Prediction method for thermal runaway of energy storage battery and related system

Abstract

The invention belongs to the field of out-of-control prediction, and discloses an energy storage battery thermal out-of-control prediction method and a related system, which comprehensively incorporates various sensor data such as temperature, voltage, current and the like into a model, and can more comprehensively describe the running state of a battery. Compared with the mode that only a single feature or a small number of features are adopted, the decision-making tree can reflect the health condition and the potential abnormal trend of the battery more accurately after the multi-source features are fused, and therefore the prediction accuracy of thermal runaway is improved. According to the method, the decision tree model is constructed by using long-time-span historical data, slicing processing can be carried out according to real-time data at the current moment and in a short period, the data are used as initial input, and the current data and the historical trend are combined, so that the decision tree model is constructed. According to the prediction method, the temperature change trend of the battery in a future period of time can be efficiently predicted in time, and the early warning delay is shortened.

Description

Technical Field

[0001] The present invention belongs to the field of out-of-control prediction, and particularly relates to a prediction method and related system for thermal runaway of energy storage batteries. Background Art

[0002] Energy storage batteries play an important role in the development of modern power systems and renewable energy, especially in coping with power demand fluctuations and optimizing energy utilization. With the instability of renewable energy (such as solar energy and wind energy) power generation, the demand for energy storage batteries is particularly important. They can store excess electrical energy and release it during peak demand, thus balancing the grid load and enhancing the stability of power supply. In addition, energy storage batteries are also widely used in electric vehicles, home energy storage systems, and industrial energy storage and other fields.

[0003] However, with the increasing application of energy storage batteries, their safety issues have attracted increasing attention, especially the phenomenon of thermal runaway. Thermal runaway refers to the rapid increase in temperature of the battery under conditions such as overcharge, over-discharge, short circuit, or physical damage, triggering self-heating and further deterioration, and ultimately may lead to battery fire or explosion. This process usually begins with the decomposition of the electrolyte inside the battery or the chemical reaction of the electrode material, and is accelerated by the release of a large amount of heat. Thermal runaway not only endangers the safety of equipment, but also poses a serious threat to the surrounding environment and personal safety. Therefore, it becomes particularly important to study how to detect the abnormal state of the battery in advance, issue a warning signal for battery thermal runaway, and reduce the occurrence of thermal runaway accidents.

[0004] Existing battery thermal runaway warning algorithms mainly analyze the relevant signal characteristics in the energy storage battery to determine whether the battery will undergo thermal runaway by whether they conform to the laws of relevant signal changes during the thermal runaway process. The analysis of temperature and temperature rise rate is an important judgment process and detection method in the process of battery thermal runaway detection. Due to various factors such as overcharge, over-discharge, and extrusion, the internal temperature of the battery gradually increases, and the process of temperature increase has a corresponding relationship with the thermal runaway stage of the battery. Application No. 202210052756.5 discloses a thermal runaway detection method based on establishing an early characteristic value equation of lithium battery thermal runaway based on the temperature and voltage change rates during the thermal runaway process of the lithium battery. In addition, battery gas is also one of the indicators for battery thermal runaway detection. During the thermal runaway process of the battery, , and other gases are generated. As time goes by, the , inside the battery will gradually increase. Application No. 202310259507.8 discloses a method and system for quantitatively calculating hydrogen after thermal runaway of lithium-ion batteries, which can be used to measure the change process of hydrogen inside the battery after thermal runaway, and an early warning of thermal runaway can be achieved by setting a detection threshold.

[0005] However, the existing methods do not make full use of the battery data itself. Using only temperature or gas will result in a high false alarm rate for the early warning system. When the sudden increase or decrease of the ambient temperature affects the internal temperature of the battery, the early warning system will issue an alarm when the battery does not experience thermal runaway. Similarly, due to the different usage times of batteries of different models and materials, the internal gas changes will also vary when thermal runaway occurs, making it difficult for some methods to adapt to the changing battery environment. Summary of the Invention

[0006] The object of the present invention is to overcome the problem that due to the inability to combine various signal characteristics, the thermal runaway of the battery cannot be efficiently and accurately predicted, and to provide a prediction method and related system for the thermal runaway of energy storage batteries.

[0007] To achieve the above object, the present invention adopts the following technical solutions: In the first aspect, the present invention provides a prediction method for the thermal runaway of energy storage batteries, including the following steps: Obtain the historical data of the temperature, voltage and current of the batteries in the battery pack, and construct a decision tree using the historical data; Obtain the current data of the temperature, voltage and current of the batteries in the battery pack, slice the obtained data to obtain the data at the current moment and the required time period, and use it as the initial data; Process the initial data using the decision tree to predict the temperature change data of the battery in the required future time period; Analyze the temperature change data of the battery in the required future time period, and judge whether the battery will experience thermal runaway in the required future time period according to the analysis result.

[0008] A further improvement of the present invention is that the specific method for obtaining the historical data of the temperature, voltage and current of the batteries in the battery pack and constructing a decision tree using the historical data is as follows: Obtain the objective function of the decision tree without leaves as the original objective function; Obtain the historical data of the temperature, voltage and current of the batteries in the battery pack, sort the historical data, traverse all possible decision tree structures in the historical data, and calculate the current objective function of all decision tree structures; Obtain the difference between the current objective function and the original objective function. If there is a positive value in the difference, select the decision tree structure corresponding to the largest positive value as the splitting node; if there is no positive value in the difference, do not perform splitting; Traverse all the initial data and split the decision tree to obtain the final decision tree.

[0009] A further improvement of the present invention lies in obtaining the temperature, voltage, and current data of the batteries in the battery pack, slicing the obtained data to obtain the data at the current moment and the required time period, and the specific method for obtaining the initial data is as follows: Obtain the temperature of the batteries in the battery pack as , the voltage as , and the current data as , where is the number of data samples; Slice the data samples, , is the data sample after slicing, is the data interval at the current moment and within the required time period The data sample after slicing is used as the initial data.

[0010] A further improvement of the present invention lies in using a decision tree to process the initial data and predicting the temperature change data of the battery within the required time period in the future. The specific method is as follows: Construct weakly classified decision trees, preset the penalty term for each weakly classified decision tree, and integrate all the weakly classified decision trees into a decision tree model; Input the initial data into the decision tree model to obtain the temperature change data of the battery within the required time period in the future.

[0011] A further improvement of the present invention lies in constructing weakly classified decision trees, preset the penalty term for each weakly classified decision tree, and the specific method for integrating all the weakly classified decision trees into a decision tree model is as follows: Construct weakly classified decision trees, and obtain the final predicted value as the output value of the decision tree model by summing the prediction results of each weakly classified decision tree:

[0012] where, is the final predicted value, is the th among the weakly classified decision trees, is:

[0013] where, is the number of leaves in the weakly classified decision tree, is the The weights of the leaf nodes, is the square penalty of the weights of the leaf nodes, is the weighted proportion of the number of leaves in the decision tree of the weak classification, is the weighted proportion of the sum of the weights of the leaf nodes; Preset the objective function of the th decision tree

[0014] Among them, represents the error of all temperature prediction values, represents the sum of the penalty terms of all weak classification decision trees, is the loss function, is the true value of the th sample, is the predicted value of the th sample; Simplify the objective function by using the second-order Taylor expansion and take the derivative to obtain the optimal node weights

[0015] Among them, is the first derivative of the loss function with respect to the predicted value, is the second derivative of the loss function with respect to the predicted value, is the set of all sample indices of the

[0016] A further improvement of the present invention lies in analyzing the temperature change data of the battery within the future required time period, and the specific method for judging whether the battery will have a thermal runaway within the future required time period according to the analysis results is as follows: Preset the alarm threshold and the number of consecutive early warnings; Obtain the temperature change data of the battery within the future required time period, and calculate the mean value of the temperature change data of the battery within the future required time period; Subtract the temperature change data of the battery within the future required time period from the mean value to obtain a sequence with a mean value of 0; Calculate the cumulative sum of the sequence and obtain the minimum value of the cumulative sum; Calculate the difference between all other cumulative sums and the minimum value of the cumulative sum, and compare all the differences with the alarm threshold. If the difference is greater than the alarm threshold, record it as an early warning; When the number of recorded early warnings is greater than the preset number of consecutive early warnings, it is determined that the battery is about to have a thermal runaway within the future required time period.

[0017] In a second aspect, the present invention provides a prediction system for energy storage battery thermal runaway, including: A decision tree construction module, which is used to obtain historical data of the temperature, voltage, and current of the batteries in the battery pack, and construct a decision tree using the historical data; An initial data acquisition module, which is used to obtain the current data of the temperature, voltage, and current of the batteries in the battery pack, slice the acquired data, and obtain the data at the current moment and the required time period as the initial data; A decision tree processing module, which is used to process the initial data using the decision tree and predict the temperature change data of the battery in the future required time period; A thermal runaway prediction module, which is used to analyze the temperature change data of the battery in the future required time period and determine whether the battery will experience thermal runaway in the future required time period according to the analysis results.

[0018] A further improvement of the present invention is that the function of the decision tree construction module is realized by the following method: Obtain the objective function of the decision tree without leaves as the original objective function; Obtain the historical data of the temperature, voltage, and current of the batteries in the battery pack, sort the historical data, traverse all possible decision tree structures in the historical data, and calculate the current objective functions of all decision tree structures; Obtain the difference between the current objective function and the original objective function. If there are positive values in the difference, select the decision tree structure corresponding to the largest positive value as the splitting node; if there are no positive values in the difference, no splitting is performed; Traverse all the initial data and split the decision tree to obtain the final decision tree.

[0019] A further improvement of the present invention is that the function of the initial data acquisition module is realized by the following method: Obtain that the temperature of the battery in the battery pack is , the voltage is , and the current data is , where is the number of data samples; Slice the data samples, , is the data sample after slicing, is the data interval at the current moment and within the required time period The data sample after slicing is used as the initial data.

[0020] A further improvement of the present invention is that the function of the decision tree processing module is realized by the following method: Construct weakly classified decision trees, preset the penalty term for each weakly classified decision tree, and integrate all the weakly classified decision trees into a decision tree model; Input the initial data into the decision tree model to obtain the temperature change data of the battery within the required future time period.

[0021] A further improvement of the present invention lies in that the function of the thermal runaway prediction module is implemented by the following method: Preset an alarm threshold and the number of consecutive early warnings; Obtain the temperature change data of the battery within the required future time period, and calculate the average value of the temperature change data of the battery within the required future time period; Subtract the temperature change data of the battery within the required future time period from the average value to obtain a sequence with a zero mean; Calculate the cumulative sum of the sequence and obtain the minimum value of the cumulative sum; Calculate the difference between all other cumulative sums and the minimum value of the cumulative sum, and compare all the differences with the alarm threshold. If the difference is greater than the alarm threshold, record it as an early warning; When the recorded number of early warnings is greater than the preset number of consecutive early warnings, it is determined that the battery is about to experience thermal runaway within the required future time period.

[0022] In a third aspect, the present invention provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of a prediction method for thermal runaway of an energy storage battery are implemented.

[0023] In a fourth aspect, the present invention provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of a prediction method for thermal runaway of an energy storage battery are implemented.

[0024] Compared with the prior art, the present invention has the following beneficial effects: The present invention comprehensively incorporates various sensor data such as temperature, voltage, and current into the model, enabling a more comprehensive characterization of the battery's operating state. Compared with using only a single or a small number of features, the decision tree can more accurately reflect the battery's health status and potential abnormal trends after fusing multi-source features, thereby improving the prediction accuracy of thermal runaway. The present invention not only utilizes historical data with a long time span to construct a decision tree model but also can perform slicing processing on real-time data at the current moment and in the short term as the initial input. By combining the current data and historical trends, this prediction method can efficiently predict the temperature change trend of the battery in the future for a period of time, shortening the warning delay. Due to its hierarchical partitioning method, the decision tree model of the present invention has relatively good interpretability of feature importance and decision paths. Maintenance personnel can understand which features (temperature, voltage, or current) have a greater impact on the prediction result based on the splitting nodes in the decision tree, facilitating subsequent targeted adjustment, troubleshooting, or optimization. The present invention simulates the temperature evolution trend of the battery in the future period. If there is a sudden increase in temperature or other signs that may trigger thermal runaway, a forward-looking risk warning can be achieved. For energy storage systems, taking countermeasures (such as load reduction, heat dissipation, shutdown, etc.) in advance against possible thermal runaway risks can effectively reduce the probability of fault escalation and ensure the overall safety of the system. In summary, the present invention realizes a more efficient and accurate warning of thermal runaway of energy storage batteries through multi-feature fusion, model interpretability, and forward-looking prediction of future temperature changes, and has significant practical value in ensuring the safety and reliability of energy storage systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is the flowchart of Embodiment 1; Figure 2 is the system diagram of Embodiment 2; Figure 3 is the flowchart of Embodiment 3; Figure 4 is the flowchart of Embodiment 6; Figure 5 is the system diagram of Embodiment 7. DETAILED DESCRIPTION OF THE EMBODIMENTS To further understand the content of the present invention, the following provides a detailed description of the present invention in combination with the accompanying drawings and specific embodiments. It should be understood that the embodiments are only for explaining the present invention and not for limiting it.

[0026] Embodiment 1: Refer to Figure 1 , a prediction method for thermal runaway of energy storage batteries, comprising the following steps: S1, obtain the historical data of the temperature, voltage, and current of the batteries in the battery pack, and construct a decision tree using the historical data.

[0027] S2. Obtain the current data of the temperature, voltage, and current of the batteries in the battery pack, slice the obtained data, and obtain the data at the current moment and the required time period as the initial data.

[0028] S3. Use a decision tree to process the initial data and predict the temperature change data of the battery in the required future time period.

[0029] S4. Analyze the temperature change data of the battery in the required future time period, and judge whether the battery undergoes thermal runaway in the required future time period according to the analysis results.

[0030] Embodiment 2: See Figure 2 , a prediction system for energy storage battery thermal runaway, including: A decision tree construction module, which is used to obtain the historical data of the temperature, voltage, and current of the batteries in the battery pack and construct a decision tree using the historical data.

[0031] An initial data acquisition module, which is used to obtain the current data of the temperature, voltage, and current of the batteries in the battery pack, slice the obtained data, and obtain the data at the current moment and the required time period as the initial data.

[0032] A decision tree processing module, which is used to process the initial data using a decision tree and predict the temperature change data of the battery in the required future time period.

[0033] A thermal runaway prediction module, which is used to analyze the temperature change data of the battery in the required future time period and judge whether the battery undergoes thermal runaway in the required future time period according to the analysis results.

[0034] Embodiment 3: This embodiment further limits the steps of S1 and the functions of the decision tree construction module on the basis of the above embodiment, specifically as follows: S11. Obtain the objective function of the decision tree without leaves as the original objective function.

[0035] S12. Obtain the historical data of the temperature, voltage, and current of the batteries in the battery pack, sort the historical data, traverse all possible decision tree structures in the historical data, and calculate the current objective functions of all decision tree structures.

[0036] S13. Obtain the difference between the current objective function and the original objective function. If there are positive values in the difference, select the decision tree structure corresponding to the largest positive value as the splitting node; if there are no positive values in the difference, do not perform splitting.

[0037] S14. Traverse all the initial data, split the decision tree, and obtain the final decision tree.

[0038] Further: Refer to Figure 3 to obtain the objective function of the leafless decision tree as the original objective function.

[0039] Obtain the historical data of the temperature, voltage, and current of the batteries in the battery pack, sort the historical data, traverse all possible decision tree structures in the historical data, and calculate the current objective function of all decision tree structures , The value of is . Calculate . If there is a positive value among these values, select the decision tree structure corresponding to the largest positive value as the first node. If there is no positive value among these

[0040] values, do not perform splitting and return the structure of the best tree. Based on the above splitting method, sort the data of the left subtree and the right subtree according to the voltage magnitude. Taking the left subtree as an example, traverse all possible decision tree structures and calculate . Calculate

[0041] . If there is a positive value among them, select the tree structure corresponding to the largest positive value as the first node. If there is no positive value, do not perform splitting and return the structure of the best tree. Similarly, perform the same operation on the right subtree. Return the optimal decision tree after splitting until the decision tree reaches the maximum number of constructed trees and save the shapes, nodes, and weights of the leaves of trees.

[0042] Embodiment 4: This embodiment further limits the steps of S2 and the functions of the initial data acquisition module on the basis of the above embodiment, specifically as follows: S21. Obtain that the temperature of the batteries in the battery pack is , the voltage is , and the current data is , where is the number of data samples.

[0043] S22. Slice the data samples , preferably , is the data samples after slicing is the data interval within the current time and the required time period.

[0044] The data samples after slicing As initial data.

[0045] Example 5: Based on the above example, this example further defines the steps of S3 and the functions of the decision tree processing module, specifically as follows: S31, construct weak classification decision trees, preset the penalty term for each weak classification decision tree, and integrate all weak classification decision trees into a decision tree model. Specifically: The goal of the algorithm is the structural parameters of the decision tree, enabling it to correctly predict to the temperature between, taking the temperature at as an example, using the data from to to train the model, enabling it to correctly predict the temperature at. Define the loss function for the temperature prediction value at a single time point:

[0046] Among them, represents the true temperature value at, represents the temperature value predicted by the algorithm. Squaring is to offset the sign impact of the loss function due to too low or too high predicted values.

[0047] Construct weak classification decision trees, and through the prediction results of each weak classification decision tree, then sum to obtain the final predicted value as the output value of the decision tree model:

[0048] Among them, is the final predicted value, is the th in the weak classification decision trees, preset the penalty term for each weak classification decision tree as:

[0049] Among them, is the number of leaves in the weak classification decision tree, is the weight of the th leaf node in the weak classification decision tree, is the square penalty of the leaf node weight,

[0050] is the weighted ratio of the number of leaves in the weak classification decision tree, is the weighted ratio of the leaf node weight sum; The preset objective function of the th decision tree is:

[0050] wherein, represents the error of all temperature prediction values, represents the sum of penalty terms of all weak classification decision trees, is the loss function, is the true value of the th sample, is the predicted value of the th sample;

[0051] wherein, represents the first derivative of the loss function, represents the second derivative of the loss function.

[0052] Derive this function to obtain the optimal node weight :

[0053] wherein, is the set of all sample indices of the

[0054] S32. Input the initial data into the decision tree model to obtain the temperature change data of the battery in the future required time period.

[0055] Example 6: Refer to Figure 4 . This example further limits the steps of S4 and the functions of the thermal runaway prediction module on the basis of the above example, specifically as follows: S41. Preset the alarm threshold and the continuous warning times .

[0056] S42. Obtain the temperature change data of the battery in the future required time period , and calculate the mean value of the temperature change data of the battery in the future required time period:

[0057] S43. Subtract the temperature change data of the battery in the future required time period from the mean value to obtain a sequence with a mean value of 0:

[0058] S44, Calculate the cumulative sum of the sequence and obtain the minimum value of the cumulative sum , The cumulative sum of the sequence is:

[0059] S45, Calculate the difference between all other cumulative sums and the minimum value of the cumulative sum , And all the differences are compared with the alarm threshold . If the difference is greater than the alarm threshold , it is recorded as a warning.

[0060] S46, When the number of recorded warnings is greater than the preset number of consecutive warnings , it is determined that the battery is about to experience thermal runaway within the required time period in the future.

[0061] Embodiment 7: Refer to Figure 5 , The present invention also provides an electronic device 100 for a prediction method of energy storage battery thermal runaway; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0062] The memory 101 can be used to store the computer program 103. The processor 102 realizes the steps of a prediction method for energy storage battery thermal runaway described in Embodiment 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 can include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one magnetic disk storage device, a flash device, or other non-volatile solid-state storage devices.

[0063] The at least one processor 102 can be a Central Processing Unit (CPU), or can also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 can be a microprocessor or the processor 102 can also be any conventional processor, etc. The processor 102 is the control center of the electronic device 100, and connects various parts of the entire electronic device 100 using various interfaces and lines.

[0064] The memory 101 in the electronic device 100 stores a plurality of instructions to implement a prediction method for thermal runaway of an energy storage battery. The processor 102 can execute the plurality of instructions to implement: Obtain historical data of the temperature, voltage, and current of the batteries in the battery pack, and construct a decision tree using the historical data; Obtain the current data of the temperature, voltage, and current of the batteries in the battery pack, slice the obtained data to obtain the data at the current moment and the required time period, and use it as the initial data; Process the initial data using the decision tree to predict the temperature change data of the battery in the required future time period; Analyze the temperature change data of the battery in the required future time period, and determine whether the battery will experience thermal runaway in the required future time period according to the analysis results.

[0065] Embodiment 8: If the modules / units integrated in the electronic device 100 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, and read-only memory (ROM, Read-Only Memory).

[0066] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, system, or computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code.

[0067] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0068] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0069] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable apparatus to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable apparatus provide steps for realizing the functions specified in one process or a plurality of processes and / or boxes. Figure 1 One process or a plurality of processes and / or boxes Figure 1 Steps for realizing the functions specified in one box or a plurality of boxes.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific implementation manners of the present invention, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A prediction method for thermal runaway of energy storage batteries, characterized in that, It includes the following steps: Obtain the historical data of the temperature, voltage and current of the batteries in the battery pack, and construct a decision tree using the historical data; Obtain the current data of the temperature, voltage and current of the batteries in the battery pack, slice the obtained data to get the data at the current moment and the required time period, and use it as the initial data; Process the initial data using the decision tree to predict the temperature change data of the battery in the required future time period; Analyze the temperature change data of the battery in the required future time period, and judge whether the battery will experience thermal runaway in the required future time period according to the analysis results.

2. The prediction method for thermal runaway of an energy storage battery according to claim 1, characterized in that The specific method of obtaining the historical data of the temperature, voltage and current of the batteries in the battery pack and constructing a decision tree using the historical data is as follows: Obtain the objective function of the decision tree without leaves as the original objective function; Obtain the historical data of the temperature, voltage and current of the batteries in the battery pack, sort the historical data, traverse all possible decision tree structures in the historical data, and calculate the current objective function of all decision tree structures; Obtain the difference between the current objective function and the original objective function. If there is a positive value in the difference, select the decision tree structure corresponding to the largest positive value as the splitting node; If there is no positive value in the difference, no splitting is performed; Traverse all the initial data, split the decision tree to obtain the final decision tree.

3. A prediction method for thermal runaway of energy storage batteries according to claim 1, characterized in that, The specific method of obtaining the data of the temperature, voltage and current of the batteries in the battery pack, slicing the obtained data to get the data at the current moment and the required time period, and using it as the initial data is as follows: Obtain the temperature of the battery in the battery pack as , the voltage is , the current data is , where is the number of data samples; Slice the data sample, , where the sliced data sample is the data range for the current moment and the required time period Data sample after slicing As the initial data.

4. A prediction method for thermal runaway of energy storage batteries according to claim 1, characterized in that, The specific method of processing the initial data using the decision tree to predict the temperature change data of the battery in the required future time period is as follows: Construct a number of weak classification decision trees, preset the penalty term for each weak classification decision tree, and integrate all the weak classification decision trees into a decision tree model; Input the initial data into the decision tree model to obtain the temperature change data of the battery in the required future time period.

5. A prediction method for thermal runaway of energy storage batteries according to claim 4, characterized in that Construct a number of decision trees for weak classification, preset a penalty term for each decision tree for weak classification, and the specific method for integrating all decision trees for weak classification into a decision tree model is as follows: Construct weak classification decision trees, and obtain the final predicted value as the output value of the decision tree model by summing the prediction results of each weak classification decision tree: in, is the final predicted value, for The first Tree, is the initial data; Penalty term for each weak classification decision tree preset It is as follows: Among them, is the number of leaves in the decision tree of weak classification, is the weight of the th leaf node in the decision tree of weak classification, is the square penalty of the leaf node weight, is the weighted ratio of the number of leaves in the decision tree of weak classification, is the weighted ratio of the sum of leaf node weights; The objective function of the preset th decision tree is as follows: Among them, represents the error of all temperature prediction values, represents the sum of penalty terms of all weak classification decision trees, is the loss function, is the true value of the th sample, is the predicted value of the th sample; The objective function is simplified by using second-order Taylor expansion and then differentiated to obtain the optimal node weights : wherein, is the first-order derivative of the loss function with respect to the predicted value, is the second-order derivative of the loss function with respect to the predicted value, is the set of all sample indices of the 6. The prediction method for thermal runaway of energy storage batteries according to claim 1, wherein The specific method of analyzing the temperature change data of the battery in the required future time period and judging whether the battery will experience thermal runaway in the required future time period according to the analysis results is as follows: Preset the alarm threshold and the number of consecutive early warnings; Obtain the temperature change data of the battery in the required future time period, and calculate the average value of the temperature change data of the battery in the required future time period; Subtract the temperature change data of the battery in the required future time period from the average value to obtain a sequence with a mean of 0; Calculate the cumulative sum of the sequence and obtain the minimum value of the cumulative sum; Calculate the difference between all other cumulative sums and the minimum value of the cumulative sum, compare all the differences with the alarm threshold. If the difference is greater than the alarm threshold, record it as an early warning; When the number of recorded early warnings is greater than the preset number of consecutive early warnings, it is judged that the battery will soon experience thermal runaway in the required future time period.

7. A prediction system for thermal runaway of energy storage batteries, characterized in that, It includes: A decision tree construction module, which is used to obtain the historical data of the temperature, voltage and current of the batteries in the battery pack and construct a decision tree using the historical data; An initial data acquisition module, which is used to obtain the current data of the temperature, voltage and current of the batteries in the battery pack, slice the obtained data to get the data at the current moment and the required time period, and use it as the initial data; A decision tree processing module, which is used to process the initial data using the decision tree to predict the temperature change data of the battery in the required future time period; A thermal runaway prediction module, which is used to analyze the temperature change data of the battery within the required future time period, and determine whether the battery will undergo thermal runaway within the required future time period according to the analysis results.

8. The prediction system for thermal runaway of energy storage batteries according to claim 7, characterized in that, The function of the decision tree construction module is implemented by the following method: Obtain the objective function of the decision tree without leaves as the original objective function; Obtain the historical data of the temperature, voltage and current of the batteries in the battery pack, sort the historical data, traverse all possible decision tree structures in the historical data, and calculate the current objective functions of all decision tree structures; Obtain the difference between the current objective function and the original objective function. If there are positive values in the difference, select the decision tree structure corresponding to the largest positive value as the splitting node; If there are no positive values in the difference, no splitting is performed; Traverse all the initial data, split the decision tree, and obtain the final decision tree.

9. A prediction system for thermal runaway of energy storage batteries according to claim 7, characterized in that The function of the initial data acquisition module is implemented by the following method: Obtain the temperature of the battery in the battery pack as , the voltage is , the current data is , where is the number of data samples; Slice the data sample, , be the sliced data sample, be the data range for the current moment and the required time period Data sample after slicing As the initial data.

10. A prediction system for thermal runaway of energy storage batteries according to claim 7, characterized in that, The function of the decision tree processing module is implemented by the following method: Construct weakly classified decision trees, preset the penalty terms for each of the weakly classified decision trees, and integrate all the weakly classified decision trees into a decision tree model; Input the initial data into the decision tree model to obtain the temperature change data of the battery within the required future time period.

11. A prediction system for thermal runaway of energy storage batteries according to claim 7, characterized in that, The function of the thermal runaway prediction module is implemented by the following method: Preset an alarm threshold and the number of consecutive early warnings; Obtain the temperature change data of the battery within the required future time period, and calculate the average value of the temperature change data of the battery within the required future time period; Subtract the temperature change data of the battery within the required future time period from the average value to obtain a sequence with a zero mean; Calculate the cumulative sum of the sequence and obtain the minimum value of the cumulative sum; Calculate the difference between all other cumulative sums and the minimum value of the cumulative sum, compare all the differences with the alarm threshold. If the difference is greater than the alarm threshold, record it as an early warning; When the recorded number of early warnings is greater than the preset number of consecutive early warnings, it is determined that the battery is about to undergo thermal runaway within the required future time period.

12. An electronic device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of a prediction method for thermal runaway of an energy storage battery according to any one of claims 1 to 6.

13. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of a prediction method for thermal runaway of an energy storage battery according to any one of claims 1 to 6.

Images