Energy storage safety early warning method and system based on deep learning

Abstract

The invention discloses an energy storage safety early warning method and system based on deep learning, and relates to the technical field of electrical safety monitoring, and the method comprises the steps: inputting historical operation state data, current operation state data, historical environment feature data and current environment feature data into a data prediction network, performing operation state data prediction and environment characteristic data prediction based on the data prediction network, and generating an operation state data prediction quantity and an environment characteristic data prediction quantity; the safety state deviation degree of the energy storage equipment at the current moment is determined according to the running state data predictive quantity and the environment characteristic data predictive quantity, and the pyrolysis particle target index at the current moment is determined according to the current environment characteristic data and the safety state deviation degree; according to the pyrolysis particle target index, the early warning mode of the energy storage equipment at the current moment is determined, and the technical problem that in the prior art, the early warning monitoring accuracy of the energy storage equipment is low is solved.

Description

Technical Field

[0001] This application relates to the technical field of electrical safety monitoring, and particularly to a method and system for energy storage safety early warning based on deep learning. Background Art

[0002] Traditional fire monitoring technologies mainly rely on means such as temperature sensors, smoke detectors or infrared monitoring, and have problems such as response lag, susceptibility to interference, and inability to achieve early warning. With the increasingly wide application of new technologies such as energy storage devices, the safety monitoring of energy storage devices has attracted more and more attention. During the operation of energy storage devices, traditional safety monitoring methods mainly rely on fixed thresholds and single-dimensional data analysis, and it is difficult to comprehensively and accurately evaluate the safety status of devices, resulting in late warning or high false alarm rates, and potential safety hazards cannot be detected in time. Especially in the complex and changeable actual operating environment, the safety risks of energy storage devices become more difficult to predict and control. Therefore, there is currently a lack of a method that can comprehensively process multi-dimensional data and provide accurate early warnings for energy storage devices.

[0003] The above content is only used to assist in understanding the technical solution of this application, and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of this application is to provide a method and system for energy storage safety early warning based on deep learning, aiming to solve the technical problem of low accuracy of early warning monitoring for energy storage devices in the prior art.

[0005] To achieve the above purpose, this application proposes a method for energy storage safety early warning based on deep learning. The method for energy storage safety early warning based on deep learning includes: Obtain the historical operating status data and historical environmental characteristic data of the energy storage device at historical moments, and obtain the current operating status data and current environmental characteristic data of the energy storage device at the current moment; Input the historical operating status data, the historical environmental characteristic data, the current operating status data, and the current environmental characteristic data into a data prediction network, and obtain the predicted values of the operating status data and environmental characteristic data of the energy storage device at a future moment generated by the data prediction network; the data prediction network is obtained by training a long short-term memory network (LSTM) based on a training sample set; Determine the deviation degree of the safety state of the energy storage device at the current moment according to the predicted value of the operating status data and the predicted value of the environmental characteristic data; Determine the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the deviation degree of the safety state; Determine the early warning method of the energy storage device at the current moment according to the pyrolysis particle target index.

[0006] In some embodiments, determining the safety state deviation degree of the energy storage device at the current moment according to the predicted operation state data and the predicted environmental characteristic data includes: Obtain the preset operation state data reference range and the preset environmental characteristic data reference range corresponding to the energy storage device; Calculate the difference between the predicted operation state data and the preset operation state data reference range to obtain the operation state data deviation; Calculate the difference between the predicted environmental characteristic data and the preset environmental characteristic data reference range to obtain the environmental characteristic data deviation; Determine the safety state deviation degree according to the operation state data deviation and the environmental characteristic data deviation.

[0007] In some embodiments, determining the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the safety state deviation degree includes: Determine the pyrolysis particle initial index at the current moment according to the current environmental characteristic data; Based on the safety state deviation degree, correct the pyrolysis particle initial index to obtain the pyrolysis particle target index at the current moment.

[0008] In some embodiments, determining the pyrolysis particle initial index at the current moment according to the current environmental characteristic data includes: Determine the current time-sharing stage category at the current moment according to the current environmental temperature change rate; According to the first correspondence between the preset time-sharing stage category and the corresponding relation table, determine the target corresponding relation table corresponding to the current time-sharing stage category, and the target corresponding relation table includes the correspondence between various parameters and weighting coefficients; According to the target corresponding relation table, determine the weighting coefficients of various parameters in the current environmental characteristic data; Based on the weighting coefficients of various parameters in the current environmental characteristic data and the current environmental characteristic data, determine the pyrolysis particle initial index, where the current environmental data includes the current environmental temperature change rate.

[0009] In some embodiments, based on the safety state deviation degree, correcting the pyrolysis particle initial index to obtain the pyrolysis particle target index at the current moment includes: If the safety state deviation degree is less than the first deviation threshold, determine the pyrolysis particle initial index as the pyrolysis particle target index; If the safety state deviation degree is greater than or equal to the first deviation threshold and less than the second deviation threshold, a weighted sum of the safety state deviation degree and the initial pyrolysis particle index is obtained to get a first summation result, and the first summation result is determined as the pyrolysis particle target index, where the second deviation threshold is greater than the first deviation threshold; If the safety state deviation degree is greater than or equal to the second deviation threshold, a weighted sum of the safety state deviation degree and the initial pyrolysis particle index is obtained to get a second summation result, and the values of the second summation result and a preset pyrolysis particle index are compared, and the larger one of them is determined as the pyrolysis particle target index.

[0010] In some embodiments, determining the warning method of the energy storage device at the current moment according to the pyrolysis particle target index includes: Obtain a first warning threshold and a second warning threshold, where the second warning threshold is greater than the first warning threshold; Obtain an energy storage device image, which includes the energy storage device; When the pyrolysis particle target index is greater than or equal to the first warning threshold and less than the second warning threshold, determine the warning method as the first warning method, and control to display a first warning message at the user end, where the first warning message includes the pyrolysis particle target index and the energy storage device image; When the pyrolysis particle target index is greater than or equal to the second warning threshold, determine the warning method as the second warning method, control to display a second warning message at the user end and control the user end to emit a warning sound for prompting the second warning method, and control the energy storage device to emit a warning sound for prompting the second warning method, where the second warning message includes the pyrolysis particle target index and the energy storage device image.

[0011] In some embodiments, after determining the warning method as the first warning method, it further includes: Generate a first control instruction, and based on the first control instruction, control the ventilation device connected to the energy storage device to be turned on; and, Generate a second control instruction, and based on the second control instruction, control the load of the energy storage device to migrate from the energy storage device to a standby line.

[0012] In some embodiments, the operation state data includes voltage and current, and the environmental characteristic data includes environmental temperature, target volatile gas concentration, and target smoke particle concentration. Inputting the historical operation state data, the historical environmental characteristic data, the current operation state data, and the current environmental characteristic data into the data prediction network includes: The Kalman filter algorithm is used to smooth the historical operation state data, the historical environmental characteristic data, the current operation state data, and the current environmental characteristic data to obtain a filtered data sequence; Normalize various parameters in the filtered data sequence to obtain a normalized data sequence, which includes a normalized voltage data sequence, a normalized current data sequence, a normalized environmental temperature data sequence, a normalized target volatile gas concentration data sequence, and a normalized target smoke particle concentration data sequence; Determine the normalized data sequence as the input sequence and input the input sequence into the data prediction network.

[0013] In some embodiments, the historical environmental characteristic data and the current environmental characteristic data are obtained through the following steps: Obtain the material parameters of the energy storage device, and determine the target volatile gas concentration and target smoke particles according to the material parameters of the energy storage device; Obtain the historical environmental temperature before the current moment and the current environmental temperature at the current moment collected by the temperature sensor, and determine the current environmental temperature change rate based on the historical environmental temperature and the current environmental temperature; Obtain the historical target volatile gas concentration before the current moment and the current target volatile gas concentration at the current moment collected by the electrochemical sensor, and determine the current target volatile gas concentration change rate based on the historical target volatile gas concentration and the current target volatile gas concentration; Obtain the historical target smoke particle concentration before the current moment and the current target smoke particle concentration at the current moment collected by the laser particle sensor, and determine the current target smoke particle concentration change rate based on the historical target smoke particle concentration and the current target smoke particle concentration; Determine the historical environmental temperature, the historical target volatile gas concentration, and the historical target smoke particle concentration as the historical environmental characteristic data; Determine the current environmental temperature, the current environmental temperature change rate, the current target volatile gas concentration, the current target volatile gas concentration change rate, the current target smoke particle concentration change rate, and the current target smoke particle concentration as the current environmental characteristic data.

[0014] In some embodiments, the obtaining of the first warning threshold and the second warning threshold includes: Obtain a first preset warning initial value; Determine a current first correction amount corresponding to the current ambient temperature according to the current ambient temperature and a second mapping relationship, where the second mapping relationship is used to represent the mapping relationship between the ambient temperature and the first correction amount, and the ambient temperature and the first correction amount are positively correlated; Correct the first preset warning initial value according to the current first correction amount to obtain the first warning threshold; Obtain a second preset warning initial value; Determine a current second correction amount corresponding to the current ambient temperature according to the current ambient temperature and a third mapping relationship, where the third mapping relationship is used to represent the mapping relationship between the ambient temperature and the second correction amount, and the ambient temperature and the second correction amount are positively correlated; Correct the second preset warning initial value according to the current second correction amount to obtain the second warning threshold.

[0015] In addition, to achieve the above object, the present application also proposes an energy storage safety warning system based on deep learning, and the energy storage safety warning system based on deep learning includes: A multi-source data acquisition module, configured to acquire historical operation state data and historical environmental feature data of the energy storage device at a historical moment, and acquire current operation state data and current environmental feature data of the energy storage device at the current moment; A prediction module, configured to input the historical operation state data, the historical environmental feature data, the current operation state data, and the current environmental feature data into a data prediction network, and obtain predicted values of the operation state data and environmental feature data of the energy storage device at a future moment generated by the data prediction network; the data prediction network is obtained by training a long short-term memory network (LSTM) based on a training sample set; A deviation degree determination module, configured to determine a safety state deviation degree of the energy storage device at the current moment according to the predicted value of the operation state data and the predicted value of the environmental feature data; An index determination module, configured to determine a pyrolysis particle target index at the current moment according to the current environmental feature data and the safety state deviation degree; A warning determination module, configured to determine a warning method of the energy storage device at the current moment according to the pyrolysis particle target index.

[0016] In addition, to achieve the above object, the present application also proposes an energy storage safety warning system based on deep learning, and the system includes: a memory, a processor, and an energy storage safety warning program based on deep learning stored on the memory and executable on the processor, where the energy storage safety warning program based on deep learning is configured to implement the steps of the energy storage safety warning method based on deep learning as described above.

[0017] In addition, to achieve the above object, the present application also proposes a storage medium, the storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the energy storage safety warning method based on deep learning as described above are implemented.

[0018] In addition, to achieve the above object, the present application also provides a computer program product, the computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the energy storage safety warning method based on deep learning as described above are implemented.

[0019] One or more technical solutions proposed in this application have at least the following technical effects: By obtaining the historical operating status data of the energy storage device at a historical moment and the current operating status data at the current moment, as well as obtaining the historical environmental characteristic data at the historical moment and the current environmental characteristic data at the current moment; inputting the historical operating status data, the current operating status data, the historical environmental characteristic data, and the current environmental characteristic data into a data prediction network, and performing prediction on the operating status data and the environmental characteristic data based on the data prediction network to generate a predicted value of the operating status data and a predicted value of the environmental characteristic data of the energy storage device at a future moment. The data prediction network is obtained by training a Long Short-Term Memory Network (LSTM) based on a training sample set, which can capture long-term dependencies in time series and predict future data. Determine the safety state deviation degree of the energy storage device at the current moment according to the predicted value of the operating status data and the predicted value of the environmental characteristic data, evaluate the deviation degree between the current device and the normal state, and determine the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the safety state deviation degree. The pyrolysis particle target index reflects the current safety risk level of the energy storage device. Different index ranges correspond to different warning methods. According to the pyrolysis particle target index, determine the warning method of the energy storage device at the current moment and take corresponding response measures. The energy storage safety warning method based on deep learning provided in this application integrates multi-source heterogeneous data (operating status data and environmental characteristic data), uses the trained LSTM network (data prediction network) for prediction and analysis, combines historical data and current moment data for trend prediction, predicts the state change and environmental change of the energy storage device in advance, and calculates the safety state deviation degree of the energy storage device at the current moment according to the prediction results (predicted value of environmental characteristic data and predicted value of operating status data). Combining the current environmental characteristic data and the safety state deviation degree, determine the warning method of the energy storage device at the current moment, realize efficient early warning and accurate safety state evaluation of the energy storage device, solve the technical problem of low warning monitoring accuracy for energy storage devices in the prior art, not only improve the accuracy and timeliness of energy storage safety warning, but also enhance the adaptability to the characteristics of energy storage devices, and better cope with the complex working conditions of energy storage devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings here are incorporated into the specification and form a part of this specification, showing embodiments consistent with this application, and are used together with the specification to explain the principles of this application.

[0021] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0022] Figure 1 FIG. 4 is a schematic flowchart provided for Embodiment 1 of the energy storage safety early warning method based on deep learning of the present application; Figure 2 FIG. 5 is a schematic module structure diagram of the energy storage safety early warning system based on deep learning according to an embodiment of the present application; Figure 3 FIG. 6 is a schematic system structure diagram of the hardware operating environment involved in the energy storage safety early warning method based on deep learning in an embodiment of the present application. Detailed Embodiment

[0023] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0024] To better understand the technical solutions of the present application, the following will be described in detail in conjunction with the accompanying drawings of the specification and specific embodiments.

[0025] The main solution of the embodiment of the present application is as follows: obtaining the historical operation state data and historical environmental characteristic data of the energy storage device at a historical moment, and obtaining the current operation state data and current environmental characteristic data of the energy storage device at the current moment; inputting the historical operation state data, historical environmental characteristic data, current operation state data, and current environmental characteristic data into a data prediction network to obtain the predicted operation state data and predicted environmental characteristic data of the energy storage device at a future moment generated by the data prediction network; the data prediction network is obtained by training a long short-term memory network (LSTM) based on a training sample set; determining the safety state deviation degree of the energy storage device at the current moment according to the predicted operation state data and predicted environmental characteristic data; determining the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the safety state deviation degree; and determining the early warning method of the energy storage device at the current moment according to the pyrolysis particle target index.

[0026] In this embodiment, for the convenience of description, the following will be described with the energy storage safety early warning system based on deep learning as the execution subject.

[0027] Traditional electrical fire monitoring technologies mainly rely on means such as temperature sensors, smoke detectors, or infrared monitoring. Although these methods can provide fire alarms to a certain extent, they generally have problems such as response lag and susceptibility to environmental interference. In recent years, the application of pyrolysis particle detectors has gradually increased. Although pyrolysis particle detectors can detect micron-sized particles and gases generated by the pyrolysis of insulating materials and can provide a more refined monitoring means compared to temperature sensors, they lack the ability to perform intelligent fusion analysis on multi-source heterogeneous data and are difficult to adapt to complex and changing actual application scenarios. In addition, existing technologies lack targeted solutions for the safety early warning of new application scenarios such as energy storage devices. Energy storage devices have unique operating characteristics, including high energy density, fast charging and discharging, etc., which require higher safety monitoring. Traditional safety monitoring methods can no longer meet the requirements of energy storage devices under complex working conditions, and effective and accurate safety monitoring cannot be carried out through existing single sensors or simple combinations.

[0028] To address the above problems, this application proposes a deep learning-based energy storage safety early warning method. The deep learning-based energy storage safety early warning method integrates multi-source heterogeneous data (operating state data and environmental characteristic data), uses a trained LSTM network (data prediction network) for prediction and analysis, combines historical data and current moment data for trend prediction, predicts the state changes and environmental changes of energy storage devices in advance, and calculates the safety state deviation degree of the energy storage device at the current moment according to the prediction results (predicted values of environmental characteristic data and predicted values of operating state data). Combining the current environmental characteristic data and the safety state deviation degree, the early warning method for the energy storage device at the current moment is determined, realizing efficient early warning of the energy storage device and accurate safety state assessment, solving the technical problem of low accuracy in early warning monitoring for energy storage devices in the existing technology, not only improving the accuracy and timeliness of energy storage safety early warning, but also enhancing the adaptability to the characteristics of energy storage devices and better coping with the complex working conditions of energy storage devices.

[0029] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program running functions, or an energy storage device that can implement the above functions. Hereinafter, taking the deep learning-based energy storage safety early warning system as an example, this embodiment and the following embodiments will be described.

[0030] Based on this, the embodiments of this application provide a deep learning-based energy storage safety early warning method, referring to Figure 1 , Figure 1 is the flowchart of the first embodiment of the deep learning-based energy storage safety early warning method of this application.

[0031] In this embodiment, the deep learning-based energy storage safety early warning method includes steps 101 to 105: Step 101: Obtain the historical operation status data and historical environmental characteristic data of the energy storage device at historical moments, and obtain the current operation status data and current environmental characteristic data of the energy storage device at the current moment.

[0032] Specifically, obtain the first preset number of historical operation status data of the energy storage device before the current moment and the current operation status data at the current moment, and obtain the first preset number of historical environmental characteristic data before the current moment and the current environmental characteristic data at the current moment. The energy storage device is a device for storing electrical energy, such as an energy storage battery cabinet, an energy storage battery cabinet, etc. The operation status and safety of the energy storage device directly affect the reliability and stability of the entire system. The current moment is the time point monitored by the system in real time and is the reference time for data collection and analysis. The first preset number is a set fixed quantity, which is the first preset number of sampling points within a preset time window (such as 10 minutes). For example, 1000 sampling points collected in the past 10 minutes. This application does not make specific numerical limitations on the first preset number. In this application, the value of the first preset number is greater than or equal to two. The historical operation status data is the operation parameters of the energy storage device in the past period of time, such as current, voltage, etc. The first preset number of historical operation status data can reflect the historical operation trend of the energy storage device. The current operation status data is the real-time operation parameters of the energy storage device at the current moment, such as the current temperature value, current, voltage. The current operation status data can reflect the immediate status of the device. The environmental characteristic data is the physical quantity parameters indicating fire hazards in the environment where the energy storage device is located, such as environmental temperature, target volatile gas concentration, and target smoke particle concentration, etc. The first preset number of historical environmental characteristic data is the physical quantity parameters indicating fire hazards in the environment where the energy storage device is located in the past period of time, which can reflect the change trend of the environment where the energy storage device is located in the past period of time. The current environmental characteristic data is the physical quantity parameters indicating fire hazards in the environment where the energy storage device is located at the current moment, which can reflect the immediate status of the current environment.

[0033] In some embodiments, multi-source sensors (such as voltage sensors, current sensors, etc.) can be used to collect the operating state data of the energy storage device, sample it at a fixed frequency (such as 100 times per second), and store it in a database. At the same time, environmental sensors (such as electrochemical sensors, temperature sensors, and laser particle sensors, etc.) are used to collect the environmental characteristic data of the environment where the energy storage device is located, sample it at a fixed frequency, and store it synchronously with the operating state data. The first preset number of historical operating state data (for example, the operating state data in the past 10 minutes) can be extracted from the database. The first preset number of historical operating state data can be represented as a time series, and the operating state data at the current moment is collected as a real-time input. The first preset number of historical environmental characteristic data (for example, data such as temperature, gas concentration, and particle concentration in the past 10 minutes) can be extracted from the database, and the environmental characteristic data at the current moment is collected to supplement the real-time information. The above historical and current operating state data, historical and current environmental characteristic data are integrated to obtain a multi-dimensional time series data set as the input for subsequent deep learning models (such as data prediction networks). By combining historical and current operating state data, the changing trend of the device's state is fully reflected. The historical operating state data can provide context information, which helps to better understand whether the current state has a long-term trend in the future. By integrating multiple sensors (such as electrochemical sensors, temperature sensors, and laser particle sensors) to real-time monitor multiple environmental parameters (environmental characteristic data) of the environment where the energy storage device is located, the environmental characteristic data can include environmental temperature, target volatile gas concentration, and target smoke particle concentration, etc., comprehensively covering the physical quantities that may indicate fire hazards. Combining environmental characteristic data and operating state data can more comprehensively evaluate the safety state of the device, avoid misjudgment caused by single data, and improve the accuracy and pertinence of early warning.

[0034] By obtaining the historical and current operating state data and environmental characteristic data of the energy storage device, a comprehensive and multi-dimensional data foundation can be constructed. Through the fusion analysis of multi-dimensional data, it helps to more accurately identify potential safety hazards, solve the problems of late warning, high false alarm rate, and inability to fuse multi-dimensional data in traditional technologies, and improve the early hazard prediction ability and intelligent level of the energy storage safety monitoring system.

[0035] Step 102: Input the historical operating state data, historical environmental characteristic data, current operating state data, and current environmental characteristic data into the data prediction network to obtain the predicted values of the operating state data and environmental characteristic data of the energy storage device at a future moment generated by the data prediction network; the data prediction network is obtained by training a long short-term memory network (LSTM) based on a training sample set.

[0036] Specifically, input the first preset number of historical operation status data, current operation status data, first preset number of historical environmental feature data, and current environmental feature data into the data prediction network. Based on the data prediction network, perform operation status data prediction and environmental feature data prediction to generate the second preset number of predicted operation status data after the current moment and the second preset number of predicted environmental feature data after the current moment. The data prediction network is obtained by training a long short-term memory network (LSTM) based on a training sample set. The data prediction network is obtained by training a deep learning model (such as LSTM) based on a training sample set. The data prediction network analyzes and predicts the first preset number of historical operation status data, current operation status data, first preset number of historical environmental feature data, and current environmental feature data to generate predicted operation status data and predicted environmental feature data at a future time point. The second preset number of predicted operation status data is the predicted operation status data of the energy storage device within a future period of time (such as the next 30 minutes or the next 1000 sampling points) based on historical and current operation status data through the data prediction network. The second preset number of predicted environmental feature data is the predicted environmental feature data within a future period of time (such as the next 30 minutes or the next 1000 sampling points) based on historical and current environmental feature data through the data prediction network. This application does not make specific numerical limitations on the second preset. In this application, the value of the first preset is greater than or equal to two.

[0037] In some embodiments, the training process of the data prediction network may be as follows: Obtain multiple sets of data, where each set of data includes operating status data and environmental feature data. Use ten-fold cross-validation to randomly allocate the multiple sets of data into a training sample set and a test sample set. The training sample set includes multiple training samples, and each training sample includes multiple historical operating status data of the energy storage device at historical moments, multiple historical environmental feature data at historical moments, multiple predicted values of the operating status data of the energy storage device after historical moments, and multiple predicted values of the environmental feature data. Use the training sample set to train the LSTM. The input of the LSTM is multiple historical operating status data and multiple historical environmental feature data, and the output is multiple predicted values of the operating status data and multiple predicted values of the environmental feature data for the relevant input. The optimizer selected for LSTM training can be Adam. The performance of the model is evaluated using the root mean square error to optimize the deviation between the predicted value and the true value. The trained LSTM (i.e., the data prediction network) can predict the predicted value of the operating status data of the energy storage at a future moment and the predicted value of the environmental feature data of the environment where the energy storage device is located through multiple historical operating status data and multiple historical environmental feature data. Through its internal memory units and gating mechanisms (input gate, forget gate, output gate), the LSTM can effectively retain and forget information, thereby capturing long-term dependencies in the input sequence. Even if the time series is very long, it can remember important information segments. Therefore, the data prediction network obtained by training the LSTM based on the training sample set also has this characteristic. By learning the long-term patterns in the training sample set through the training process, the data prediction network can handle long-term dependencies in time series data, thereby more accurately predicting future operating status data and future environmental feature data. In addition, the data prediction network has high computational efficiency in prediction and control and can meet real-time requirements. The data prediction network algorithm can be processed in parallel on hardware accelerators such as GPUs to improve real-time performance. By applying the LSTM to the multi-parameter time series correlation analysis of the energy storage system, it helps to solve the problem of low accuracy of traditional methods for energy storage safety warning.

[0038] In some embodiments, the first preset number of historical operating state data, current operating state data, first preset number of historical environmental characteristic data, and current environmental characteristic data are input into a data prediction network. Based on the trained data prediction network, using the learned patterns, the input first preset number of historical operating state data, current operating state data, first preset number of historical environmental characteristic data, and current environmental characteristic data are processed to predict the future operating state data of the energy storage device and the future environmental characteristic data of the environment where the energy storage device is located. By using multiple historical operating state data, current operating state data, multiple historical environmental characteristic data, and current environmental characteristic data, the data prediction network can capture the trends and patterns of the operating state data and environmental characteristic data of the energy storage device evolving over time, and obtain accurate prediction results, that is, the second preset number of predicted values of the operating state data of the energy storage device after the current moment and the second preset number of predicted values of the environmental characteristic data after the current moment.

[0039] By using LSTM to comprehensively analyze multi-source heterogeneous data, extracting complex non-linear relationships from it, and realizing intelligent fusion analysis of multi-dimensional data to predict future operating states and environmental characteristics, abnormal trends can be identified more accurately, and false alarms caused by single data fluctuations can be reduced. At the same time, the energy storage safety warning system based on deep learning can issue warnings at the early stage of potential hazards, thereby improving the response speed and warning ability. In addition, the trained data prediction network can better adapt to the complex and changeable actual application scenarios of energy storage devices, which helps to provide more accurate safety monitoring and warnings.

[0040] Step 103, determine the safety state deviation degree of the energy storage device at the current moment according to the predicted values of the operating state data and the predicted values of the environmental characteristic data.

[0041] Specifically, determine the safety state deviation degree of the energy storage device at the current moment according to the second preset number of predicted values of the operating state data and the second preset number of predicted values of the environmental characteristic data. The safety state deviation degree is the degree of deviation between the predicted operating state data and environmental characteristic data of the energy storage device and the safe operating state data and environmental characteristic data in the safe state, which can be calculated by comparing the predicted values with the safe state baseline value. The calculated safety state deviation degree of the energy storage device at the current moment can quantify the deviation degree of the current state of the energy storage device from the safe normal baseline and reflect the cumulative effect of potential risks.

[0042] In some embodiments, the second preset number of predicted operating state data quantities and the second preset number of predicted environmental characteristic data quantities obtained by predicting through a data prediction network are compared with the safety state baseline value, and the deviation degree of each parameter is calculated to comprehensively obtain the safety state deviation degree of the energy storage device at the current moment. The safety state baseline value can be established based on historical normal data, for example, the normal ranges of temperature, gas concentration, etc.

[0043] Step 104: Determine the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the safety state deviation degree.

[0044] Specifically, the pyrolysis particle target index is a dynamic risk score that comprehensively combines the safety deviation degree and the current environmental characteristic data, which can reflect the pyrolysis degree of the reaction insulation material and quantify the current safety risk level of the energy storage device, and is used for hierarchical early warning decision-making. The higher the calculated pyrolysis particle target index, the higher the risk at the current moment, the greater the probability of an electrical fire occurring, and the higher the level of the subsequent early warning method to be adopted.

[0045] In some embodiments, by combining the current environmental characteristic data and the safety state deviation degree, the pyrolysis particle target index can be determined through a specific algorithm (such as weighted average, machine learning model, etc.), which reflects the current safety risk level of the energy storage device and the pyrolysis degree of the insulation material. By calculating the safety state deviation degree and combining the current environmental characteristic data to determine the pyrolysis particle target index, a comprehensive assessment of the safety state of the energy storage device is realized, which helps to improve the accuracy and timeliness of early warning and enhance the adaptability to complex application scenarios, so as to solve the technical problems of early warning lag, high false alarm rate and inability to predict early hidden dangers in the existing methods.

[0046] Step 105: Determine the early warning method of the energy storage device at the current moment according to the pyrolysis particle target index.

[0047] Specifically, the pyrolysis particle target index is a dynamic risk score generated by fusing the safety state deviation degree and the real-time environmental data (current environmental characteristic data), and the numerical range can be 0 - 4000, which can reflect the pyrolysis degree of the insulation material and quantify the current safety risk level of the energy storage device. The higher the value, the more urgent the risk. The early warning method is to determine the corresponding early warning level and response measures according to different value ranges of the pyrolysis particle target index. The early warning method can include level 1 early warning, level 2 early warning, etc., and each early warning level corresponds to different response measures. According to the pyrolysis particle target index, the early warning method of the energy storage device at the current moment is determined, and the early warning methods with different emergency levels are matched according to the value of the pyrolysis particle target index to avoid overreaction.

[0048] Based on the energy storage safety warning method based on deep learning provided in this application, by obtaining the historical operation state data of the energy storage device at historical moments and the current operation state data at the current moment, as well as obtaining the historical environmental characteristic data at historical moments and the current environmental characteristic data at the current moment; inputting the historical operation state data, the current operation state data, the historical environmental characteristic data, and the current environmental characteristic data into the data prediction network, and performing operation state data prediction and environmental characteristic data prediction based on the data prediction network to generate the predicted values of the operation state data and the environmental characteristic data of the energy storage device at future moments. The data prediction network is obtained by training the LSTM based on the training sample set, which can capture the long-term dependencies in the time series and predict future data. Determine the safety state deviation degree of the energy storage device at the current moment according to the predicted values of the operation state data and the environmental characteristic data, evaluate the deviation degree between the current device and the normal state, and determine the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the safety state deviation degree. The pyrolysis particle target index reflects the current safety risk level of the energy storage device. Different index ranges correspond to different warning methods. According to the pyrolysis particle target index, determine the warning method of the energy storage device at the current moment and take corresponding response measures. The energy storage safety warning method based on deep learning provided in this application integrates multi-source heterogeneous data (operation state data and environmental characteristic data), uses the trained LSTM (data prediction network) for prediction and analysis, combines historical data and current moment data for trend prediction, predicts the state change and environmental change of the energy storage device in advance, and calculates the safety state deviation degree of the energy storage device at the current moment according to the prediction results (predicted values of environmental characteristic data and operation state data). Combining the current environmental characteristic data and the safety state deviation degree, determine the warning method of the energy storage device at the current moment, realize the efficient early warning of the energy storage device and the accurate safety state assessment, solve the technical problem of low warning monitoring accuracy for energy storage devices in the prior art, not only improve the accuracy and timeliness of energy storage safety warning, but also enhance the adaptability to the characteristics of energy storage devices and better cope with the complex working conditions of energy storage devices.

[0049] In some embodiments, determining the safety state deviation degree of the energy storage device at the current moment according to the predicted values of the operation state data and the environmental characteristic data includes: Obtain the preset operation state data reference range and the preset environmental characteristic data reference range corresponding to the energy storage device; Calculate the difference between the predicted value of the operation state data and the preset operation state data reference range to obtain the operation state data deviation; Calculate the difference between the predicted value of the environmental characteristic data and the preset environmental characteristic data reference range to obtain the environmental characteristic data deviation; Determine the safety state deviation degree based on the operation status data deviation and the environmental characteristic data deviation.

[0050] Specifically, the number of predicted operation status data can be the second preset number. The second preset number of predicted operation status data are the predicted values of the operation status data of the energy storage device within a future period of time (such as the next 10 minutes) generated by the data prediction network, including voltage, current, etc. The number of predicted environmental characteristic data can be the second preset number. The second preset number of predicted environmental characteristic data are the predicted values of the environmental characteristic data within a future period of time generated by the data prediction network, including environmental temperature, target volatile gas concentration, target smoke particle concentration, etc. The preset operation status data reference range can be obtained according to the design specifications of the energy storage device and the reasonable ranges of various operation parameters under normal operation conditions statistically obtained from historical data. For example, the normal range of voltage is [48V, 52V]. The preset environmental characteristic data reference range can be obtained according to the safety requirements of the environment where the energy storage device is located and the normal environmental parameter ranges statistically obtained from historical data. For example, the normal range of environmental temperature is [20°C, 30°C].

[0051] As an example, the difference between the predicted operation status data and the preset operation status data reference range can be calculated through the following formula to obtain the operation status data deviation: .

[0052] Where is the deviation of the i-th operation status data, is the predicted value of the i-th operation status, is the median value of the corresponding preset operation status data reference range.

[0053] The difference between the predicted environmental characteristic data and the preset environmental characteristic data reference range can be calculated through the following formula to obtain the environmental characteristic data deviation: .

[0054] Where, is the deviation of the j-th environmental characteristic data, is the predicted value of the j-th environmental characteristic, is the lower limit of the corresponding preset environmental characteristic data reference range, is the upper limit of the corresponding preset environmental characteristic data reference range.

[0055] The safety state deviation degree at a future moment can be determined through the following formula based on the deviation of the j-th environmental characteristic data and the deviation of the i-th operation status data: .

[0056] Where, is the safety state deviation degree, is the dynamic weight of the operation status data deviation, is the dynamic weight of the environmental characteristic data deviation, n is the number of items of the operating state data, and m is the number of items of the environmental characteristic data. According to the above method, the safety state deviation degree at the second preset future moment is calculated, and the safety state deviation degrees at the second preset future moments are weighted and summed to obtain the safety state deviation degree of the energy storage device at the current moment. The closer the safety state deviation degree is to 0, the closer the energy storage device is to the normal state. The larger the safety state deviation degree is, the higher the safety risk of the energy storage device is. By calculating the deviation of the predicted values of the operating state data and environmental characteristic data of the energy storage device, a comprehensive safety state deviation degree index, that is, the safety state deviation degree, is generated to comprehensively evaluate the overall safety state of the energy storage device, avoid misjudgment caused by single-dimensional analysis, and quantify the deviation degree by integrating multi-dimensional data, which helps to more accurately identify potential safety hazards.

[0057] In some embodiments, according to the current environmental characteristic data and the safety state deviation degree, determining the pyrolysis particle target index at the current moment includes: Determining the initial pyrolysis particle index at the current moment according to the current environmental characteristic data; Based on the safety state deviation degree, correcting the initial pyrolysis particle index to obtain the pyrolysis particle target index at the current moment.

[0058] Specifically, the initial pyrolysis particle index is an index preliminarily calculated based on the current environmental characteristic data, which can quantify the pyrolysis particle concentration situation of the energy storage device at the current moment and reflect the current safety risk degree of the energy storage device. The pyrolysis particle target index is the final index corrected by comprehensively considering the initial pyrolysis particle index and the safety state deviation degree, which can more accurately reflect the current safety risk degree of the energy storage device.

[0059] As an example, obtain the current environmental characteristic data, which may include the current environmental temperature, the current environmental temperature change rate, the current target volatile gas concentration, the current target volatile gas concentration change rate, the current target smoke particle concentration change rate, and the current target smoke particle concentration. The current environmental characteristic data directly reflects the state of physical quantities in the environment where the energy storage device is located that may indicate potential fire hazards. A predefined algorithm or formula can be used to convert the current environmental characteristic data into a preliminary pyrolysis particle index, providing a basic reference value for subsequent correction. And based on the safety state deviation degree, a preset correction algorithm is used to correct the initial pyrolysis particle index to obtain the target pyrolysis particle index at the current moment. By combining multi-dimensional data (including environmental characteristic data and the safety state deviation degree of the overall device), a more comprehensive and accurate safety risk assessment index, namely the target pyrolysis particle index, is provided. Obtaining the target pyrolysis particle index can not only reflect the potential danger level in the current environment but also take into account the possible future development trend of the energy storage device, thus supporting maintenance personnel to make more scientific and reasonable maintenance and emergency response decisions, significantly improving the warning ability and safety of the energy storage safety warning system based on deep learning.

[0060] In some embodiments, determining the initial pyrolysis particle index at the current moment according to the current environmental characteristic data includes: Determine the current time-sharing stage category at the current moment according to the current environmental temperature change rate; According to the first correspondence between the preset time-sharing stage category and the corresponding relationship table, determine the target corresponding relationship table corresponding to the current time-sharing stage category. The target corresponding relationship table includes the corresponding relationship between various parameters and weighting coefficients; According to the target corresponding relationship table, determine the weighting coefficients of various parameters in the current environmental characteristic data; Based on the weighting coefficients of various parameters in the current environmental characteristic data and the current environmental characteristic data, determine the initial pyrolysis particle index.

[0061] Specifically, the various parameters in the current environmental characteristic data include: the current environmental temperature, the current environmental temperature change rate, the current target volatile gas concentration, the current target volatile gas concentration change rate, the current target smoke particle concentration change rate, and the current target smoke particle concentration. The time-sharing stage category is a monitoring stage dynamically divided according to the temperature change rate, reflecting different levels of pyrolysis risk. The time-sharing stage category can include the steady state, abnormal temperature rise, and violent decomposition stages. For example, in the steady state stage, the equipment operates normally with small temperature fluctuations, and the weights corresponding to various parameters can be evenly distributed at this time; in the abnormal temperature rise stage: the equipment shows an abnormal temperature rise trend, there may be potential safety hazards, and the temperature change is dominant. At this time, a higher weight (such as 50%) can be given to the temperature change rate weight; in the violent decomposition stage, the equipment undergoes a violent chemical reaction, gases and particles are violently released, and the safety risk is extremely high. Higher weights can be given to the target volatile gas concentration, the target volatile gas concentration change rate, and the target smoke particle concentration change rate. According to the characteristics of different time-sharing stages, there are different corresponding relationship tables for different time-sharing stage categories. The corresponding relationship is a table containing various parameters of environmental characteristic data (such as temperature, gas concentration, particle concentration, etc.) and their weighting coefficients, and the weight distribution is dynamically adjusted according to different time-sharing stage categories.

[0062] As an example, determine the current environmental temperature change rate, and map the current environmental temperature change rate to the corresponding time-sharing stage category according to the preset threshold range. For example, if the environmental temperature change rate is low (such as less than 0.1 °C / min), determine that the current time-sharing stage category at the current moment is the steady state stage; if the temperature change rate is high (such as 0.1 °C / min to 1 °C / min), determine that the current time-sharing stage category at the current moment is the abnormal temperature rise stage; and when the temperature change rate is extremely high (such as >1 °C / min), determine that the current time-sharing stage category at the current moment is the violent decomposition stage. Determine the time-sharing stage category at the current moment according to the current environmental temperature change rate, identify which monitoring stage the energy storage device is currently in, and provide a basis for subsequent analysis. Specifically, the following table can represent the corresponding relationship between the time-sharing stage category and the weighting coefficients of various parameters in the current environmental characteristic data

[0063] According to the first correspondence relationship between the preset time-sharing stage category and the corresponding relationship table for the current time-sharing stage category, search for and determine the target corresponding relationship table corresponding to the current time-sharing stage category, which is used to indicate the selection of the weighting coefficients of various subsequent parameters. In the case of determining the target corresponding relationship table, the weighting coefficients of the current various parameters applicable to the current time-sharing stage can be extracted from the target corresponding relationship table, and the weighting coefficients of the current various parameters are combined with the current ambient temperature, the current ambient temperature change rate, the current target volatile gas concentration, the current target volatile gas concentration change rate, the current target smoke particle concentration change rate, and the current target smoke particle concentration. The initial pyrolysis particle index can be calculated by weighted summation. For example, when the current ambient temperature is 78 °C, the current ambient temperature change rate is 1.5 °C / min, the current target volatile gas concentration is 350 ppm, the current target volatile gas concentration change rate is 40 ppm / min, the current target smoke particle concentration change rate is 70 (μg / m³) / min, and the current target smoke particle concentration is 480 μg / m³, the initial pyrolysis particle index is calculated as 3610 by weighted summation in combination with the weighting coefficients of various parameters of the current ambient characteristic data.

[0064] In some embodiments, the initial pyrolysis particle index is corrected based on the safety state deviation degree to obtain the target pyrolysis particle index at the current moment, including: If the safety state deviation degree is less than the first deviation threshold, the initial pyrolysis particle index is determined as the target pyrolysis particle index; If the safety state deviation degree is greater than or equal to the first deviation threshold and less than the second deviation threshold, the safety state deviation degree and the initial pyrolysis particle index are weighted and summed to obtain a first summation result, and the first summation result is determined as the target pyrolysis particle index, and the second deviation threshold is greater than the first deviation threshold; If the safety state deviation degree is greater than or equal to the second deviation threshold, the safety state deviation degree and the initial pyrolysis particle index are weighted and summed to obtain a second summation result, and the numerical values of the second summation result and the preset pyrolysis particle index are compared, and the larger of the two numerical values is determined as the target pyrolysis particle index.

[0065] Specifically, the safety state deviation degree is a comprehensive index obtained by calculating the difference between the predicted values of the energy storage device operation state data and the environmental characteristic data and the preset reference range, and can be used to measure the current safety risk level of the device. The first deviation threshold is the low-risk critical value of the set safety state deviation degree, which can indicate that the state of the energy storage device is slightly abnormal and needs attention but does not require immediate intervention. The second deviation threshold is the high-risk critical value of the set safety state deviation degree, which can indicate that the energy storage device is approaching a dangerous state. The preset pyrolysis particle index can be the high-risk critical value of the empirically set pyrolysis particle index. For example, the preset pyrolysis particle index can be 4000. When the pyrolysis particle target index exceeds the preset pyrolysis particle index, the highest-level warning needs to be directly triggered.

[0066] As an example, the safety state deviation degree can reflect the degree of difference between the predicted values of the energy storage device operation state data and the environmental characteristic data and the preset reference range. If the safety state deviation degree is less than the first deviation threshold (for example, 0.2), it can be considered that the energy storage device is in a normal or near-normal operating state. At this time, no additional correction is required for the initial pyrolysis particle index, and the initial pyrolysis particle index is directly determined as the pyrolysis particle target index. If the safety state deviation degree is greater than or equal to the first deviation threshold and less than the second deviation threshold (for example, 0.6), it can be considered that the energy storage device has certain safety hazards but has not reached a serious level. At this time, appropriate correction is required for the initial pyrolysis particle index, and the safety state deviation degree and the initial pyrolysis particle index are weighted and summed to obtain the first summation result. Specifically, the first summation result = the initial pyrolysis particle index * (1 + the safety state deviation degree), and the first summation result is determined as the pyrolysis particle target index. If the safety state deviation degree is greater than or equal to the second deviation threshold, it can be considered that the energy storage device has serious safety hazards. At this time, the safety state deviation degree and the initial pyrolysis particle index are weighted and summed to obtain the second summation result, and the values of the second summation result and the preset pyrolysis particle index are compared, and the larger one of them is selected as the pyrolysis particle target index. If the preset pyrolysis particle index is 3000, the safety state deviation degree and the initial pyrolysis particle index are weighted and summed to obtain the second summation result of 2800, and the larger value of the two is taken as the final pyrolysis particle target index: the pyrolysis particle target index = max (the second summation result, the preset pyrolysis particle index) = max(2800, 3000) = 3000. Ensure that in high-risk situations, the pyrolysis particle target index can fully reflect the potential severity, so as to issue a higher-level warning signal in a timely manner. By dynamically adjusting the pyrolysis particle index according to the safety state deviation degree, combined with different levels of the safety state deviation degree, the pyrolysis particle target index at the current moment is obtained, providing a more accurate and comprehensive safety risk assessment.

[0067] In some embodiments, determining a warning method for an energy storage device at the current moment according to the pyrolysis particle target index includes: Obtaining a first warning threshold and a second warning threshold, where the second warning threshold is greater than the first warning threshold; Obtaining an energy storage device image, which includes the energy storage device; When the pyrolysis particle target index is greater than or equal to the first warning threshold and less than the second warning threshold, determining the warning method as the first warning method, and controlling to display a first warning message at the user end, where the first warning message includes the pyrolysis particle target index and the energy storage device image; When the pyrolysis particle target index is greater than or equal to the second warning threshold, determining the warning method as the second warning method, controlling to display a second warning message at the user end and controlling the user end to emit a warning sound for prompting the second warning method, and controlling the energy storage device to emit a warning sound for prompting the second warning method, where the second warning message includes the pyrolysis particle target index and the energy storage device image.

[0068] Specifically, the first warning threshold is the low-risk warning critical value (such as 1000) of the preset pyrolysis particle index, which can indicate that there are potential risks in the energy storage device. At this time, the user can be prompted to pay attention but no immediate operation is required. The second warning threshold is the high-risk warning critical value (such as 3000) of the preset pyrolysis particle index. When the pyrolysis particle index is greater than the second warning threshold, it means that the device is close to or in a dangerous state and immediate intervention is required to trigger an emergency response (such as sound and light alarm, device shutdown). The first warning method is a low-risk warning mode, where the risk information and the real-time image of the device are displayed at the user end, and the sound alarm at the device end is not triggered. It can be used to remind the operation and maintenance personnel to check the device status. For example, perform a standardized inspection, including: checking the connection status of the battery pack (using infrared thermal imaging to assist in locating abnormal temperature rise points); detecting whether the load current exceeds the limit (comparing with historical operating conditions data); tightening loose joints (using a torque wrench for standardized operation); after the hidden danger is eliminated, the system automatically resumes monitoring to form a safety closed-loop. The second warning method is a high-risk warning mode, where the emergency information is displayed at the user end and a warning sound is emitted, and at the same time, the sound alarm is started at the device end, which can force attention or immediately take emergency measures such as shutdown. The energy storage device image is an image captured by a camera connected to the energy storage device. The camera is connected to the energy storage device, and the energy storage device is within the shooting range of the camera. The energy storage device is photographed regularly at a preset frequency to obtain the image data of the energy storage device.

[0069] As an example, when the pyrolysis particle target index is less than the first warning threshold, there is no need to take a warning method, and the continuous monitoring state is maintained. When the pyrolysis particle target index is greater than or equal to the first warning threshold and less than the second warning threshold, it is determined that the warning method is the first warning method, and the first warning information is controlled to be displayed on the user side. The first warning information includes the pyrolysis particle target index and the image data of the energy storage device. For example, on the user side interface, it is displayed as "Current pyrolysis particle target index: 1500, please pay attention to the device operation status", and the real-time image of the energy storage device is attached. Through visual cues, it can attract the attention of users and prompt the operation and maintenance personnel to further check the device status. When the pyrolysis particle target index is greater than or equal to the second warning threshold, it is determined that the warning method is the second warning method, and the second warning information is controlled to be displayed on the user side. The second warning information includes the pyrolysis particle target index (such as 3400) and the image data of the energy storage device. The image data of the energy storage device included in the second warning information is the image data collected by the most recent shooting. In addition, more obvious visual cues can also be displayed, such as a red alarm sign or a flashing icon, to enhance the warning effect. In addition to visual cues, a specific warning sound is played on the user side, such as a continuous beeping sound or a voice prompt "Emergency warning: The pyrolysis particle target index is too high, please check the device immediately". At the same time, the energy storage device is controlled to emit a warning sound in the second warning method, such as a continuous beeping sound or a voice prompt "Device anomaly, please handle it immediately", so that even if the user is not near the user side, the warning information can be received in a timely manner. Through the hierarchical warning mechanism, combined with the device status information and multiple reminder means, the comprehensive monitoring and timely response to the safety status of the energy storage device are realized. The first warning method provides a mild but clear reminder to identify potential problems at an early stage. The second warning method uses strong visual and auditory signals to quickly attract the attention of users and prompt them to take actions in high-risk situations. Through the multi-level warning strategy, not only the intelligent level and safety of the system are improved, but also the risk of fire accidents of the energy storage device can be effectively reduced, ensuring the stable operation of the energy storage device.

[0070] After determining that the warning method is the first warning method, it further includes: Generating a first control instruction, and based on the first control instruction, controlling the ventilation device connected to the energy storage device to turn on; and, Generating a second control instruction, and based on the second control instruction, controlling the load of the energy storage device to migrate from the energy storage device to the standby line.

[0071] Specifically, after determining that the early warning method is the first early warning method, a first control instruction is generated to activate and start the ventilation equipment connected to the energy storage device. The first control instruction may include specific parameters for turning on the ventilation equipment, such as wind speed, running time, etc. Specifically, the first control instruction can be sent to the ventilation equipment controller through a non-polar two-wire communication method to instruct it to start immediately. By turning on the ventilation equipment, it helps to quickly reduce the concentration of volatile gases and temperature in the surrounding environment of the energy storage device, thereby reducing the risk of fire or other safety accidents.

[0072] In addition, a second control instruction is generated to control the load of the energy storage device to migrate from the energy storage device to the standby line, reducing the working load of the energy storage device and avoiding more serious problems caused by overload. Specifically, determine the current load status of the energy storage device, including key parameters such as current and power, to judge whether load transfer is required. Based on the first early warning method and the load assessment result, a second control instruction is generated to instruct to switch part or all of the load from the energy storage device to the standby line. The generated second control instruction may include a detailed load distribution plan to ensure a smooth transition and not affect normal power supply. The system can execute the second control instruction through the controller to gradually transfer the load from the energy storage device to the standby line. During the execution process, non-critical loads can be cut off first, and the load ratio of the standby line can be gradually increased. Through the coordinated control of ventilation and load migration, it helps to block the vicious cycle of temperature and current, extend the safe window period of the equipment, multi-line load balancing can improve power supply reliability, effectively reduce the pressure on the energy storage device, avoid failures or damages caused by overload, improve the working environment of the energy storage device, and reduce the potential risk of combustion or explosion. Through the above multi-level early warning and control strategies, the intelligent level and safety of the system can be improved, the response ability to emergencies can be significantly enhanced, and it is ensured that the energy storage device can operate stably for a long time.

[0073] In some embodiments, the operating state data includes voltage and current, and the environmental characteristic data includes environmental temperature, target volatile gas concentration, and target smoke particle concentration. Inputting the historical operating state data, historical environmental characteristic data, current operating state data, and current environmental characteristic data into the data prediction network includes: Using the Kalman filter algorithm to smooth the historical operating state data, historical environmental characteristic data, current operating state data, and current environmental characteristic data to obtain a filtered data sequence; Normalize various parameters in the filtered data sequence to obtain a normalized data sequence, which includes a voltage data normalization sequence, a current data normalization sequence, an environmental temperature data normalization sequence, a target volatile gas concentration data normalization sequence, and a target smoke particle concentration data normalization sequence; Extract features from the normalized data sequence to generate short-term features, long-term features, and cross-category correlation features; Determine the normalized data sequence, short-term features, long-term features, and cross-category correlation features as the input sequence, and input the input sequence into the data prediction network.

[0074] Specifically, it includes the voltage and current of the energy storage device, which can reflect the electrical operating state of the device. The environmental feature data includes environmental temperature, target volatile gas concentration (such as the gas generated by thermal decomposition), and target smoke particle concentration (such as micron-sized particles), which can reflect the state of physical quantities in the environment where the energy storage device is located that may indicate potential fire hazards. The short-term features are features extracted from the data within a short period of time (such as a few seconds), which can reflect instantaneous state changes, and the long-term features are features extracted from the data within a long period of time (such as a few minutes), which can reflect trend changes. The cross-category correlation features are correlation features between different categories of data. For example, there may be an association relationship between voltage and environmental temperature.

[0075] As an example, the Kalman filter algorithm can be used to perform noise suppression and trend smoothing on the first preset historical operating state data, the current operating state data, the first preset historical environmental characteristic data, and the current environmental characteristic data, removing noise and smoothing the time series data. By filtering the historical and current data, the influence of sensor noise and other interference factors can be reduced, thereby obtaining a more accurate data sequence, i.e., the filtered data sequence. Normalize various parameters in the filtered data sequence. Calculate the mean and standard deviation of various parameters based on a sliding window (such as 24 hours), which can be updated every 30 minutes. Apply the normalization formula to convert data with different dimensions to the same scale range (such as [0,1] or [-1,1]) to generate a normalized data sequence, which helps each feature to have the same importance during subsequent model training. Specifically, the voltage, current, environmental temperature, target volatile gas concentration, and target smoke particle concentration can be normalized respectively, so that the data can be compared and analyzed under the same scale, eliminating the influence brought by dimensional differences, and improving the stability and efficiency of model training. Perform multi-scale feature extraction on the normalized data sequence (the normalized data sequence) to generate short-term features, long-term features, and cross-category correlation features. Short-term features are statistical features extracted from the data within a short period (such as a few seconds), such as mean, variance, etc., which can characterize instantaneous state changes; long-term features are trend features extracted within a longer time period (such as within a few minutes), such as slope, periodic fluctuations, etc., which are used to characterize long-term behavior patterns; cross-category correlation features are correlation features between different category data, such as the correlation between voltage and environmental temperature, which helps to discover potential complex relationships. The short-term features, long-term features, and cross-category correlation features together constitute a more comprehensive data description, providing richer information input for the data prediction network. Integrate the normalized data sequence, short-term features, long-term features, and cross-category correlation features into a complete input sequence to obtain the above input sequence, and input the input sequence into the data prediction network. Through the above multi-dimensional data fusion, making full use of all available information, it helps to enhance the adaptability of the energy storage safety warning method based on deep learning provided by this application to complex application scenarios, helps the energy storage safety warning system based on deep learning to identify potential safety hazards at an early stage, and take corresponding warning measures, thereby significantly improving the safety and reliability of energy storage devices, solving the problems of warning lag, high false alarm rate, and inability to fuse multi-dimensional data existing in traditional technologies, and achieving more accurate and timely safety monitoring.

[0076] In some embodiments, the historical environmental characteristic data and the current environmental characteristic data are obtained through the following steps: Obtain the material parameters of the energy storage device, and determine the target volatile gas concentration and target smoke particles according to the material parameters of the energy storage device; Obtain the historical ambient temperature before the current moment and the current ambient temperature at the current moment collected by the temperature sensor, and determine the current ambient temperature change rate based on the historical ambient temperature and the current ambient temperature; Obtain the historical target volatile gas concentration before the current moment and the current target volatile gas concentration at the current moment collected by the electrochemical sensor, and determine the current target volatile gas concentration change rate based on the historical target volatile gas concentration and the current target volatile gas concentration; Obtain the historical target smoke particle concentration before the current moment and the current target smoke particle concentration at the current moment collected by the laser particle sensor, and determine the current target smoke particle concentration change rate based on the historical target smoke particle concentration and the current target smoke particle concentration; Determine the historical ambient temperature, historical target volatile gas concentration, and historical target smoke particle concentration as historical ambient characteristic data; Determine the current ambient temperature, current ambient temperature change rate, current target volatile gas concentration, current target volatile gas concentration change rate, current target smoke particle concentration change rate, and current target smoke particle concentration as current ambient characteristic data.

[0077] Specifically, the material parameters of the energy storage device are the physical and chemical properties of the materials used in the energy storage device, such as insulation materials, electrolyte components, etc. The material parameters of the energy storage device directly determine the types of target volatile gases and smoke particles that may be released when the device decomposes due to heat. The target volatile gas concentration is the concentration of specific gases (such as hydrogen fluoride, carbon monoxide, etc.) released due to the thermal decomposition of the energy storage device materials, which can be measured by an electrochemical sensor. The target smoke particle concentration is the concentration of micron-sized or nano-sized smoke particles generated by the thermal decomposition of the energy storage device materials, which can be measured by a laser particle sensor.

[0078] As an example, according to the design document or material specification of the energy storage device, the parameters of its main constituent materials (such as the cathode material, anode material, electrolyte composition, etc.) can be obtained, and the types of gases that may be released and the characteristics of smoke particles during the thermal decomposition of the materials can be analyzed to determine the specific categories of the target volatile gases and target smoke particles. For example, the electrolyte of a lithium battery may decompose to produce volatile gases such as hydrogen and carbon monoxide, along with micron-sized smoke particles. The ambient temperature is one of the important indicators for evaluating the operating state of the energy storage device. Obtain the first preset number of historical ambient temperatures before the current moment and the current ambient temperature collected by the temperature sensor. Specifically, it includes: extracting the ambient temperature data in the past period (such as the past 10 minutes) from the database, and collecting the ambient temperature value at the current moment in real time through the temperature sensor. Based on the historical ambient temperature and the current ambient temperature, calculate the temperature change rate per unit time to obtain the current ambient temperature change rate. The formula for calculating the current ambient temperature change rate is as follows: current ambient temperature change rate = (current ambient temperature - average temperature of the previous period) / time interval. The ambient temperature change rate can reflect the trend of temperature rise or fall and help determine whether there is an abnormal temperature rise phenomenon. At the same time, the concentration of the target volatile gas directly reflects the degree of thermal decomposition of the internal materials of the energy storage device. Obtain the first preset number of historical target volatile gas concentrations before the current moment and the current target volatile gas concentration at the current moment collected by the electrochemical sensor. Specifically, it includes: extracting the target volatile gas concentration data in the past period from the database, and collecting the current target volatile gas concentration in real time through the electrochemical sensor. Based on the historical target volatile gas concentration and the current target volatile gas concentration, calculate the concentration change rate per unit time to obtain the current ambient temperature change rate. The concentration of the target smoke particles reflects the thermal decomposition situation of the materials of the energy storage device. Obtain the first preset number of historical target smoke particle concentrations before the current moment and the current target smoke particle concentration at the current moment collected by the laser particle sensor. Specifically, it includes: extracting the target smoke particle concentration data in the past period from the database, and collecting the current target smoke particle concentration value at the current moment in real time through the laser particle sensor. Based on the historical target smoke particle concentration and the current target smoke particle concentration, calculate the concentration change rate per unit time to obtain the current target smoke particle concentration change rate. Determine the first preset number of historical ambient temperature, the first preset number of historical target volatile gas concentrations, and the first preset number of historical target smoke particle concentrations as the first preset number of historical ambient characteristic data, and determine the current ambient temperature, the current ambient temperature change rate, the current target volatile gas concentration, the current target volatile gas concentration change rate, the current target smoke particle concentration change rate, and the current target smoke particle concentration as the current ambient characteristic data.

[0079] By collecting historical and current ambient temperatures, target volatile gas concentrations, and target smoke particle concentrations, and calculating their rates of change, the dynamic change trend of the environmental state can be comprehensively captured. The target volatile gas concentration and the target smoke particle concentration can directly reflect the degree of pyrolysis of the energy storage device materials. Combining with the rate of change of the ambient temperature, the safety state of the device can be evaluated more accurately. Through the collection and analysis of multi-dimensional data, a comprehensive environmental feature dataset is constructed, significantly enhancing the early hidden danger prediction ability and intelligent level of the energy storage safety warning system based on deep learning, and helping to solve the problems of warning lag, high false alarm rate, and inability to fuse multi-dimensional data existing in traditional technologies.

[0080] In some embodiments, obtaining the first warning threshold and the second warning threshold includes: Obtaining a first preset warning initial value; According to the current ambient temperature and the second mapping relationship, determining the current first correction amount corresponding to the current ambient temperature, where the second mapping relationship is used to represent the mapping relationship between the ambient temperature and the first correction amount, and the ambient temperature and the first correction amount are positively correlated; According to the current first correction amount, correcting the first preset warning initial value to obtain the first warning threshold; Obtaining a second preset warning initial value; According to the current ambient temperature and the third mapping relationship, determining the current second correction amount corresponding to the current ambient temperature, where the third mapping relationship is used to represent the mapping relationship between the ambient temperature and the second correction amount, and the ambient temperature and the second correction amount are positively correlated; According to the current second correction amount, correcting the second preset warning initial value to obtain the second warning threshold.

[0081] It should be noted that the above examples are only for understanding the present application and do not constitute a limitation on the energy storage safety warning method based on deep learning of the present application. More simple transformations in various forms based on this technical concept are within the protection scope of the present application.

[0082] The present application also provides an energy storage safety warning system based on deep learning. Please refer to Figure 2 , the energy storage safety warning system based on deep learning includes: A multi-source data acquisition module 201, configured to acquire historical operation state data and historical environmental feature data of the energy storage device at a historical moment, and acquire current operation state data and current environmental feature data of the energy storage device at a current moment; The prediction module 202 is configured to input historical operation status data, historical environment feature data, current operation status data, and current environment feature data into a data prediction network, and obtain the predicted operation status data and predicted environment feature data of the energy storage device at a future moment generated by the data prediction network; the data prediction network is obtained by training a long short-term memory network (LSTM) based on a training sample set. The deviation determination module 203 is configured to determine the safety status deviation of the energy storage device at the current moment according to the predicted operation status data and the predicted environment feature data. The index determination module 204 is configured to determine the pyrolysis particle target index at the current moment according to the current environment feature data and the safety status deviation. The early warning determination module 205 is configured to determine the early warning method of the energy storage device at the current moment according to the pyrolysis particle target index.

[0083] The energy storage safety early warning system based on deep learning provided by this application adopts the energy storage safety early warning method based on deep learning in the above embodiment, and can solve the technical problem of low accuracy of early warning monitoring for energy storage devices in the prior art. Compared with the prior art, the beneficial effects of the energy storage safety early warning system based on deep learning provided by this application are the same as those of the energy storage safety early warning method based on deep learning provided by the above embodiment, and other technical features in the energy storage safety early warning system based on deep learning are the same as the features disclosed in the method of the above embodiment, and will not be elaborated here.

[0084] This application provides an energy storage safety early warning system based on deep learning. The energy storage safety early warning system based on deep learning includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the energy storage safety early warning method based on deep learning in the first embodiment above.

[0085] Next, refer to Figure 3 , which shows a schematic structural diagram of an energy storage safety early warning system based on deep learning suitable for implementing the embodiments of this application. Figure 3 The shown energy storage safety early warning system based on deep learning is only an example, and should not bring any limitation to the functions and usage scopes of the embodiments of this application.

[0086] As Figure 3As shown, the energy storage safety warning system based on deep learning may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in the Read Only Memory (ROM) 1002 or the program loaded from the storage device 1003 into the Random Access Memory (RAM) 1004. In the random access memory 1004, various programs and data required for the operation of the energy storage safety warning system based on deep learning are also stored. The processing device 1001, the read only memory 1002, and the random access memory 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the input / output interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the energy storage safety warning system based on deep learning to communicate with other devices wirelessly or wiredly to exchange data. Although the figure shows an energy storage safety warning system with various systems, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be implemented or had alternatively.

[0087] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device, or installed from the storage device 1003, or installed from the read only memory 1002. When the computer program is executed by the processing device 1001, the above functions defined in the methods of the embodiments disclosed in the present application are executed.

[0088] The energy storage safety warning system based on deep learning provided by this application adopts the energy storage safety warning method based on deep learning in the above-mentioned embodiment, and can solve the technical problem of low accuracy of warning and monitoring for energy storage devices in the prior art. Compared with the prior art, the beneficial effects of the energy storage safety warning system based on deep learning provided by this application are the same as those of the energy storage safety warning method based on deep learning provided by the above-mentioned embodiment, and other technical features in the energy storage safety warning system based on deep learning are the same as those disclosed in the method of the previous embodiment, which will not be elaborated here.

[0089] It should be understood that each part disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0090] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0091] This application provides a computer-readable storage medium with computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the energy storage safety warning method based on deep learning in the above-mentioned embodiment.

[0092] The computer-readable storage medium provided by the present application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. The program code contained on the computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0093] The above computer-readable storage medium can be included in the energy storage safety warning system based on deep learning; it can also exist independently and not be assembled into the energy storage safety warning system based on deep learning.

[0094] The above computer-readable storage medium carries one or more programs. When the one or more programs are executed by the energy storage safety warning system based on deep learning, the energy storage safety warning system based on deep learning is enabled to: obtain the historical operating state data and historical environmental characteristic data of the energy storage device at a historical moment, and obtain the current operating state data and current environmental characteristic data of the energy storage device at the current moment; input the historical operating state data, historical environmental characteristic data, current operating state data, and current environmental characteristic data into a data prediction network to obtain the predicted values of the operating state data and environmental characteristic data of the energy storage device at a future moment generated by the data prediction network; the data prediction network is obtained by training a long short-term memory network (LSTM) based on a training sample set; determine the safety state deviation degree of the energy storage device at the current moment according to the predicted values of the operating state data and environmental characteristic data; determine the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the safety state deviation degree; and determine the warning method of the energy storage device at the current moment according to the pyrolysis particle target index.

[0095] Computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The above-mentioned programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or it can be connected to an external computer (for example, by connecting through an Internet service provider using the Internet).

[0096] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of the code, and this module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutively represented blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0097] The modules involved in the embodiments described in this application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation to the unit itself in some cases.

[0098] The readable storage medium provided by this application is a computer-readable storage medium. The computer-readable storage medium stores computer-readable program instructions (i.e., computer programs) for performing the above-mentioned energy storage safety warning method based on deep learning, and can solve the technical problem of low accuracy of warning monitoring for energy storage devices in the prior art. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by this application are the same as those of the energy storage safety warning method based on deep learning provided in the above embodiments, and will not be elaborated here.

[0099] The present application also provides a computer program product, including a computer program which, when executed by a processor, implements the steps of the above-mentioned energy storage safety warning method based on deep learning.

[0100] The computer program product provided by the present application can solve the technical problem of low accuracy of warning and monitoring for energy storage devices in the prior art. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as those of the energy storage safety warning method based on deep learning provided in the above embodiments, and will not be elaborated here.

[0101] The above are only partial embodiments of the present application, and do not limit the patent scope of the present application accordingly. Any equivalent structural transformation made by using the content of the specification and drawings of the present application under the technical concept of the present application, or any direct / indirect application in other related technical fields shall be included within the patent protection scope of the present application.

Claims

1. A storage energy safety early warning method based on deep learning, characterized in that: The energy storage safety warning based on deep learning includes: Obtaining historical operating status data and historical environmental characteristic data of the energy storage device at historical moments, and obtaining current operating status data and current environmental characteristic data of the energy storage device at the current moment; Input the historical operating state data, the historical environmental characteristic data, the current operating state data and the current environmental characteristic data into a data prediction network to obtain the predicted operating state data and the predicted environmental characteristic data of the energy storage device at a future time generated by the data prediction network; the data prediction network is obtained by training a long short-term memory network LSTM based on a training sample set; Determining the safety state deviation of the energy storage device at the current moment according to the predicted amount of the operating state data and the predicted amount of the environmental characteristic data; Determining the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the safety state deviation; According to the pyrolysis particle target index, the early warning mode of the energy storage device at the current moment is determined.

2. The energy storage safety early warning method based on deep learning according to claim 1, characterized in that: The step of determining the safety state deviation of the energy storage device at the current moment according to the predicted amount of the operating state data and the predicted amount of the environmental characteristic data includes: Obtaining a preset operating status data reference range and a preset environmental characteristic data reference range corresponding to the energy storage device; Calculating the difference between the predicted running state data and the preset running state data reference range to obtain the running state data deviation; Calculating the difference between the predicted amount of environmental characteristic data and the preset reference range of environmental characteristic data to obtain an environmental characteristic data deviation; The safety state deviation is determined according to the operating state data deviation and the environmental characteristic data deviation.

3. The energy storage safety early warning method based on deep learning according to claim 1, characterized in that: The step of determining the pyrolysis particle target index at the current moment according to the current environmental characteristic data and the safety state deviation degree includes: Determine the initial index of pyrolysis particles at the current moment according to the current environmental characteristic data; The pyrolysis particle initial index is corrected based on the safety state deviation to obtain the pyrolysis particle target index at the current moment.

4. The energy storage safety early warning method based on deep learning according to claim 3 is characterized in that: The determining the initial index of pyrolysis particles at the current moment according to the current environmental characteristic data includes: Determine the current time-sharing stage category at the current moment according to the current ambient temperature change rate; Determine a target corresponding relationship table corresponding to the current time-sharing stage category according to a preset first corresponding relationship between the time-sharing stage category and the corresponding relationship table, wherein the target corresponding relationship table includes corresponding relationships between various parameters and weighting coefficients; Determine the weighting coefficients of various parameters in the current environmental feature data according to the target correspondence table; The pyrolysis particle initial index is determined based on the weighted coefficients of various parameters of the current environmental characteristic data and the current environmental characteristic data, wherein the current environmental data includes the current environmental temperature change rate.

5. The energy storage safety early warning method based on deep learning according to claim 3, characterized in that: The step of correcting the pyrolysis particle initial index based on the safety state deviation to obtain the pyrolysis particle target index at the current moment includes: If the safety state deviation is less than a first deviation threshold, determining the pyrolysis particle initial index as a pyrolysis particle target index; If the safety state deviation is greater than or equal to the first deviation threshold and less than the second deviation threshold, weighted summing the safety state deviation and the pyrolysis particle initial index is performed to obtain a first summation result, and the first summation result is determined as the pyrolysis particle target index, and the second deviation threshold is greater than the first deviation threshold; If the safety state deviation is greater than or equal to the second deviation threshold, the safety state deviation and the pyrolysis particle initial index are weightedly summed to obtain a second summation result, and the second summation result is compared with the value of the preset pyrolysis particle index, and the larger value is determined as the pyrolysis particle target index.

6. The energy storage safety early warning method based on deep learning according to claim 1, characterized in that: Determining the early warning mode of the energy storage device at the current moment according to the pyrolysis particle target index includes: Acquire a first warning threshold and a second warning threshold, wherein the second warning threshold is greater than the first warning threshold; Acquire an energy storage device image, wherein the energy storage device image includes the energy storage device; When the pyrolysis particle target index is greater than or equal to the first warning threshold and less than the second warning threshold, determining that the warning mode is the first warning mode, and controlling the display of the first warning information on the user terminal, wherein the first warning information includes the pyrolysis particle target index and the energy storage device image; When the pyrolysis particle target index is greater than or equal to the second warning threshold, the warning mode is determined to be the second warning mode, the second warning information is controlled to be displayed on the user terminal, the user terminal is controlled to emit a warning sound for prompting the second warning mode, and the energy storage device is controlled to emit a warning sound for prompting the second warning mode, and the second warning information includes the pyrolysis particle target index and the energy storage device image.

7. The energy storage safety early warning method based on deep learning according to claim 6, characterized in that: After determining that the warning mode is the first warning mode, the method further includes: generating a first control instruction, and controlling the ventilation device connected to the energy storage device to open based on the first control instruction; and, A second control instruction is generated, and based on the second control instruction, the load of the energy storage device is controlled to migrate from the energy storage device to the backup line.

8. The energy storage safety early warning method based on deep learning according to claim 1, characterized in that: The operating state data includes voltage and current, the environmental characteristic data includes environmental temperature, target volatile gas concentration and target smoke particle concentration, and the historical operating state data, the historical environmental characteristic data, the current operating state data and the current environmental characteristic data are input into a data prediction network, including: Using a Kalman filter algorithm to smooth the historical operating state data, the historical environmental characteristic data, the current operating state data, and the current environmental characteristic data to obtain a filtered data sequence; Normalizing various parameters in the filtered data sequence to obtain a normalized data sequence, wherein the normalized data sequence includes a voltage data normalized sequence, a current data normalized sequence, an ambient temperature data normalized sequence, a target volatile gas concentration data normalized sequence, and a target smoke particle concentration data normalized sequence; Performing feature extraction on the normalized data sequence to generate short-term features, long-term features, and cross-category correlation features; The normalized data sequence, the short-term features, the long-term features and the cross-category association features are determined as an input sequence, and the input sequence is input into the data prediction network.

9. The energy storage safety early warning method based on deep learning according to claim 1, characterized in that: The historical environmental characteristic data and the current environmental characteristic data are obtained through the following steps: Acquiring material parameters of the energy storage device, and determining a target volatile gas concentration and a target smoke particle according to the material parameters of the energy storage device; Acquire the historical ambient temperature before the current moment and the current ambient temperature at the current moment collected by the temperature sensor, and determine the current ambient temperature change rate based on the historical ambient temperature and the current ambient temperature; Acquire a historical target volatile gas concentration before the current moment and a current target volatile gas concentration at the current moment collected by an electrochemical sensor, and determine a current target volatile gas concentration change rate based on the historical target volatile gas concentration and the current target volatile gas concentration; Acquire the historical target smoke particle concentration before the current moment and the current target smoke particle concentration at the current moment collected by the laser particle sensor, and determine the current target smoke particle concentration change rate based on the historical target smoke particle concentration and the current target smoke particle concentration; Determining the historical environmental temperature, the historical target volatile gas concentration, and the historical target smoke particle concentration as the historical environmental characteristic data; The current environment temperature, the current environment temperature change rate, the current target volatile gas concentration, the current target volatile gas concentration change rate, the current target smoke particle concentration change rate and the current target smoke particle concentration are determined as the current environment characteristic data.

10. A deep learning-based energy storage safety early warning system, characterized in that: The deep learning-based energy storage safety warning system includes: a memory, a processor, and a deep learning-based energy storage safety warning program stored in the memory and executable on the processor, wherein the deep learning-based energy storage safety warning program is configured to implement the steps of the deep learning-based energy storage safety warning method as described in any one of claims 1 to 9.

Images