Energy storage lithium ion battery combustion and explosion risk detection method and device, computer equipment, readable storage medium and program product

Abstract

The application relates to a lithium ion battery explosion risk detection method and device for energy storage, computer equipment, a computer readable storage medium and a computer program product, relates to the technical field of energy storage batteries, and can improve the risk detection efficiency of the lithium ion battery. The method comprises the following steps: acquiring real-time operation data of the lithium ion battery and environment parameters of an energy storage environment; inputting the real-time operation data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining service life of the lithium ion battery; determining a weight optimization target according to the health state prediction value, the remaining service life and the environment parameters, and performing iterative screening in a preset index weight solution space of an explosion risk detection index to determine target index weights meeting the weight optimization target; and determining an explosion risk grade of the lithium ion battery according to the target index weights and the explosion risk detection index.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage batteries, in particular to a lithium ion battery explosion risk detection method and device for energy storage, computer equipment, computer readable storage medium and computer program product. BACKGROUND

[0002] With the wide application of lithium ion batteries in the field of energy storage, the safety problem of lithium ion batteries has become increasingly prominent.

[0003] In the traditional technology, the detection of the explosion risk of lithium ion batteries usually adopts a static detection method. For example, some methods only rely on a single or a few parameters (such as battery thermal runaway temperature) to detect the risk. However, in the face of the complex operating environment and real-time changes of working conditions of current lithium ion batteries for energy storage, the detection method using fixed weights or static indicators in the prior art cannot adapt well, resulting in low accuracy of the detection of the explosion risk of lithium ion batteries. SUMMARY

[0004] Therefore, it is necessary to provide a lithium ion battery explosion risk detection method and device for energy storage, computer equipment, computer readable storage medium and computer program product to solve the above technical problems.

[0005] In a first aspect, the present application provides a lithium ion battery explosion risk detection method for energy storage, comprising:

[0006] obtaining real-time operating data of the lithium ion battery and environmental parameters of the energy storage environment;

[0007] inputting the real-time operating data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining service life of the lithium ion battery; the health state prediction model is trained based on time sequence features, battery aging labels and service life labels of historical operating data;

[0008] determining a weight optimization target according to the health state prediction value, the remaining service life and the environmental parameters, and iteratively screening in a preset index weight solution space of an explosion risk detection index to determine a target index weight meeting the weight optimization target; the explosion risk detection index includes the health state prediction value and the remaining service life;

[0009] determining an explosion risk level of the lithium ion battery according to the target index weight and the explosion risk detection index.

[0010] In one embodiment, the explosion risk detection index further includes a voltage fluctuation index and a temperature gradient index; the voltage fluctuation index and the temperature gradient index are determined by the following steps:

[0011] determine a voltage time sequence characteristic parameter according to the voltage drop amplitude and the voltage drop duration of the lithium ion battery;

[0012] generate a voltage fluctuation abnormality signal as the voltage fluctuation indicator if the voltage time sequence characteristic parameter meets a preset voltage abnormality condition;

[0013] determine a temperature distribution characteristic parameter according to the temperature difference change rate between adjacent temperature detection points;

[0014] generate a temperature gradient change signal as the temperature gradient indicator if the temperature distribution characteristic parameter meets a preset temperature abnormality condition.

[0015] In one of the embodiments, the explosion risk detection indicator includes a flammability indicator;

[0016] After the target indicator weight meeting the weight optimization objective is determined through the iterative screening in the preset indicator weight solution space of the explosion risk detection indicator, the method further includes:

[0017] obtain the charge-discharge cycle number of the lithium ion battery;

[0018] if the charge-discharge cycle number exceeds a preset cycle threshold, adjust the indicator weight corresponding to the health state prediction value according to a first preset proportion;

[0019] obtain the flammable gas concentration in the environmental parameter;

[0020] if the flammable gas concentration exceeds a preset concentration threshold, adjust the indicator weight corresponding to the flammability indicator according to a second preset proportion;

[0021] based on the adjusted indicator weight corresponding to the health state prediction value and the adjusted indicator weight corresponding to the flammability indicator, return to execute the step of determining the target indicator weight meeting the weight optimization objective through the iterative screening in the preset indicator weight solution space of the explosion risk detection indicator.

[0022] In one of the embodiments, the explosion risk detection indicator includes a temperature indicator and a humidity indicator;

[0023] After the target indicator weight meeting the weight optimization objective is determined through the iterative screening in the preset indicator weight solution space of the explosion risk detection indicator, the method further includes:

[0024] obtain the temperature in the environmental parameter;

[0025] if the temperature exceeds a preset temperature threshold, adjust the indicator weight corresponding to the temperature indicator according to a third preset proportion;

[0026] acquiring humidity in the environmental parameter;

[0027] if the humidity exceeds a preset humidity threshold, adjusting an index weight corresponding to the humidity index according to a fourth preset proportion;

[0028] based on the adjusted index weight corresponding to the temperature index and the index weight corresponding to the humidity index, returning to perform the step of iteratively screening in the preset index weight solution space of the explosion risk detection index to determine the target index weight meeting the weight optimization target.

[0029] In one of the embodiments, the acquiring the real-time operation data of the lithium ion battery comprises:

[0030] acquiring a current state of charge of the lithium ion battery;

[0031] if the state of charge is greater than a preset state of charge threshold, generating a collection strategy adjustment instruction and sending it to a data collection terminal; the collection strategy adjustment instruction is used to instruct the data collection terminal to adjust the collection frequency for the real-time operation data to a first collection frequency; wherein the first collection frequency is higher than a second collection frequency of the lithium ion battery in a normal state of charge.

[0032] In one of the embodiments, the determining the explosion risk level of the lithium ion battery according to the target index weight and the set of explosion risk detection indexes comprises:

[0033] performing weighted calculation on the explosion risk detection indexes according to the target index weight to obtain a comprehensive risk value, and determining a candidate explosion risk level based on the comprehensive risk value;

[0034] if the health state prediction value is less than a preset health risk threshold, performing a risk level transition operation based on the candidate explosion risk level to obtain a target explosion risk level;

[0035] based on the target explosion risk level, determining early warning information and risk response strategies to respond to the explosion risk.

[0036] In a second aspect, the present application further provides a lithium ion battery explosion risk detection device for energy storage, comprising:

[0037] a data acquisition module, configured to acquire real-time operation data of the lithium ion battery and environmental parameters of an energy storage environment;

[0038] a state prediction module, configured to input the real-time operation data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining service life of the lithium ion battery; the health state prediction model is trained based on time sequence features, battery aging labels and service life labels of historical operation data;

[0039] an index weight optimization module, configured to determine a weight optimization target according to the health state prediction value, the remaining service life and the environmental parameter, and perform iterative screening in a preset index weight solution space of a combustion and explosion risk detection index to determine a target index weight meeting the weight optimization target; the combustion and explosion risk detection index includes the health state prediction value and the remaining service life;

[0040] a risk determination module, configured to determine a combustion and explosion risk grade of the lithium ion battery according to the target index weight and the combustion and explosion risk detection index.

[0041] In a third aspect, the present application further provides a computer device, comprising a memory and a processor, the memory stores a computer program, and the processor realizes the following steps when executing the computer program:

[0042] obtaining real-time operation data of the lithium ion battery and an environmental parameter of an energy storage environment;

[0043] inputting the real-time operation data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining service life of the lithium ion battery; the health state prediction model is trained based on time sequence features, battery aging labels and service life labels of historical operation data;

[0044] determining a weight optimization target according to the health state prediction value, the remaining service life and the environmental parameter, and performing iterative screening in a preset index weight solution space of a combustion and explosion risk detection index to determine a target index weight meeting the weight optimization target; the combustion and explosion risk detection index includes the health state prediction value and the remaining service life;

[0045] determining a combustion and explosion risk grade of the lithium ion battery according to the target index weight and the combustion and explosion risk detection index.

[0046] In a fourth aspect, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the following steps:

[0047] obtaining real-time operation data of the lithium ion battery and an environmental parameter of an energy storage environment;

[0048] inputting the real-time operation data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining service life of the lithium ion battery; the health state prediction model is trained based on time sequence features, battery aging labels and service life labels of historical operation data;

[0049] determining a weight optimization target according to the health state prediction value, the remaining service life and the environmental parameter, and performing iterative screening in a preset index weight solution space of a combustion risk detection index to determine a target index weight satisfying the weight optimization target; the combustion risk detection index includes the health state prediction value and the remaining service life;

[0050] determining a combustion risk level of the lithium ion battery according to the target index weight and the combustion risk detection index.

[0051] In a fifth aspect, the present application further provides a computer program product comprising a computer program which, when executed by a processor, implements the following steps:

[0052] obtaining real-time operation data of the lithium ion battery and environmental parameters of an energy storage environment;

[0053] inputting the real-time operation data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining service life of the lithium ion battery; the health state prediction model is trained based on time sequence features, battery aging labels and service life labels of historical operation data;

[0054] determining a weight optimization target according to the health state prediction value, the remaining service life and the environmental parameter, and performing iterative screening in a preset index weight solution space of a combustion risk detection index to determine a target index weight satisfying the weight optimization target; the combustion risk detection index includes the health state prediction value and the remaining service life;

[0055] determining a combustion risk level of the lithium ion battery according to the target index weight and the combustion risk detection index.

[0056] The energy storage lithium ion battery explosion risk detection method, device, computer equipment, computer readable storage medium and computer program product, by acquiring real-time running data of the lithium ion battery and environment parameters of the energy storage environment; inputting the real-time running data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining service life of the lithium ion battery; the health state prediction model is trained based on time sequence features, battery aging labels and service life labels of historical running data; a weight optimization target is determined according to the health state prediction value, the remaining service life and the environment parameters, and iterative screening is performed in a preset index weight solution space of the explosion risk detection index to determine a target index weight meeting the weight optimization target; the explosion risk detection index includes the health state prediction value and the remaining service life; and a lithium ion battery explosion risk level is determined according to the target index weight and the explosion risk detection index. In the present application, by inputting the real-time running data into the pre-trained neural network model, the health state and the remaining service life of the battery are accurately predicted by using the time sequence feature mapping relationship of the historical running data, and the influence of the aging degree of the battery on the safety is effectively quantified; and then the weight optimization target is determined in combination with the environment parameters, the target weight coefficient is dynamically determined in the preset weight coefficient solution space by using the iterative screening algorithm, the adaptive adjustment of the evaluation index weight with the battery aging state and the external environment change is realized, and the accuracy of the explosion risk detection of the energy storage lithium ion battery is improved. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the description of the embodiments of the present application or the related art will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other related drawings can be obtained by those skilled in the art without creative labor.

[0058] Figure 1 A flowchart of the energy storage lithium ion battery explosion risk detection method in one embodiment;

[0059] Figure 2 A schematic diagram of the battery surface temperature, voltage and temperature rise rate curve at full load in one embodiment;

[0060] Figure 3 A schematic diagram of the battery thermal runaway gas production rate change with time at different state of charge in one embodiment;

[0061] Figure 4 A schematic diagram of the battery surface maximum temperature and maximum pressure change with state of charge in one embodiment;

[0062] Figure 5Flowchart of the lithium-ion battery explosion risk detection method for energy storage in another embodiment;

[0063] Figure 6 Flowchart of the lithium-ion risk monitoring and early warning method for energy storage in an embodiment;

[0064] Figure 7 Block diagram of the lithium-ion battery explosion risk detection device for energy storage in an embodiment;

[0065] Figure 8 Internal structure diagram of the computer device in an embodiment. DETAILED DESCRIPTION

[0066] In order to make the purposes, technical solutions and advantages of the present application clearer, further detailed description will be made to the present application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0067] It should be noted that the terms "first", "second", etc. used in the present application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "include" and "have" used in the present application and any variations thereof are intended to cover non-exclusive inclusion. The term "multiple" used in the present application refers to two and more than two. The term "and / or" used in the present application refers to one of the options, or any combination of multiple options.

[0068] In an embodiment, as shown in Figure 1 A lithium-ion battery explosion risk detection method for energy storage is provided. The embodiment illustrates the method applied to a terminal. It should be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and can be realized through the interaction of the terminal and the server. The terminal can be a data acquisition end deployed on the site of an energy storage system. The data acquisition end is used to perceive multi-dimensional data of the lithium-ion battery for energy storage. The data acquisition end can include but is not limited to a sensor network or an edge computing node. The sensor network includes sensors for collecting battery monomer or module electrochemical data (such as voltage, current, temperature, etc.), and environmental monitoring sensors for collecting energy storage environmental parameters (such as environmental temperature, humidity, air pressure, gas concentration, etc.). It should be noted that the above-mentioned sensors can be contact sensors or non-contact sensors, and the present application embodiment does not make specific limitation thereto. In the present embodiment, the method includes the following steps:

[0069] Step S102, acquiring real-time running data of the lithium-ion battery and environmental parameters of the energy storage environment.

[0070] The real-time operation data can be physical quantity data representing internal electrochemical characteristics and external working characteristics of the lithium-ion battery in a charging and discharging cycle or a static state, including but not limited to voltage, current, temperature, internal resistance, state of charge, etc., and can be obtained in real time based on a battery management system or a sensing component deployed at a battery monomer, a module or a cluster level, according to a preset sampling frequency or a triggering strategy.

[0071] The environmental parameters can be physical quantities describing the environmental state of the physical space where the energy storage system is located, such as environmental temperature, environmental humidity, atmospheric pressure and concentration of specific gases (such as hydrogen, carbon monoxide, etc.), which are used to assist in analyzing the potential impact of external environmental stress on the safety of the battery.

[0072] Specifically, for the acquisition of real-time operation data, electrochemical parameters during battery operation can be continuously monitored. These data are usually presented in the form of time series, which can reflect the voltage fluctuations, current loads and temperature changes of the battery at different time points. During the acquisition process, the battery monomers can be sampled one by one, or the battery modules can be sampled as a whole, thereby obtaining a multi-dimensional operation data stream containing time stamps.

[0073] At the same time, for the acquisition of environmental parameters, data from environmental monitoring equipment can be received synchronously. These data reflect the external conditions during battery operation, such as real-time temperature field distribution, humidity level and whether there is abnormal gas leakage in the energy storage cabin, etc. The acquired real-time operation data and environmental parameters can be preprocessed (such as denoising, normalization, etc.) and stored in a local database or cache, so as to provide data support for subsequent input of health state prediction model and determination of weight optimization target.

[0074] In step S104, the real-time operation data is input into the pre-trained health state prediction model to obtain the health state prediction value and the remaining useful life of the lithium-ion battery. The health state prediction model is trained based on the time series features of historical operation data, battery aging labels and useful life labels.

[0075] The health state prediction model can be a time series analysis model based on an artificial intelligence algorithm (e.g., a Long Short-Term Memory (LSTM), a Gated Recurrent Unit (GRU), or a Recurrent Neural Network (RNN)), which is used to represent the complex nonlinear evolution mapping relationship between the external observable electrochemical operation parameters of the lithium ion battery and the internal unobservable health state (SOH) and remaining useful life (RUL), and can be generated and deployed based on a large amount of lithium ion battery full life cycle historical operation data (covering time series features of charge-discharge curves in different aging stages) and corresponding real aging labels (such as capacity retention rate) and life labels after deep learning training and parameter optimization.

[0076] Specifically, the acquired real-time operation data can be first preprocessed, such as data cleaning, outlier rejection, and normalization, to construct time series data segments conforming to the model input specification. Subsequently, key feature indicators that can sensitively reflect the battery aging degree can be further mined based on the processed data. Specifically, these feature indicators can include the slope change rate of the charge-discharge curve (Slope Change Rate, SCR) ), the internal resistance growth rate (Internal Resistance Growth Rate, IRGR) ), and the self-discharge rate (Self-Discharge Rate, SDR), etc. The slope change rate of the charge-discharge curve can effectively reflect the evolution of the internal polarization effect of the battery, the internal resistance growth rate directly reflects the increase of the internal electrochemical impedance of the battery, and the self-discharge rate is closely related to the degree of internal micro-short circuit or side reaction of the battery.

[0077] Next, the above extracted multi-dimensional aging feature vector is input into the pre-trained health state prediction model. Taking the Long Short-Term Memory (LSTM) as an example, this model can effectively capture the time series dependence relationship in the long-term operation data of the battery by using its unique memory unit and gating mechanism (such as input gate, forget gate, and output gate), thereby calculating and outputting the health state prediction value (e.g., SOH represented in percentage) and the remaining useful life prediction value (e.g., RUL represented by the number of remaining cycles or the remaining service time) of the lithium ion battery at the current time based on the fusion of the above-mentioned multiple aging feature indicators. In this way, the digital and quantitative representation of the battery aging degree can be realized, and precise data input is provided for subsequent inclusion of the battery aging state as a key dimension into the explosion risk assessment system.

[0078] Step S106, determine a weight optimization target according to the health state prediction value, the remaining use life and the environmental parameter, and perform iterative screening in a preset index weight solution space of the explosion risk detection index to determine a target index weight satisfying the weight optimization target. The explosion risk detection index includes the health state prediction value and the remaining use life.

[0079] The weight optimization target can be a mathematical function or evaluation criterion for measuring rationality of different risk index weight distribution schemes based on the current internal and external states of the battery, aiming to enable the final evaluation result to dynamically adapt to the aging degree and environmental stress level of the battery, and can be constructed based on the health state prediction value, the remaining use life and the environmental parameter and through a preset logical rule or mapping relationship.

[0080] The preset index weight solution space can be a multi-dimensional vector space containing all possible weight value combinations of the explosion risk detection index, used to limit the search range of weight optimization, and can be pre-set according to expert experience or historical data statistical characteristics.

[0081] Specifically, the health state prediction value, the remaining use life and the real-time collected environmental parameter obtained according to the previous steps can be used to dynamically construct or adjust the weight optimization target. Specifically, the optimization target can be designed as a fitness function, which can reflect the rationality of the contribution of each risk index to the overall explosion risk under the current specific aging stage (characterized by SOH and RUL) and environmental conditions (characterized by environmental parameters). For example, when the battery is severely aged (low SOH) or in extreme environment (such as high temperature and high humidity), the optimization target can guide the search direction to tilt towards the direction of increasing the weight of the related index (such as SOH or environmental index), to ensure that the evaluation model maintains sufficient sensitivity to the current dominant risk factor.

[0082] Subsequently, an iterative screening process can be started in the preset index weight solution space containing the health state prediction value, the remaining use life and other explosion risk detection indexes. This process can simulate swarm intelligence optimization behavior, generate multiple potential weight combinations (particles) in the solution space, and calculate the fitness value of each weight combination according to the weight optimization target determined above. In the iterative process, the search position and speed can be updated continuously to gradually approach the optimal solution satisfying the optimization target (such as the maximum or minimum of the fitness function value). Finally, when the preset convergence condition (such as the number of iterations reaching the upper limit or the fitness change being less than a threshold value) is met, the current global optimal solution can be determined as the target index weight, thereby completing the dynamic assignment of the weights of the detection indexes.

[0083] Step S108, determine the explosion risk level of the lithium ion battery according to the target index weight and the explosion risk detection index.

[0084] The explosion risk level can be a quantitative classification mark representing the probability and severity of thermal runaway or combustion explosion of the lithium ion battery in the current operating state, used to guide the energy storage system to take graded early warning or differentiated safety control measures, and can be determined based on the comparison between the weighted calculation result of each detection index and the preset risk threshold interval, usually in the form of numerical score, grade number (such as I, II, III) or semantic label (such as low risk, high risk).

[0085] The explosion risk detection index can be a multi-dimensional parameter set involved in the weighted calculation, including the health state prediction value, the remaining useful life and possibly other electrochemical or environmental characteristic quantities obtained or calculated in the previous steps.

[0086] Specifically, based on the selected target index weight and the corresponding explosion risk detection index value, the normalized values of the health state prediction value, the remaining useful life and other indicators are multiplied by their corresponding dynamic weights and accumulated by using weighted summation or other fusion algorithms, so as to obtain a comprehensive risk score reflecting the overall safety condition of the battery at the current time. Subsequently, the comprehensive risk score can be mapped to the preset risk classification standard. For example, the comprehensive risk score can be compared with a plurality of preset risk determination thresholds: if the score is in the first interval, it can be determined as a low risk level, corresponding to a normal monitoring state; if the score is in the second interval, it can be determined as a medium risk level, corresponding to an attention or warning state; if the score exceeds the critical threshold, it can be determined as a high risk level, corresponding to an emergency disposal state. In addition, the determined explosion risk level can be packaged as a state packet to trigger subsequent safety response mechanisms such as audible and visual alarms, system shutdown or fire linkage.

[0087] In this embodiment, by inputting real-time operating data into a pre-trained neural network model, the health state and the remaining useful life of the battery are accurately predicted using the time sequence feature mapping relationship of historical operating data, effectively quantifying the influence of the aging degree of the battery on safety; further, the weight optimization target is determined in combination with the environmental parameters, the target weight coefficient is dynamically determined in the preset weight coefficient solution space using an iterative screening algorithm, realizing adaptive adjustment of the evaluation index weight with the change of the battery aging state and the external environment, and improving the accuracy of the explosion risk detection of the lithium ion battery for energy storage.

[0088] In one exemplary embodiment, the explosion risk detection index further includes a voltage fluctuation index and a temperature gradient index; the voltage fluctuation index and the temperature gradient index are determined by the following steps:

[0089] The voltage time sequence characteristic parameter is determined according to the voltage drop amplitude and the duration of the lithium ion battery;

[0090] If the voltage time sequence characteristic parameter meets the preset voltage abnormality condition, a voltage fluctuation abnormality signal is generated as a voltage fluctuation index; a temperature distribution characteristic parameter is determined according to a temperature difference change rate between adjacent temperature detection points; if the temperature distribution characteristic parameter meets a preset temperature abnormality condition, a temperature gradient change signal is generated as a temperature gradient index.

[0091] The voltage fluctuation index can be a quantitative parameter representing an unexpected large drop or oscillation of the lithium ion battery terminal voltage in a short time, and is used to identify early fault features such as internal micro-short circuit or separator failure of the battery. The voltage fluctuation index can be extracted based on real-time collected voltage time sequence data through a sliding window algorithm or differential calculation.

[0092] The temperature gradient index can be a physical quantity describing the degree of non-uniformity of the temperature field distribution on the surface of the battery monomer or inside the module and the temperature rise rate, and is used to capture the evolution trend of local thermal runaway. The temperature gradient index can be calculated based on the difference between adjacent temperature sensors or the derivative of temperature with respect to time at a single detection point. The preset voltage abnormality condition and the preset temperature abnormality condition are determination logics set based on the safety operation boundary of the battery, such as a voltage drop amplitude threshold or a temperature rise rate threshold, and are used to distinguish between normal working condition fluctuations and potential thermal runaway precursors.

[0093] For example, for voltage data analysis, the terminal can first use wavelet transform or empirical mode decomposition (EMD) to denoise and modal separate the original voltage signal to extract the subtle change signal in the battery operation process. Then, the voltage drop amplitude (ΔV) in the sliding time window and the duration of the drop process are calculated. In one embodiment, as shown in FIG. 2, the X-axis (left) is time (Time) in minutes (min), the Y-axis (left) indicates voltage (Voltage) in volts (V), and the Y-axis (right) indicates temperature (Temperature) in degrees Celsius (°C). The X-axis (right) indicates the temperature rise rate (Temperature rise rate) in °C / min. Figure 2 The voltage curve corresponding to the left Y-axis voltage in the figure shows a clear voltage drop characteristic (Voltage drop) in the early stage of thermal runaway (before reaching the exothermic reaction starting temperature ). This means that a micro-short circuit has occurred inside the battery. The internal pressure of the battery accumulates to a certain extent and breaks through the pressure relief valve. This is usually accompanied by the release of gas (evaporation or decomposition of electrolyte). The pressure relief process will take away a small amount of heat, causing a temporary fluctuation in temperature, thereby causing the safety valve to open (Safety valve open). Due to the opening of the safety valve, the temperature rises to ​​​After the thermal runaway trigger temperature, the temperature starts to deviate from the linear rise and enters the self-acceleration stage, and finally reaches the highest temperature At this time, the temperature change rate starts to change from low noise to a clear upward trend.

[0094] Based on the above, the terminal can compare the extracted voltage characteristics with the preset quantization standard. If the voltage drop amplitude AV>0.5V and the duration of this state is >2s (i.e., the preset voltage abnormal condition is met), it is determined that a micro-short circuit or insulation failure may occur inside the battery, and a voltage fluctuation abnormal signal is generated as a voltage fluctuation indicator.

[0095] At the same time, for the analysis of temperature data, the terminal can calculate the temperature difference change rate between adjacent temperature detection points or the temperature rise rate of a single point based on the data collected by the high-precision temperature sensor. In one embodiment, as shown in Figure 2 The temperature curve and temperature rise rate ( ) curve corresponding to the right Y-axis of the figure intuitively show the thermal behavior in the battery thermal runaway process. Before the thermal runaway trigger, the temperature rise rate will reach a peak value and be accompanied by a sharp rise in local temperature. Based on this, the terminal can monitor the temperature difference change between adjacent sensors in real time: if the temperature difference change rate (or temperature rise rate) between adjacent sensors is >15℃ / min (i.e., the preset temperature abnormal condition is met), it indicates that a local hot spot may be forming inside the battery module. The terminal immediately triggers a local hot spot alarm and generates a temperature gradient change signal as a temperature gradient indicator.

[0096] In this embodiment, through the above method, voltage and temperature characteristics with high physical interpretability can be accurately extracted from real-time running data. This method can capture dangerous signals in the early stage of thermal runaway, so as to include the voltage fluctuation indicator and the temperature gradient indicator as high-sensitivity input factors into the subsequent risk assessment system, significantly improving the timeliness of the explosion risk detection and the advance of the early warning.

[0097] In an exemplary embodiment, after iterating and screening in the preset indicator weight solution space of the explosion risk detection indicator, and determining the target indicator weight that meets the weight optimization goal, the method further includes:

[0098] The number of charge-discharge cycles of the lithium ion battery is obtained; if the number of charge-discharge cycles exceeds a preset cycle threshold, the index weight corresponding to the state of health prediction value is adjusted according to a first preset proportion; the concentration of combustible gas in the environmental parameter is obtained; if the concentration of combustible gas exceeds a preset concentration threshold, the index weight corresponding to the flammability index is adjusted according to a second preset proportion; based on the adjusted index weight corresponding to the state of health prediction value and the index weight corresponding to the flammability index, the step of performing iterative screening in the preset index weight solution space of the combustion risk detection index is returned to execute to determine the target index weight that meets the weight optimization target.

[0099] The number of charge-discharge cycles can be a cumulative value recorded for one complete charge-discharge process of the lithium ion battery, used to represent the aging process and capacity attenuation degree of the battery during long-term service.

[0100] The flammability index can be a quantitative parameter reflecting the accumulation degree of combustible gas (such as hydrogen and alkane gas) generated by side reactions during battery thermal runaway, which is a key basis for determining whether the battery is in a critical state of combustion and explosion.

[0101] The first preset proportion can be a weight correction factor for representing the negative influence of battery aging degree on thermal stability, used to increase the weight proportion of the state of health index when the number of battery cycles is high, to compensate for the risk of safety margin decrease of old battery.

[0102] The second preset proportion can be a high-risk response coefficient set based on the gas explosion limit, used to forcibly increase the weight of the flammability index when combustible gas leakage is monitored, to ensure high sensitivity to critical combustion risk.

[0103] Illustratively, after the terminal preliminarily determines the target index weight that meets the optimization target, the life cycle state and extreme environmental characteristics of the battery can be further introduced as correction factors. For the correction of the battery aging dimension, the terminal can obtain the number of charge-discharge cycles of the lithium ion battery, and compare the cycle number with a preset cycle threshold (e.g., 300 times). If it is monitored that the number of charge-discharge cycles exceeds 300 times, it indicates that the battery has entered the later stage of aging, the stability of the internal material has decreased, and the trigger threshold of thermal runaway may be lowered. At this time, in order to strengthen the risk control of the aging battery, the terminal can adjust the index weight corresponding to the state of health prediction value (SOH) according to a first preset proportion (e.g., an increase of 10%), so that its proportion in risk assessment increases.

[0104] Secondly, for the correction of the combustible gas dimension, the terminal can obtain the concentration of combustible gas in the environmental parameter. In one embodiment, as Figure 3As shown in the figure, the dramatic change in gas production over time during thermal runaway is evident. With increasing State of Charge (SOC), the onset time of gas production from internal side reactions is significantly advanced, and the gas production rate and final total gas production increase substantially. This suggests that rapid gas release is often a direct precursor to impending thermal runaway. Based on this, the terminal can monitor combustible gases (hydrogen). The concentration is compared with the preset concentration threshold in real time: if gas production is detected... If the volume fraction is greater than 6% (i.e., exceeding the preset concentration threshold), it indicates that the battery may be at a critical point of venting or imminent combustion and explosion. At this time, the terminal can automatically increase the weight of the indicator corresponding to the flammability indicator according to the second preset ratio to ensure that this key indicator plays a dominant role in the risk level calculation.

[0105] Based on the adjusted weights of the health status prediction and flammability index, the terminal can use these revised weights as new constraints or initial values ​​to return to the step of iterative selection in the preset index weight solution space. Through this secondary iteration mechanism, under the premise of satisfying new safety constraints (such as aging weighting and gas weighting), the globally optimal target index weights are searched again.

[0106] In this embodiment, the above steps can effectively solve the problem of insufficient sensitivity of traditional static evaluation methods under extreme conditions such as late battery aging or gas leakage. At the same time, with the secondary iterative screening of weights, the adaptability and safety of the risk assessment model in the whole life cycle and complex failure scenarios are improved.

[0107] In an exemplary embodiment, after iteratively filtering through the preset index weight solution space of the combustion and explosion risk detection index to determine the target index weight that satisfies the weight optimization objective, the method further includes:

[0108] The process involves: acquiring the temperature from the environmental parameters; adjusting the weight of the temperature index according to a third preset ratio if the temperature exceeds a preset temperature threshold; acquiring the humidity from the environmental parameters; adjusting the weight of the humidity index according to a fourth preset ratio if the humidity exceeds a preset humidity threshold; and then, based on the adjusted weights of the temperature and humidity indexes, returning to the preset weight solution space for the explosion risk detection index to iteratively filter and determine the target index weight that meets the weight optimization objective.

[0109] The temperature index and the humidity index can be quantitative parameters representing the potential influence of the external physical environment of the energy storage system on the thermal stability and electrical insulation performance of the battery, wherein the temperature index is related to the heat dissipation efficiency and heat accumulation risk of the battery, and the humidity index is related to the insulation strength and condensation short circuit risk of the high-voltage components.

[0110] The preset temperature threshold and the preset humidity threshold are warning values set according to the optimal operating environment interval and the limit tolerance boundary of the lithium ion battery.

[0111] For example, for the correction of the environmental temperature dimension, the terminal can obtain real-time temperature data in the environmental parameters. Since the environmental temperature directly determines the heat dissipation capacity of the battery system, a high temperature environment can significantly reduce the thermal runaway trigger threshold of the battery. The obtained temperature can be compared with the preset temperature threshold (for example, 45°C). If it is monitored that the environmental temperature exceeds the threshold, it indicates that the battery is in a high-risk area of heat accumulation. At this time, the terminal can adjust the index weight corresponding to the temperature index according to a third preset proportion, so that it occupies a larger proportion in the comprehensive risk calculation, thereby more sensitively reflecting the influence of thermal stress on safety.

[0112] At the same time, for the correction of the environmental humidity dimension, the terminal can obtain real-time humidity data in the environmental parameters. A high-humidity environment can easily cause condensation in the battery pack, thereby causing a decrease in insulation resistance and even high-voltage arc short circuit, and can also accelerate the corrosion failure of the circuit board. The terminal can compare the monitored humidity (such as relative humidity) with the preset humidity threshold (for example, 60%). If it is found that the environmental humidity exceeds the threshold, the terminal can identify that there is a high electrical failure risk at present, and then adjust the index weight corresponding to the humidity index according to a fourth preset proportion (for example, directly set the weight to 0.15 or increase a certain proportion based on the original weight), in order to prevent risk misjudgment due to neglecting environmental humidity.

[0113] Finally, based on the temperature index weight and the humidity index weight corrected by the environmental stress, the terminal can return to the step of performing iteration and screening in the preset index weight solution space of the explosion risk detection index, taking these weight values as locked constraint conditions or penalty terms. Through this mechanism, the optimization algorithm is driven to search for the optimal weight combination that can simultaneously consider the battery state (SOH, RUL) and the external harsh environment (high temperature, high humidity) in the solution space.

[0114] In this embodiment, by constructing a weight dynamic correction mechanism based on environmental temperature and humidity, the derived safety risks of the energy storage system under complex climate conditions can be effectively addressed, ensuring that the explosion risk assessment result can be dynamically adjusted in real time following the environmental changes, avoiding the risk assessment lag or underestimation due to the use of conventional weights in harsh environments, and significantly enhancing the safety perception ability of the system in all-weather environments.

[0115] In one exemplary embodiment, acquiring real-time operating data of a lithium-ion battery includes:

[0116] The current state of charge (SOC) of the lithium-ion battery is obtained. If the SOC is greater than a preset SOC threshold, a data acquisition strategy adjustment instruction is generated and sent to the data acquisition terminal. The data acquisition strategy adjustment instruction is used to instruct the data acquisition terminal to adjust the acquisition frequency for real-time operating data to a first acquisition frequency. The first acquisition frequency is higher than the second acquisition frequency of the lithium-ion battery under normal SOC.

[0117] Among them, the state of charge (SOC) is a physical quantity that reflects the ratio of the remaining charge of a battery to its rated capacity. It is used to characterize the level of chemical energy stored inside the battery and is an important reference for assessing the potential energy release intensity of a battery's combustion and explosion.

[0118] The data acquisition strategy adjustment command can be a control signal used to change the working mode of the sensor network, instructing the underlying hardware to switch the granularity of data sampling under different risk levels, and can be sent to edge computing nodes or data acquisition terminals through communication protocols.

[0119] The first and second acquisition frequencies can be the data sampling rates corresponding to high-risk and normal states, respectively, to ensure that transient voltage changes or temperature rise signals can be captured at critical moments.

[0120] For example, since SOC is directly related to the energy density inside the battery, its level is positively correlated with the destructive force after thermal runaway. In a specific embodiment, such as... Figure 4 As shown in the figure, the severity of thermal runaway in a battery is strongly correlated with the State of Charge (SOC). For example, as the SOC gradually increases from 25% to 100%, the highest temperature generated during thermal runaway (…) ) and maximum pressure ( All of these show a significant upward trend, especially when the SOC is high (e.g., above 75% or 80%), the energy released instantaneously during thermal runaway increases sharply, and the peak values ​​of temperature and pressure reach extremely high levels. This means that once a battery in a high SOC state fails, its evolution is extremely fast and its destructive power is enormous.

[0121] Based on the above physical law, the application sets a preset charge threshold (for example, 80%), when the terminal obtains the real-time SOC value exceeding the preset charge threshold, it can be determined that the battery is currently in a high-risk state of high energy accumulation. At this time, in order to avoid missing the transient characteristic signal (such as millisecond-level voltage drop) in the initial stage of thermal runaway due to too large sampling interval, the terminal can generate a collection strategy adjustment instruction and send it to the front-end data collection terminal. The instruction is used to instruct the collection terminal to adjust the collection frequency of key data such as voltage and temperature from the second collection frequency under normal working conditions to a higher first collection frequency (for example, increased to 1Hz or higher). Conversely, when the SOC is lower than the threshold, the second lower collection frequency is maintained to balance the data transmission bandwidth and storage pressure.

[0122] In the embodiment, by constructing a variable frequency data collection mechanism based on SOC, the sampling point can be automatically encrypted when the battery is in a high-risk energy state; this strategic frequency adjustment not only ensures low-power operation in normal state, but also ensures that the fleeting pre-fault signal can be captured with sufficient time resolution in high-risk state, effectively improving the success rate of early warning.

[0123] In one exemplary embodiment, according to the target index weight and the set of explosion risk detection indexes, the explosion risk level of the lithium ion battery is determined, comprising:

[0124] According to the target index weight, the explosion risk detection index is weighted and calculated to obtain a comprehensive risk value, and based on the comprehensive risk value, a candidate explosion risk level is determined; if the health state prediction value is less than a preset health risk threshold, a risk level transition operation is performed based on the candidate explosion risk level to obtain a target explosion risk level; based on the target explosion risk level, warning information and risk response strategy are determined to respond to the explosion risk.

[0125] Among them, the comprehensive risk value can be a dimensionless value or a percentage score calculated based on the multi-dimensional explosion risk index and its corresponding dynamic weight, which is used to intuitively quantify the overall safety situation of the current battery system.

[0126] The candidate explosion risk level can be a preliminary classification result obtained by comparing the comprehensive risk value with a standard risk classification table, which is used to represent the general risk level before considering the aging-specific correction.

[0127] The risk level transition operation can be a logic control strategy for correcting the potential safety margin of the old battery, which compensates for the risk of thermal stability decline due to battery aging by forcibly increasing the risk level. It can be set based on the positive correlation between battery state of health (SOH) and thermal runaway critical temperature.

[0128] The target explosion risk level is the final safety state determination result after aging correction, and is used to directly associate with specific risk response strategies such as system shutdown or sound-light alarm.

[0129] Exemplarily, the terminal can normalize the health state prediction value, the remaining service life, and the explosion risk detection indexes such as the voltage fluctuation, the temperature gradient, and the flammability that can be generated in the previous step. Then, the terminal can perform weighted calculation (for example, weighted summation) on the normalized index values by using the determined target index weight, so as to obtain a comprehensive risk value reflecting the real-time safety condition of the current battery system.

[0130] The terminal can map the comprehensive risk value to a preset risk level interval table, so as to determine a candidate explosion risk level (for example, low risk, medium risk, and high risk). On this basis, in order to solve the problem of sudden increase in risk caused by battery aging that is ignored by the existing evaluation method, the terminal can further introduce a risk level transition mechanism.

[0131] In a specific embodiment, as the number of battery cycles increases, the state of health (SOH) decreases, resulting in poor stability of the internal diaphragm of the battery and a corresponding decrease in the initial triggering temperature of thermal runaway. Therefore, the terminal can compare the currently predicted health state prediction value (SOH) with a preset health risk threshold (for example, 80%). If the SOC is greater than 80%, it indicates that the battery has entered a deep aging stage, and its ability to resist thermal runaway is significantly weaker than that of a new battery. At this time, even if the candidate explosion risk level calculated above is low (for example, low risk), the terminal can perform a risk level transition operation to automatically increase the risk level by one level (for example, adjust to medium risk), so as to obtain the final target explosion risk level.

[0132] Finally, the terminal can match corresponding warning information and risk response strategies in the local database based on the target explosion risk level. For example, when the target explosion risk level is the highest, the terminal can generate a response strategy of cutting off the main circuit contactor, starting the fire extinguishing device, and sending an emergency alarm, so as to respond to the explosion risk in a timely manner.

[0133] In this embodiment, by using the above steps, the fusion analysis capability of the dynamic weight algorithm for multi-dimensional data is fully utilized, the current comprehensive risk is accurately quantified, and the poor thermal stability of the aged battery is compensated, so that the safety and reliability of the energy storage system during the whole life cycle are significantly improved.

[0134] In an exemplary embodiment, the risk detection method provided by the present application can run in an Internet of Things architecture environment including a sensor network, edge computing, and cloud evaluation, wherein the sensor network can be composed of voltage, current, and temperature sensor arrays deployed at the battery cluster or module level and gas, temperature and humidity environment monitoring units distributed in the energy storage cabin, for capturing the electrochemical transient characteristics of the battery and the environmental stress state; edge computing can transmit data streams to the cloud while performing part of the preprocessing logic through communication links such as industrial Ethernet; the cloud can be a cloud server or a local monitoring host, which is built-in with a deep learning prediction model and a swarm intelligence optimization algorithm, for performing complex weight optimization and risk deduction, and finally responding to the risk state through a visual terminal or an execution mechanism. The method specifically includes the following steps, as shown in Figure 5

[0135] Step S501, frequency data acquisition and preprocessing. The acquisition operation of real-time running data of lithium ion batteries can be performed. In this process, in order to balance the data processing load and the timeliness of risk capture, a dynamic acquisition strategy based on state of charge (SOC) can be introduced. Specifically, the current state of charge of the battery can be acquired, and the state of charge is compared with the preset charge threshold. If it is found that the current state of charge is greater than the preset charge threshold, it indicates that the battery is in a high energy accumulation state, at which time an acquisition strategy adjustment instruction can be generated to instruct the data acquisition terminal to adjust the acquisition frequency of key data such as voltage and temperature to a higher first acquisition frequency, so as to ensure that transient abnormalities can be captured; otherwise, a lower second acquisition frequency can be maintained. Based on the determined acquisition frequency, the voltage, current, temperature of the battery and the environmental parameters (such as temperature, humidity, gas concentration, etc.) of the energy storage environment are acquired in real time.

[0136] Step S502, feature extraction of multi-dimensional risk indicators. After acquiring the basic running data, more representative risk indicators can be further extracted from the time domain dimension. Specifically, the voltage drop in a specific window and the duration of the drop can be calculated to determine the voltage time sequence characteristic parameter; if the parameter meets the preset voltage abnormality condition, a voltage fluctuation abnormal signal is generated as a voltage fluctuation indicator. At the same time, the temperature difference change rate between adjacent temperature detection points can be calculated to determine the temperature distribution characteristic parameter; if the parameter meets the preset temperature abnormality condition, a temperature gradient change signal is generated as a temperature gradient indicator. These two indicators will be used as input dimensions for subsequent evaluation.

[0137] ​Step S503, health status and remaining life prediction. Synchronously, the aging state of the battery can be quantitatively evaluated by using an artificial intelligence model. The collected real-time running data can be input into a pre-trained health status prediction model. The model outputs the health status prediction value and the remaining useful life of the lithium ion battery at the current time by analyzing the slope change rate, internal resistance growth rate and other deep features in the data. These two parameters constitute the core indicators of the evaluation system about the physical condition of the battery.

[0138] Step S504, weight dynamic optimization based on multi-dimensional constraints. After obtaining all the detection indicators mentioned above, enter the core weight distribution stage. The initial weight optimization target can be determined according to the health status prediction value, the remaining useful life and the environmental parameters, and the iterative screening is started in the preset solution space. In this process, in order to improve the robustness of the evaluation, a secondary correction mechanism based on environment and aging can be introduced:

[0139] Aging and gas correction, obtain the charge and discharge cycle number of the battery and the concentration of combustible gas in the environment. If the cycle number exceeds the preset cycle threshold, the weight of the health status prediction value is increased by a first preset proportion; if the concentration of combustible gas exceeds the preset concentration threshold, the weight of the flammability index is increased by a second preset proportion.

[0140] Temperature and humidity correction, obtain the environmental temperature and humidity. If the temperature exceeds the preset temperature threshold, the weight of the temperature index is increased by a third preset proportion; if the humidity exceeds the preset humidity threshold, the weight of the humidity index is increased by a fourth preset proportion. Based on the weight parameters corrected above, the iterative screening step can be returned to execute, and finally the target index weight meeting the optimization target is determined.

[0141] Step S505, risk level determination and transition. Finally, based on the determined weight and index value, the final quantification and grading of risk can be performed. The target index weight can be used to calculate the comprehensive risk value by weighting each explosion risk detection index, and a candidate explosion risk level is determined accordingly. On this basis, the health status prediction value of the battery can be further verified. If it is found that the prediction value is less than the preset health risk threshold, it indicates that the battery is seriously aged, and at this time the risk level transition operation (such as automatically increasing by one level) can be performed based on the candidate explosion risk level, so as to obtain the final target explosion risk level. Based on the target level, the system can determine the corresponding warning information and coping strategy to realize the precise control of the explosion risk, for example, Figure 6 As shown, when the low risk level is determined, the normal monitoring state is corresponding; when the medium risk level is determined, the warning state is corresponding; and when the high risk level is determined, the emergency disposal state is corresponding.

[0142] It should be understood that although each step in the flowchart involved in the above embodiments is shown in sequence according to the arrow, these steps are not necessarily executed in the order indicated by the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in the above embodiments can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but can be alternately executed with at least part of other steps or steps or stages in other steps. It can be understood that the steps in different embodiments can be freely combined as needed, and various non-contradictory schemes formed by the combination all belong to the scope of protection of the present application.

[0143] Based on the same inventive concept, the embodiments of the present application also provide a lithium-ion battery explosion risk detection device for energy storage for implementing the above-mentioned lithium-ion battery explosion risk detection method for energy storage. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more lithium-ion battery explosion risk detection device embodiments for energy storage provided below can refer to the limitations of the lithium-ion battery explosion risk detection method for energy storage described above, which will not be repeated here.

[0144] In one exemplary embodiment, as shown in Figure 7 A lithium-ion battery explosion risk detection device for energy storage is provided, comprising: a data acquisition module 710, a state prediction module 720, an index weight optimization module 730, and a risk determination module 740, wherein:

[0145] The data acquisition module 710 is configured to acquire real-time operation data of the lithium-ion battery and environmental parameters of the energy storage environment.

[0146] The state prediction module 720 is configured to input the real-time operation data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining useful life of the lithium-ion battery. The health state prediction model is trained based on time sequence features, battery aging labels, and useful life labels of historical operation data.

[0147] The index weight optimization module 730 is configured to determine a weight optimization target according to the health state prediction value, the remaining useful life, and the environmental parameters, and iteratively screen in a pre-set index weight solution space of the explosion risk detection index to determine a target index weight that satisfies the weight optimization target. The explosion risk detection index includes the health state prediction value and the remaining useful life.

[0148] a risk determination module 740, configured to determine a risk level of the lithium ion battery according to the target indicator weight and the explosion risk detection indicator.

[0149] In one of the embodiments, the explosion risk detection indicator further includes a voltage fluctuation indicator and a temperature gradient indicator; the data acquisition module 710 is further configured to:

[0150] determine a voltage time sequence characteristic parameter according to the voltage drop amplitude and the drop duration of the lithium ion battery;

[0151] generate a voltage fluctuation abnormal signal as the voltage fluctuation indicator if the voltage time sequence characteristic parameter meets a preset voltage abnormal condition;

[0152] determine a temperature distribution characteristic parameter according to the temperature difference change rate between adjacent temperature detection points;

[0153] generate a temperature gradient change signal as the temperature gradient indicator if the temperature distribution characteristic parameter meets a preset temperature abnormal condition.

[0154] In one of the embodiments, the explosion risk detection indicator includes a flammability indicator; the indicator weight optimization module 730 is further configured to:

[0155] acquire the charge-discharge cycle number of the lithium ion battery;

[0156] adjust the indicator weight corresponding to the health state prediction value according to a first preset proportion if the charge-discharge cycle number exceeds a preset cycle threshold;

[0157] acquire the flammable gas concentration in the environmental parameter;

[0158] adjust the indicator weight corresponding to the flammability indicator according to a second preset proportion if the flammable gas concentration exceeds a preset concentration threshold;

[0159] based on the adjusted indicator weight corresponding to the health state prediction value and the indicator weight corresponding to the flammability indicator, return to execute the step of iteratively screening in the preset indicator weight solution space of the explosion risk detection indicator to determine the target indicator weight meeting the weight optimization target.

[0160] In one of the embodiments, the explosion risk detection indicator includes a temperature indicator and a humidity indicator; the indicator weight optimization module 730 is further configured to:

[0161] acquire the temperature in the environmental parameter;

[0162] adjust the indicator weight corresponding to the temperature indicator according to a third preset proportion if the temperature exceeds a preset temperature threshold;

[0163] obtaining humidity in the environment parameters;

[0164] if the humidity exceeds a preset humidity threshold, adjusting an index weight corresponding to the humidity index according to a fourth preset proportion;

[0165] based on the adjusted index weight corresponding to the temperature index and the index weight corresponding to the humidity index, returning to perform the step of iteratively screening in the preset index weight solution space of the combustion and explosion risk detection index to determine the target index weight meeting the weight optimization target.

[0166] In one of the embodiments, the data acquisition module 710 is further configured to:

[0167] obtaining the current state of charge of the lithium ion battery;

[0168] if the state of charge is greater than a preset state of charge threshold, generating a collection strategy adjustment instruction and sending it to the data collection terminal; the collection strategy adjustment instruction is used to instruct the data collection terminal to adjust the collection frequency of the real-time operation data to a first collection frequency; wherein the first collection frequency is higher than a second collection frequency of the lithium ion battery under normal state of charge.

[0169] In one of the embodiments, the risk determination module 740 is further configured to: according to the target index weight, performing weighted calculation on the combustion and explosion risk detection index to obtain a comprehensive risk value, and based on the comprehensive risk value, determining a candidate combustion and explosion risk level;

[0170] if the health state prediction value is less than a preset health risk threshold, performing a risk level transition operation based on the candidate combustion and explosion risk level to obtain a target combustion and explosion risk level;

[0171] based on the target combustion and explosion risk level, determining early warning information and risk response strategy to respond to the combustion and explosion risk.

[0172] The above-mentioned various modules in the lithium ion battery combustion and explosion risk detection device for energy storage can be realized by software, hardware and their combinations in whole or in part. The above-mentioned various modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to the above-mentioned various modules.

[0173] In one exemplary embodiment, a computer device is provided, which can be a terminal, and its internal structure diagram can be as shown in Figure 8As shown in the figure. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. Among them, the processor, the memory and the input / output interface are connected through the system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capability. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be realized through WIFI, mobile cellular network, near field communication (Near Field Communication, NFC) or other technologies. The computer program is executed by the processor to realize a lithium ion battery explosion risk detection method for energy storage.

[0174] Those skilled in the art can understand that, Figure 8 The skilled in the art can understand that,

[0175] In one embodiment, a computer device is provided, including a memory and a processor, the memory stores a computer program, and the processor executes the computer program to realize the steps in the above method embodiments.

[0176] In one embodiment, a computer readable storage medium is provided, which stores a computer program, and the computer program is executed by the processor to realize the steps in the above method embodiments.

[0177] In one embodiment, a computer program product is provided, including a computer program, and the computer program is executed by the processor to realize the steps in the above method embodiments.

[0178] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0179] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0180] It can be understood by those skilled in the art that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing related hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments of each method. In the embodiments provided in the present application, any reference to memory, database or other medium can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., without being limited thereto.

[0181] Any technical features in the above embodiments can be combined, and for the sake of brevity, not all possible combinations are described above, however, any combination of these technical features is deemed to be within the scope of the present application.

[0182] The above embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it should not be understood as a limitation on the patent scope of the present application. It should be pointed out that, for ordinary skilled persons in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for detecting the risk of fire and explosion of a lithium-ion battery for energy storage, characterized by, The method comprises: acquiring real-time operation data of the lithium ion battery and environmental parameters of the energy storage environment; inputting the real-time operation data into a pre-trained health state prediction model to obtain a health state prediction value and a remaining useful life of the lithium ion battery; the health state prediction model is trained based on time sequence features, battery aging labels and useful life labels of historical operation data; wherein the time sequence features include key aging feature indicators reflecting the evolution of internal polarization and impedance of the battery, which are extracted based on historical operation data; the key aging feature indicators include a slope change rate of a charge-discharge curve, an internal resistance growth rate and a self-discharge rate; determining a weight optimization target according to the health state prediction value, the remaining useful life and the environmental parameters, and iteratively screening in a preset index weight solution space of a combustion risk detection index to determine a target index weight meeting the weight optimization target; the combustion risk detection index includes the health state prediction value, a flammability index and the remaining useful life; after the iteratively screening in the preset index weight solution space of the combustion risk detection index to determine the target index weight meeting the weight optimization target, the method further comprises: acquiring a charge-discharge cycle number of the lithium ion battery; if the charge-discharge cycle number exceeds a preset cycle threshold, adjusting an index weight corresponding to the health state prediction value according to a first preset proportion; acquiring a flammable gas concentration in the environmental parameters; if the flammable gas concentration exceeds a preset concentration threshold, adjusting an index weight corresponding to the flammability index according to a second preset proportion; based on the adjusted index weight corresponding to the health state prediction value and the index weight corresponding to the flammability index, returning to perform the step of iteratively screening in the preset index weight solution space of the combustion risk detection index to determine the target index weight meeting the weight optimization target; determining a combustion risk level of the lithium ion battery according to the target index weight and the combustion risk detection index.

2. The method of claim 1, wherein, The combustion risk detection index further includes a voltage fluctuation index and a temperature gradient index; the voltage fluctuation index and the temperature gradient index are determined by the following steps: determining a voltage time sequence feature parameter according to a voltage drop amplitude and a drop duration of the lithium ion battery; if the voltage time sequence feature parameter meets a preset voltage abnormality condition, generating a voltage fluctuation abnormality signal as the voltage fluctuation index; determining a temperature distribution feature parameter according to a temperature difference change rate between adjacent temperature detection points; if the temperature distribution feature parameter meets a preset temperature abnormality condition, generating a temperature gradient change signal as the temperature gradient index.

3. The method of claim 1, wherein, The combustion risk detection index includes a temperature index and a humidity index; after the iteratively screening in the preset index weight solution space of the combustion risk detection index to determine the target index weight meeting the weight optimization target, the method further comprises: acquiring a temperature in the environmental parameters; if the temperature exceeds a preset temperature threshold, adjusting an index weight corresponding to the temperature index according to a third preset proportion; acquiring a humidity in the environmental parameters; if the humidity exceeds a preset humidity threshold, the humidity indicator corresponding to the index weight is adjusted according to a fourth preset proportion; Based on the adjusted temperature index corresponding to the index weight and the humidity index corresponding to the index weight, return to execute the step of iterating and screening in the preset index weight solution space of the combustion risk detection index to determine the target index weight meeting the weight optimization target.

4. The method of claim 1, wherein, The real-time operation data of the lithium ion battery is obtained, including: Obtain the current state of charge of the lithium ion battery; If the state of charge is greater than a preset charge threshold, generate a collection strategy adjustment instruction and send it to the data collection terminal; the collection strategy adjustment instruction is used to instruct the data collection terminal to adjust the collection frequency of the real-time operation data to a first collection frequency; wherein the first collection frequency is higher than the second collection frequency of the lithium ion battery under normal state of charge.

5. The method according to any one of claims 1 to 4, characterized in that, The target index weight and the combustion risk detection index are used to determine the combustion risk level of the lithium ion battery, including: According to the target index weight, the combustion risk detection index is weighted and calculated to obtain a comprehensive risk value, and based on the comprehensive risk value, a candidate combustion risk level is determined; If the health state prediction value is less than a preset health risk threshold, the risk level transition operation is performed based on the candidate combustion risk level to obtain a target combustion risk level; Based on the target combustion risk level, determine the early warning information and risk response strategy to respond to the combustion risk.

6. A device for detecting the risk of explosion of a lithium-ion battery for energy storage, characterized by The device includes: Data acquisition module, for acquiring real-time operation data of the lithium ion battery and environmental parameters of the energy storage environment; State prediction module, for inputting the real-time operation data into a pre-trained health state prediction model to obtain the health state prediction value and the remaining useful life of the lithium ion battery; the health state prediction model is trained based on the time sequence characteristics of the historical operation data, the battery aging label and the service life label; wherein the time sequence characteristics include key aging characteristic indicators reflecting the evolution of internal polarization and impedance of the battery based on historical operation data, the key aging characteristic indicators include the slope change rate of the charge-discharge curve, the internal resistance growth rate and the self-discharge rate; The index weight optimization module is configured to determine a weight optimization target according to the health state prediction value, the remaining use life, and the environmental parameter, and perform iterative screening in a preset index weight solution space of the explosion risk detection index to determine a target index weight satisfying the weight optimization target; the explosion risk detection index includes the health state prediction value, a flammability index, and the remaining use life; after the iterative screening in the preset index weight solution space of the explosion risk detection index to determine the target index weight satisfying the weight optimization target, the method further includes: obtaining a charge-discharge cycle number of the lithium ion battery; if the charge-discharge cycle number exceeds a preset cycle threshold, adjusting an index weight corresponding to the health state prediction value according to a first preset proportion; obtaining a combustible gas concentration in the environmental parameter; if the combustible gas concentration exceeds a preset concentration threshold, adjusting an index weight corresponding to the flammability index according to a second preset proportion; and based on the adjusted index weight corresponding to the health state prediction value and the index weight corresponding to the flammability index, returning to perform the iterative screening in the preset index weight solution space of the explosion risk detection index to determine the target index weight satisfying the weight optimization target. The risk determination module is configured to determine an explosion risk level of the lithium ion battery according to the target index weight and the explosion risk detection index. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-6 when the computer program is executed by the processor. The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 5.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 5.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 5.

Images