standard modular energy storage unit through the sensor network at a preset acquisition frequency to obtain multiple types of sensor data, and the multiple types of sensor data are entered into the energy storage unit operation database. Receive thermal runaway test command, and retrieve the data analysis window based on the thermal runaway test command; The multidimensional sensing data is obtained by extracting data from the energy storage unit's operating database using the data analysis window.

[0018] After the sensor network is established, operational data can be collected from the standard modular energy storage unit based on the sensor network. First, data is collected according to a preset acquisition frequency. The preset acquisition frequency refers to the number of data acquisitions per unit time set for different types of sensors, that is, the time interval parameter for the sensor to convert physical quantities into digital signals, such as 1 time / second. Since different physical quantities change at different rates, such as temperature changing slowly and voltage fluctuations being transient, preset differentiated frequencies can reduce the system's data transmission and storage load while ensuring the real-time performance of key data, thus avoiding resource waste.

[0019] Secondly, various types of sensor data are entered into the energy storage unit operation database. These various types of sensor data refer to the raw data collection gathered through sensor networks, reflecting different dimensions of the energy storage unit's status. The energy storage unit operation database is a structured database used to store real-time operational data of the energy storage unit, including data storage tables, indexing rules, and access interfaces. Standardizing the storage of various types of sensor data ensures data traceability and retrieval. Simultaneously, time-series indexing supports rapid data extraction along the time dimension, providing a foundation for subsequent data analysis.

[0020] Furthermore, upon receiving a thermal runaway verification command, the data analysis window is retrieved. The thermal runaway verification command is a signal that triggers the data extraction process, containing information such as the command ID, target energy storage unit ID, and data analysis window parameters. Data extraction is driven by the command, enabling on-demand analysis and avoiding the inefficiency caused by indiscriminate full data processing. The data analysis window refers to the time interval parameters used for data extraction, including the window duration (time span) and the sliding step size (time interval between adjacent windows), used to focus on key data segments in the early stages of thermal runaway. Since the precursory characteristics of thermal runaway typically evolve continuously over a certain period, such as historical data showing coordinated anomalies in temperature and gas concentration 5 minutes before thermal runaway, setting the window duration can completely capture trend changes; the sliding step size ensures continuous data coverage, avoiding missing key nodes. After receiving the thermal runaway verification command, the window parameters in the command are parsed, and the window start time is calculated in conjunction with the current timestamp.

[0021] Finally, multidimensional sensing data is extracted from the data analysis window. Based on the window time range, valid data from all sensors within the interval are retrieved through the database index and integrated into multidimensional sensing data according to time order and sensor type.

[0022] S30: Transmit the multi-dimensional sensing data to the digital twin platform to perform thermal runaway prediction on the standard modular energy storage unit and obtain the thermal runaway risk probability; Specifically, the multi-dimensional sensor data is transmitted to a digital twin platform to predict thermal runaway of the standard modular energy storage unit, obtaining the probability of thermal runaway risk, including: Obtain the input mapping format of the digital twin platform; The multidimensional sensing data is converted based on the input mapping format to obtain formatted sensing data that conforms to the input of the digital twin platform; The digital twin platform performs thermal runaway prediction based on the formatted sensor data to obtain the thermal runaway risk probability.

[0023] The construction of the digital twin platform includes: Obtain the physical structure information, electrical parameter information, and thermophysical parameters of the standard modular energy storage unit; Based on the physical structure information, the electrical parameter information, and the thermophysical parameters, an initial digital twin model of the standard modular energy storage unit is constructed. Obtain a historical sample dataset, which includes a multidimensional sensor dataset of samples and a probability set of thermal runaway risks of samples; Using the sample multidimensional sensing dataset as input and the sample thermal runaway risk probability set as supervision signal, the initial digital twin model is trained to obtain the digital twin platform.

[0024] Specifically, a digital twin platform refers to a virtual platform that is mapped in real time to a physical energy storage unit. It includes core components such as a virtual twin model, a data processing engine, and a prediction algorithm module. By integrating real-time data with physical laws, it can simulate the operating status of the energy storage unit and predict risks.

[0025] Building a digital twin platform first requires acquiring the physical structure, electrical parameters, and thermophysical parameters of standard modular energy storage units. Physical structure information refers to the geometric dimensions, structural composition, and spatial layout data of the standard modular energy storage units, including geometric parameters (length, width, height, volume, etc.), battery module arrangement and connection methods, thermal management system configuration (heat sinks, fans, liquid cooling pipes, etc.), and electrical connection topology. Electrical parameters are key parameters reflecting the electrical characteristics of the energy storage unit, including the rated capacity, rated voltage, internal resistance, charge / discharge characteristic curves and SOC-OCV relationship of individual battery cells, series and parallel configuration parameters of battery modules, and parameter settings of electrical protection devices. Thermophysical parameters describe the heat transfer and thermodynamic properties of materials, including the thermal conductivity, specific heat capacity, density, battery heat generation model parameters (ohmic heat, polarization heat, etc.), heat transfer boundary conditions and environmental parameters, and thermal runaway reaction kinetic parameters.

[0026] Physical structural information determines the spatial form of the virtual model, ensuring complete consistency between the model and the physical units' geometry and component positions, providing a spatial framework for subsequent simulations of heat transfer paths and electrical connections. Electrical parameters are the core inputs for simulating the electrochemical reactions and dynamic changes in voltage / current of the energy storage unit, directly affecting the model's simulation accuracy regarding the correlation between electrical anomalies and increased heat generation. Thermophysical parameters characterize the heat transfer patterns. In summary, physical, electrical, and thermophysical parameters provide data support for achieving accurate mapping between the virtual model and the physical energy storage units.

[0027] Furthermore, based on physical structure information, electrical parameter information, and thermophysical parameters, an initial digital twin model of a standard modular energy storage unit is constructed. Specifically, firstly, a 3D modeling tool is used to construct a virtual geometric model at a 1:1 scale with the physical unit, restoring the spatial positions of the battery pack, heat dissipation structure, and electrical connectors to ensure the morphological realism of the virtual model. Secondly, based on the heat transfer process, electrochemical reaction, and thermal runaway propagation, heat conduction equations are embedded in the geometric model to simulate the heat transfer process inside the battery caused by electrochemical reactions and resistance losses. Electrical parameters are converted into mathematical equations to simulate the heat generation and voltage fluctuations of electrochemical reactions during charging and discharging. Thermal runaway triggering conditions are preset to simulate the process of thermal runaway spreading from a single battery cell to the surrounding area.

[0028] Physical modeling, based on fundamental laws of thermodynamics and electrochemistry, transforms the physical structure, electrical parameters, and thermophysical parameters of standard modular energy storage units into calculable mathematical equations and simulation models. This enables the virtual digital model to accurately replicate the geometry and operating mechanism of the physical energy storage unit, ensuring consistency between the virtual model and the physical unit. This provides underlying support for the real-time mapping characteristics of digital twins. Furthermore, the simulation logic constructed by physical modeling provides physical constraints for machine learning algorithms, avoiding data correlation deviations that may occur in purely data-driven models, which deviate from basic physical principles.

[0029] Furthermore, a historical sample dataset is obtained, which includes a multidimensional sensor dataset of samples and a probability set of thermal runaway risks for samples, including: Based on the physical structure information, the electrical parameter information, and the thermophysical parameters, physical structure features, electrical parameter features, and thermophysical features are constructed. Energy storage units are searched using the physical structural features, electrical parameter features, and thermophysical features as search constraints to identify multiple similar energy storage units. By retrieving historical operation records from multiple similar energy storage units, a set of historical operation records is obtained; Obtain a data analysis window, and sample each historical operation record in the historical operation record set according to the data analysis window to obtain a sample multidimensional sensor dataset; The thermal runaway risk probability set of the sample multidimensional sensor dataset is labeled with the thermal runaway risk probability. The historical sample dataset is obtained by summarizing the sample multidimensional sensing dataset and the sample thermal runaway risk probability set.

[0030] First, the physical structure, electrical parameters, and thermophysical parameters of the target energy storage unit are structured, key features are extracted and standardized, and physical structure, electrical parameters, and thermophysical features are constructed. Second, the constructed three types of features are input into the retrieval interface of the energy storage unit database. Through feature similarity algorithms, such as cosine similarity and Euclidean distance, the matching degree with units in the database is calculated, and units with a matching degree higher than a preset threshold, such as 85%, are selected to form a list of similar units, thus identifying multiple similar energy storage units. Third, the historical operation record set is retrieved. Based on the list of similar units, the full operation record of the corresponding unit is retrieved from the historical database of the energy storage system, including sensor time-series data and operating conditions. Logs, fault reports, etc., are categorized and stored by unit to form a historical operation record set. Next, a data analysis window is obtained, and each historical operation record in the historical operation record set is sampled according to the data analysis window to obtain a sample multidimensional sensor dataset. Specifically, the data analysis window parameters in step S20 are used to perform sliding sampling on the time-series data in the historical operation record set. Further, risk labeling is performed on each sample multidimensional sensor data segment, combining historical fault records and expert assessments. Finally, the sample multidimensional sensor dataset and the corresponding sample thermal runaway risk probability set are associated one-to-one with each sample and label to form a structured historical sample dataset for subsequent training of the digital twin model.

[0031] Furthermore, a hybrid prediction model is constructed using multidimensional sensor data as input and the probability of thermal runaway risk as output. The hybrid model combining LSTM (Long Short-Term Memory) and GBDT (Gradient Boosting Tree) is a machine learning model that integrates temporal feature capture and multidimensional risk quantification. It is used to process multidimensional sensor time-series data of energy storage systems, and is particularly adept at capturing long-term dependencies in the thermal runaway evolution process. Because thermal runaway characteristics are continuous, current conditions such as temperature and gas concentration affect future risks, and these influencing factors change dynamically over time, such as increasing temperature gradients and accumulating gas concentrations, which may exhibit long-term dependencies, such as the susceptibility to thermal runaway due to battery aging. For example, choosing LSTM to capture temporal correlations and combining it with GBDT to quantify risk provides a more accurate prediction of the probability of thermal runaway risk than traditional single models.

[0032] Specifically, the core components of the hybrid model include the temporal processing unit of LSTM and the risk quantization unit of GBDT. The memory cells of LSTM refer to the information channels running through the network, used to store key long-term information from the thermal runaway time-series data, such as the continuous temperature rise trend. As information flows through the cells, it is selectively modified by the gating units. The LSTM gating units include: a forget gate, used to decide which irrelevant historical information in the "memory cells," such as instantaneous temperature fluctuations, needs to be discarded; an input gate, used to decide which new information needs to be stored in the "memory cells"; and an output gate, used to decide which temporal features, such as the co-current trend of temperature change rate and gas concentration, should be output to the next layer based on the current state of the "memory cells." The GBDT unit is an integration of multiple decision trees, capturing the nonlinear mapping relationship between multi-dimensional sensing features and risk probabilities through feature splitting.

[0033] Hybrid models in practical applications typically consist of a combination of multi-layer networks and ensemble trees. The structure includes: an input layer that receives multi-dimensional sensor time-series data, such as time series data for temperature, voltage, gas concentration, and pressure; an LSTM layer, the core time-series processing layer, containing multiple LSTM units, each processing time-series information through a gating mechanism to capture the temporal dependencies between features, such as the correlation between a 2°C / min temperature increase over the past 30 minutes and an increase in risk over the next 10 minutes; stacking multiple LSTM layers enhances dynamic feature extraction capabilities; a fully connected layer that converts the high-dimensional time-series features output by the LSTM layer into low-dimensional features, integrating key information such as temperature gradient, voltage fluctuation standard deviation, and gas concentration growth rate; a GBDT layer that, based on the features output by the fully connected layer, quantifies the risk probability through weighted voting by multiple decision trees, capturing the non-linear relationship between features and risk; and an output layer, typically a single neuron, that outputs the prediction result, such as the probability of thermal runaway risk over the next 10 minutes.

[0034] Specifically, multi-dimensional sensing parameters, comprising temperature time-series characteristics, voltage fluctuation characteristics, gas concentration characteristics, and pressure change characteristics, are input into a hybrid model. Corresponding actual thermal runaway risk probability labels are extracted. The sample set is divided into a training set (7:3) for model learning and a validation set for monitoring overfitting. The model uses the training set to predict risk probabilities, calculates the difference between the predicted and actual risk probabilities using a loss function, and then updates the weight parameters of the LSTM and fully connected layers, as well as the tree structure parameters of the GBDT, through backpropagation. Simultaneously, the validation set is used to evaluate model performance, iteratively optimizing until the validation set error stabilizes and reaches a warning accuracy of over 90%, ultimately forming a model capable of accurately predicting the probability of thermal runaway risk.

[0035] Finally, the real-time collected multi-dimensional sensor time-series parameters are input into the model. The model learns the mapping relationship between time-series characteristics and risk probabilities based on historical data, calculates and outputs the probability of thermal runaway risk for a preset duration in the future, and provides a quantitative basis for early warning decision-making.

[0036] In summary, the construction process of the digital twin platform includes: first, acquiring the physical structure information, electrical parameter information, and thermophysical parameters of the standard modular energy storage unit, and constructing an initial digital twin model based on the parameters; second, acquiring a historical sample dataset containing a multidimensional sensor dataset and a sample thermal runaway risk probability set; and finally, training the initial digital twin model using the multidimensional sensor dataset as input and the sample thermal runaway risk probability set as a supervision signal.

[0037] The digital twin platform can receive multi-dimensional sensor data after format conversion and predict thermal runaway for standard modular energy storage units based on the formatted sensor data, obtaining the probability of thermal runaway risk. Simultaneously, the platform supports virtual simulation and optimization functions. By simulating different operating conditions and fault scenarios, it optimizes sensor layout and early warning thresholds, further improving the accuracy and reliability of the early warning system. Maintenance personnel can use the platform to view the energy storage system's operating status in real time, receive early warning information, and formulate emergency measures based on the virtual model analysis results, such as remotely cutting off power, activating the cooling system, or isolating faulty modules, minimizing the impact of thermal runaway accidents.

[0038] S40: Based on the aforementioned thermal runaway risk probability, conduct an early warning of thermal runaway in energy storage.

[0039] Specifically, based on the aforementioned thermal runaway risk probability, early warning of thermal runaway in energy storage is provided, including: Set a risk probability threshold; The probability of thermal runaway risk is compared with the risk probability threshold. When the probability of thermal runaway risk is greater than or equal to the risk probability threshold, an early warning of thermal runaway in energy storage is issued.

[0040] The risk probability threshold refers to the pre-set critical value for triggering a thermal runaway warning, expressed as a percentage (%). It is a quantitative standard for determining whether to issue a warning. For example, setting the threshold to 60% means that a warning is triggered when the risk probability is ≥60%. The threshold transforms the abstract risk probability into a clear basis for warning decisions, avoiding subjective judgment bias and ensuring the consistency and reliability of warnings.

[0041] Furthermore, the probability of thermal runaway risk is compared with a risk probability threshold. When the probability of thermal runaway risk is greater than or equal to the risk probability threshold, an early warning for thermal runaway in energy storage is issued. The probability of thermal runaway risk refers to the likelihood, expressed as a percentage (%), of a thermal runaway event occurring within a specific timeframe, such as 10 minutes, as predicted by the digital twin platform. Comparing the probability of thermal runaway risk with the risk probability threshold, if the probability is less than the threshold, the system does not trigger an early warning but only continuously monitors for changes in risk. If the probability is greater than or equal to the threshold, the early warning mechanism is immediately activated, including sending early warning information to the operation and maintenance platform, triggering on-site audible and visual alarms, and recording early warning logs. In summary, the embodiments of this application have at least the following technical effects: Compared to existing technologies, this application firstly establishes a sensing network by deploying multiple types of sensors within a standard modular energy storage unit. This network collects multi-dimensional sensor data such as temperature, pressure, gas concentration, and electrical parameters, overcoming the limitations of traditional methods that rely on single sensor data. This enables comprehensive monitoring of the operating status and avoids biased early warnings caused by partial data. Secondly, it utilizes a digital twin platform to construct a virtual model synchronized in real time with the physical unit. This model integrates physical modeling and data-driven methods to accurately simulate heat transfer, electrochemical reactions, and thermal runaway propagation processes, solving the problem that traditional models struggle to replicate complex thermal evolution processes and providing a precise virtual mapping foundation for risk assessment. Thirdly, it employs a hybrid prediction method combining machine learning and physical models. Based on historical data learning and real-time data dynamic analysis, it achieves quantitative assessment and early feature identification of thermal runaway risk, overcoming the shortcomings of traditional experience-based threshold methods in quantifying risk and providing early warnings. Finally, it supports virtual simulation to optimize sensor layout and early warning thresholds and assists maintenance personnel in developing emergency measures, effectively improving the accuracy and reliability of early warnings, reducing the impact of thermal runaway accidents, and comprehensively enhancing the safety of standard modular energy storage systems.

[0042] Example 2, as Figure 2 As shown, based on the same inventive concept as the standard modular energy storage thermal runaway early warning method based on digital twin provided in Embodiment 1, this embodiment of the invention also provides a standard modular energy storage thermal runaway early warning system based on digital twin, including: The sensing module 11 is used to deploy multiple types of sensors in a standard modular energy storage unit to form a sensor network; The data acquisition and transmission module 12 is used to acquire operational data of the standard modular energy storage unit based on the sensor network, obtain multi-dimensional sensing data, and transmit the multi-dimensional sensing data to the digital twin platform. The digital twin prediction module 13 is used to predict thermal runaway of the standard modular energy storage unit based on the multidimensional sensing data, and obtain the probability of thermal runaway risk. The early warning execution module 14 is used to provide early warning of thermal runaway in energy storage based on the probability of thermal runaway risk.

[0043] Specifically, the sensing module 11 is used for: Deploying multiple types of sensors within standard modular energy storage units to form a sensor network, including: Obtain the unit model of the standard modular energy storage unit, and perform sensor matching in the sensor configuration library based on the unit model to determine multiple types of sensors; Sensors are configured in the standard modular energy storage unit according to multiple types of sensors to form a sensor network.

[0044] The data acquisition and transmission module 12 is specifically used for: Based on the sensor network, operational data of the standard modular energy storage unit is collected to obtain multi-dimensional sensing data, including: Data is collected from the standard modular energy storage unit through the sensor network at a preset acquisition frequency to obtain multiple types of sensor data, and the multiple types of sensor data are entered into the energy storage unit operation database. Receive a thermal runaway test command and retrieve the data analysis window based on the thermal runaway test command; The data is extracted from the energy storage unit's operating database using the data analysis window to obtain the multidimensional sensing data.

[0045] Specifically, the digital twin prediction module 13 is used for: The multidimensional sensor data is transmitted to a digital twin platform to predict thermal runaway of the standard modular energy storage unit, obtaining the probability of thermal runaway risk, including: Obtain the input mapping format of the digital twin platform; The multidimensional sensing data is converted based on the input mapping format to obtain formatted sensing data that conforms to the input of the digital twin platform; The digital twin platform performs thermal runaway prediction based on the formatted sensor data to obtain the thermal runaway risk probability.

[0046] Specifically, constructing the digital twin platform includes: Obtain the physical structure information, electrical parameter information, and thermophysical parameters of the standard modular energy storage unit; Based on the physical structure information, the electrical parameter information, and the thermophysical parameters, an initial digital twin model of the standard modular energy storage unit is constructed. Obtain a historical sample dataset, which includes a multidimensional sensor dataset of samples and a probability set of thermal runaway risks of samples; Using the sample multidimensional sensing dataset as input and the sample thermal runaway risk probability set as supervision signal, the initial digital twin model is trained to obtain the digital twin platform.

[0047] This includes acquiring a historical sample dataset, which comprises a multidimensional sensor dataset of samples and a probability set of thermal runaway risks for samples, including: Based on the physical structure information, the electrical parameter information, and the thermophysical parameters, physical structure features, electrical parameter features, and thermophysical features are constructed. Energy storage units are searched using the physical structural features, electrical parameter features, and thermophysical features as search constraints to identify multiple similar energy storage units. By retrieving historical operation records from multiple similar energy storage units, a set of historical operation records is obtained; Obtain a data analysis window, and sample each historical operation record in the historical operation record set according to the data analysis window to obtain a sample multidimensional sensor dataset; The thermal runaway risk probability set of the sample multidimensional sensor dataset is labeled with the thermal runaway risk probability. The historical sample dataset is obtained by summarizing the sample multidimensional sensing dataset and the sample thermal runaway risk probability set.

[0048] The early warning execution module 14 is specifically used for: Based on the aforementioned thermal runaway risk probability, an early warning system for energy storage thermal runaway is implemented, including: Set a risk probability threshold; The probability of thermal runaway risk is compared with the risk probability threshold. When the probability of thermal runaway risk is greater than or equal to the risk probability threshold, an early warning of thermal runaway in energy storage is issued.

[0049] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0050] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0051] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A standard modular energy storage thermal runaway early warning method based on digital twins, characterized in that, The method includes: Multiple types of sensors are deployed in standard modular energy storage units to form a sensor network; Based on the sensor network, the standard modular energy storage unit is used to collect operational data to obtain multi-dimensional sensing data; The multidimensional sensor data is transmitted to the digital twin platform to predict thermal runaway of the standard modular energy storage unit and obtain the probability of thermal runaway risk. Based on the aforementioned probability of thermal runaway risk, an early warning system for thermal runaway in energy storage is developed.

2. The method according to claim 1, characterized in that, Deploying multiple types of sensors within standard modular energy storage units to form a sensor network, including: Obtain the unit model of the standard modular energy storage unit, and perform sensor matching in the sensor configuration library based on the unit model to determine multiple types of sensors; Sensors are configured in the standard modular energy storage unit according to multiple types of sensors to form a sensor network.

3. The method according to claim 1, characterized in that, Based on the sensor network, operational data of the standard modular energy storage unit is collected to obtain multi-dimensional sensing data, including: Data is collected from the standard modular energy storage unit through the sensor network at a preset acquisition frequency to obtain multiple types of sensor data, and the multiple types of sensor data are entered into the energy storage unit operation database. Receive thermal runaway test command, and retrieve the data analysis window based on the thermal runaway test command; The multidimensional sensing data is obtained by extracting data from the energy storage unit's operating database using the data analysis window.

4. The method according to claim 1, characterized in that, The multidimensional sensor data is transmitted to a digital twin platform to predict thermal runaway of the standard modular energy storage unit, obtaining the probability of thermal runaway risk, including: Obtain the input mapping format of the digital twin platform; The multidimensional sensing data is converted based on the input mapping format to obtain formatted sensing data that conforms to the input of the digital twin platform; The digital twin platform performs thermal runaway prediction based on the formatted sensor data to obtain the thermal runaway risk probability.

5. The method according to claim 1, characterized in that, Constructing the digital twin platform includes: Obtain the physical structure information, electrical parameter information, and thermophysical parameters of the standard modular energy storage unit; Based on the physical structure information, the electrical parameter information, and the thermophysical parameters, an initial digital twin model of the standard modular energy storage unit is constructed. Obtain a historical sample dataset, which includes a multidimensional sensor dataset of samples and a probability set of thermal runaway risks of samples; Using the sample multidimensional sensing dataset as input and the sample thermal runaway risk probability set as supervision signal, the initial digital twin model is trained to obtain the digital twin platform.

6. The method according to claim 5, characterized in that, Obtain a historical sample dataset, which includes a multidimensional sensor dataset of samples and a probability set of thermal runaway risks for samples, including: Based on the physical structure information, the electrical parameter information, and the thermophysical parameters, physical structure features, electrical parameter features, and thermophysical features are constructed. Energy storage units are searched using the physical structural features, electrical parameter features, and thermophysical features as search constraints to identify multiple similar energy storage units. By retrieving historical operation records from multiple similar energy storage units, a set of historical operation records is obtained; Obtain a data analysis window, and sample each historical operation record in the historical operation record set according to the data analysis window to obtain a sample multidimensional sensor dataset; The thermal runaway risk probability set of the sample multidimensional sensor dataset is labeled with the thermal runaway risk probability. The historical sample dataset is obtained by summarizing the sample multidimensional sensing dataset and the sample thermal runaway risk probability set.

7. The method according to claim 1, characterized in that, Based on the aforementioned thermal runaway risk probability, an early warning system for energy storage thermal runaway is implemented, including: Set a risk probability threshold; The probability of thermal runaway risk is compared with the risk probability threshold. When the probability of thermal runaway risk is greater than or equal to the risk probability threshold, an early warning of thermal runaway in energy storage is issued.

8. A standard modular energy storage thermal runaway early warning system based on digital twins, characterized in that, For performing the method according to any one of claims 1-7, comprising: The sensing module is used to deploy multiple types of sensors in a standard modular energy storage unit to form a sensor network; The data acquisition and transmission module is used to acquire operational data of the standard modular energy storage unit based on the sensor network, obtain multi-dimensional sensing data, and transmit the multi-dimensional sensing data to the digital twin platform. The digital twin prediction module is used to predict thermal runaway of the standard modular energy storage unit based on the multidimensional sensing data, and obtain the probability of thermal runaway risk. The early warning execution module is used to provide early warning of thermal runaway in energy storage based on the probability of thermal runaway risk.

Images