Battery SOX estimation method based on internal sensing signal

Abstract

The invention discloses a battery SOX estimation method based on an internal sensing signal, and relates to the field of battery state estimation and safety early warning processing, and the method comprises the steps: carrying out the abuse experiment of a battery under different multiplying power and environment conditions, and determining an early warning threshold value and early warning advance time; authenticating the battery through security authentication, and acquiring a current sensing signal of the battery passing the authentication; the current sensing information comprises the current, the temperature and the internal air pressure of the battery in the operation process of the battery; determining a battery SOH predicted value and a battery SOC estimated value according to the current sensor signal; based on the battery SOH predicted value and the battery SOC estimated value, the SOX of the battery can be accurately estimated, and effective suppression can be carried out in the early stage of thermal runaway.

Description

Technical Field

[0001] This application relates to the field of battery state estimation and safety warning processing, and particularly to a method for estimating battery SOX based on internal sensing signals. Background Art

[0002] With the development of battery technology, the demand for batteries in various application scenarios has increased rapidly. Especially in the fields of electric vehicles, intelligent devices, and renewable energy, the performance of batteries directly affects the reliability and service life of the devices. Therefore, accurate estimation of the State of Health (SOH) and State of Charge (SOC) of batteries has become crucial. Most traditional battery SOH estimation methods are based on changes in voltage, current, or internal resistance. However, the changes in these parameters often lag behind the actual process of battery aging. Under complex operating conditions, these methods still face challenges in terms of accuracy and stability and cannot accurately reflect the battery state. The resulting hidden danger is that in the early stage of Thermal Runaway (TR), the changes in these parameters are often not sensitive enough or lagging, unable to provide sufficient and timely safety warnings for the management system, resulting in the inability to effectively contain the spread of thermal runaway. On the other hand, effectively managing and controlling the charging process of batteries, preventing the use of unqualified batteries, and avoiding improper charging behaviors are also crucial for avoiding thermal runaway.

[0003] In recent years, the relationship between air pressure changes and battery aging has attracted the attention of researchers. Since gases are generated during battery aging, the internal air pressure of the battery increases accordingly. The change in air pressure can be used as an important indicator of battery aging. Therefore, the SOH estimation method based on the internal air pressure of the battery has gradually become a research hotspot. At the same time, the change in the internal air pressure of a single-cycle battery is related to the change in the material state, gas generation reaction, capacity attenuation, and charge-discharge state inside the battery. If characteristic parameters can be effectively extracted from the change in the internal air pressure of the battery and a mathematical model is established, the air pressure change can be used as a new state observable to accurately estimate the SOC of the battery. Summary of the Invention

[0004] The purpose of this application is to provide a method for estimating battery SOX based on internal sensing signals, which can accurately estimate the SOX of the battery, where the SOX includes SOH and SOC.

[0005] To achieve the above purpose, this application provides the following solutions:

[0006] In the first aspect, this application provides a method for estimating battery SOX based on internal sensing signals, including:

[0007] Conduct abuse experiments on the battery under different magnification and environmental conditions to determine the warning threshold and the early warning lead time;

[0008] Authenticate the battery through safety certification and obtain the current sensing signals of the certified battery; the current sensing information includes the current, temperature, and internal air pressure during the operation of the battery;

[0009] Determine the predicted value of the battery's SOH and the estimated value of the battery's SOC based on the current sensor signals.

[0010] According to the specific embodiments provided in this application, the following technical effects are disclosed in this application:

[0011] This application determines the warning threshold and the early warning lead time through abuse experiments, and through safety certification, obtains the current sensing signals of the certified battery, so as to accurately estimate the SOC estimated value and the SOH predicted value based on the current sensing signals of the battery; this application accurately locates abnormal batteries based on the current sensing signals of the certified battery, avoiding the operation of suppressing the entire battery pack due to the abnormality of one battery, and reducing the cost of battery danger suppression.

[0012] In addition, this application can also give early warnings based on the estimated value of SOC and the predicted value of SOH in the early stage of battery thermal runaway and formulate suppression measures according to the actual scenario for early control of thermal runaway, and effectively suppress it in the early stage of thermal runaway. Description of the Drawings

[0013] In order to more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0014] Figure 1 It is a flowchart of a method for estimating battery SOX based on internal sensing signals provided in an embodiment of this application;

[0015] Figure 2 It is a schematic diagram of the dynamic changes of voltage, air pressure, and temperature based on an internal air pressure sensor provided in an embodiment of this application;

[0016] Figure 3 It is a graph of the thermal runaway warning time and threshold based on internal air pressure provided in an embodiment of this application; where, Figure 3 In (a) is a schematic diagram of the air pressure warning time; Figure 3 In (b) is a schematic diagram of the air pressure warning threshold;

[0017] Figure 4Schematic diagram of a battery authentication method based on an implanted chip provided by an embodiment of the present application; wherein, Figure 4 In (a) is a schematic diagram of a battery charging device; Figure 4 In (b) is a schematic diagram of a battery;

[0018] Figure 5 Schematic diagram of the structure of a battery top cover with a pressure sensor placed thereon provided by an embodiment of the present application; wherein, Figure 5 In (a) is a schematic diagram of the structure of a battery top cover with the pressure sensor blocked by insulating plastic; Figure 5 In (b) is a schematic diagram of the structure of a battery top cover with the insulating plastic hidden;

[0019] Figure 6 Schematic diagram of a square battery with a strain gauge pasted on an explosion-proof valve provided by an embodiment of the present application;

[0020] Figure 7 Graph of the total internal pressure, pressure caused by temperature, and pressure caused by aging in a battery under long-cycle conditions provided by an embodiment of the present application;

[0021] Figure 8 Graph of battery SOH estimation based on internal pressure signals provided by an embodiment of the present application;

[0022] Figure 9 Modeling diagram of internal pressure behavior for accurate SOC estimation provided by an embodiment of the present application;

[0023] Figure 10 Schematic diagram of a thermal runaway suppression device for an energy storage power station provided by an embodiment of the present application;

[0024] Figure 11 Diagram of a thermal runaway suppression method for an energy storage power station based on internal pressure provided by an embodiment of the present application;

[0025] Figure 12 Diagram of a thermal runaway suppression method for an electric vehicle based on internal pressure provided by an embodiment of the present application.

[0026] Reference numerals: 1, top cover; 2, groove; 3, pressure sensor; 4, explosion-proof valve; 5, strain gauge. Detailed implementation manners

[0027] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0028] To make the above objects, features, and advantages of the present application more apparent and understandable, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] An embodiment of the present application provides a method for estimating battery SOX based on internal sensing signals. This method is executed by a computer device, specifically, it can be executed alone by a computer device such as a terminal or a server, or jointly executed by a terminal and a server. In the embodiment of the present application, as Figure 1 shown, the method includes the following steps.

[0030] S1: Conduct abuse experiments on the battery under different rates and environmental conditions to determine the warning threshold and warning lead time.

[0031] S2: Authenticate the battery through security authentication and obtain the current sensing signals of the authenticated battery; the current sensing information includes the current, temperature, and internal air pressure during the operation of the battery.

[0032] S3: Determine the predicted value of battery SOH and the estimated value of battery SOC based on the current sensor signals.

[0033] In an exemplary embodiment, S1 can be replaced by the following steps.

[0034] S11: Conduct abuse experiments on the battery under different rates and environmental conditions to obtain the sensing signals during the operation of the battery; the sensing signals include voltage, temperature, and internal air pressure.

[0035] S12: Based on the sensing signals, extract the key features under different rates and environmental conditions; the key features include the inflection point of the air pressure rise in the early stage of thermal runaway and the air pressure change rate.

[0036] S13: Determine the warning threshold and warning lead time under different rates and environmental conditions based on the key features.

[0037] In practical applications, the battery is prone to thermal runaway in the high SOC and low SOH states. The change in internal gas pressure is more sensitive than traditional safety monitoring parameters (such as voltage, current, external temperature) in the early stage of thermal runaway, providing an important basis for the early warning of thermal runaway and the implementation of corresponding safety strategies. In addition, the current main thermal runaway suppression methods mainly stay in the middle and late stages of thermal runaway. The disadvantage is that they cannot obtain the internal state information of the battery, and thus cannot take timely and effective early suppression measures for the occurrence of thermal runaway events. This type of middle and late stage thermal runaway suppression method will still cause irreversible damage to the energy storage system. Therefore, how to obtain the internal state signal of the battery and simultaneously take earlier thermal runaway suppression measures has become an urgent problem to be solved.

[0038] Based on the estimated SOC and SOH above, this application can take thermal runaway suppression measures earlier. Therefore, after S3, it further includes:

[0039] S4: Based on the predicted value of the battery SOH and the estimated value of the battery SOC, according to the currently sensed signal collected in real time, the warning threshold, and the warning lead time, give a warning in the early stage of the battery thermal runaway, and formulate suppression measures according to the actual scenario to suppress the thermal runaway in the early stage.

[0040] In practical applications, as Figure 2 shown, select representative battery cells, such as laminated or wound lithium-ion batteries of the NMC / graphite system. Based on the top cover conformal design, implant a pressure sensor and embed the packaging process inside the battery. Then set the experimental conditions, set multiple groups of experiments respectively, and test the battery under conditions such as charging to exceed the rated capacity (overcharge experiment) at 0.5C, 1C, 2C, etc. C is the charge and discharge rate.

[0041] Use a high-speed data acquisition system to collect the internal air pressure signal of the battery in real time. Synchronously collect traditional monitoring parameters such as voltage, temperature, current, and SOC for comparative analysis and model construction.

[0042] This application provides a pressure warning strategy applicable to different C-rate charging, discharging, and abusive working conditions, enabling significant early warnings in different environments such as 0.5C, 1C, 2C, etc., and ensuring the safe operation of the system.

[0043] As Figure 3 shown, S1 includes a feature extraction and threshold model construction step and a real-time monitoring and online warning judgment logic.

[0044] Among them, the feature extraction and threshold model construction step: Extract key feature points from the real-time air pressure curve, such as the moment of significant rise at the beginning, the point of change in the rising rate, and the steady-state air pressure level. Analyze the time series of the air pressure change rate (dP / dt), especially pay attention to the small air pressure rising section in the early stage of thermal runaway, and statistically analyze its increment amplitude, duration, and lead time before the thermal runaway critical point.

[0045] Based on a large amount of experimental data, the warning thresholds are determined respectively for the overcharge processes at 0.5C, 1C, and 2C rates. For example: Under the condition of 0.5C, when dP / dt reaches about 0.76 kPa / s, a warning can be determined. This moment is about 29 minutes earlier than the traditional temperature rise method. Under the condition of 1C, when dP / dt reaches about 1.49 kPa / s, a warning can be issued, which is about 10 minutes earlier than the traditional scheme. Under the condition of 2C, when dP / dt reaches about 2.2 kPa / s, a warning can be triggered, which is about 4.6 minutes earlier than the traditional warning method. The threshold is further optimized according to the experimental results, so that the warning signal is issued before the thermal runaway has fermented violently, thereby improving the warning sensitivity by 2 - 3 orders of magnitude.

[0046] Real-time monitoring and online warning judgment logic: Store the established threshold model and parameters in the controller. The controller obtains the pressure and dP / dt values from the implanted pressure sensor in real time and compares them with the thresholds under the corresponding working conditions. When it is detected that the rising trend of dP / dt or the absolute pressure reaches or exceeds the set threshold, the controller immediately issues a warning signal. The warning signal can be sent to the host computer or the vehicle controller (if applied to electric vehicles) through the CAN bus, LIN bus or other communication protocols, or it can trigger a buzzer, an LED warning light, a display prompt or a wireless alarm system. After receiving the warning signal, the system can automatically take safety measures, including but not limited to: reducing the charging current, reducing the discharge power, active ventilation or opening the safety valve, emergency disconnection of the circuit, sending alarm information to the external management system, etc.

[0047] In practical applications, S1 includes internal sensor arrangement, data acquisition and processing, feature extraction and model construction, warning threshold setting, and warning output.

[0048] Internal sensor arrangement: implant a high-sensitivity pressure sensor inside the battery to monitor the change of the internal gas pressure of the battery in real time.

[0049] Data acquisition and processing: Conduct charge and discharge, abuse (such as overcharge, etc.) experiments on the battery under different C rates and environmental conditions, and record the change of the air pressure value with time and state through a high-speed data acquisition system.

[0050] Feature extraction and model construction: Analyze the collected air pressure data, extract key feature parameters such as the early air pressure rising inflection point, the air pressure change rate dP / dt, the absolute threshold, and the relative increment threshold, and establish a working condition-related warning model based on a large amount of experimental data.

[0051] Early warning threshold setting: Based on experimental data under different C-rates (such as 0.5C, 1C, 2C, etc.) and different temperature and stress conditions, the early warning threshold is increased by 2 - 3 orders of magnitude compared to traditional voltage, current, or temperature-based early warning methods, and the early warning lead time (for example, about 29 minutes under 0.5C conditions, about 10 minutes under 1C conditions, and about 4.6 minutes under 2C conditions) and the corresponding pressure change rate threshold are determined.

[0052] Early warning output: When the real-time monitored air pressure and dP / dt exceed the preset threshold, an early warning signal is immediately sent to the control system, enabling the system to have sufficient time to take safety measures such as current limiting, temperature reduction, shutdown, shunt, or activation of an explosion-proof valve.

[0053] In an exemplary embodiment, S2 can be replaced by the following steps.

[0054] S21: When the battery is connected to a battery charging device, the battery identity is authenticated, and the authenticated battery is charged; the battery identity is the identity identifier stored in the internal chip of the battery.

[0055] S22: When the authenticated battery is in an indoor charging state, control the battery charging device to stop output and prohibit the battery from charging indoors.

[0056] S23: Control the battery charging device to obtain the current sensing signal of the authenticated battery, and determine whether there is a risk of thermal runaway for the authenticated battery based on the current sensing signal.

[0057] In practical applications, S2 includes battery authentication, authentication matching, charging management, and risk early warning.

[0058] Battery authentication: The battery has a specific chip inside that stores the battery identity ID, which serves as the unique identifier of the battery. Its data cannot be tampered with and is used to identify and verify the authenticity of the battery. When the battery is connected to the charging device, the chip automatically sends a data packet containing identity authentication information to the smart meter.

[0059] Specifically, the charging device can be a smart meter, battery charger, etc., for battery authentication and charging management.

[0060] Authentication matching: After the charging device receives the authentication information sent by the battery, it compares the received battery identity ID data with the compliant battery ID data pre-stored in the smart meter; if there is information in the ID data of the charging device that matches the received battery identity ID, the system confirms that the battery is a compliant battery; otherwise, the system determines that the battery authentication fails and shuts down the charging output.

[0061] Charging management: The charging device determines whether the battery is in a charging state based on the received battery status information. If it detects that charging is taking place indoors, the charging device will stop output and prohibit the battery from charging indoors.

[0062] Risk warning: When the internal air pressure of the battery reaches a threshold and there is a risk of thermal runaway according to the battery status information received by the charging device, a warning is sent to the user.

[0063] In practical applications, as Figure 4 shown, the battery charging device communicates with the battery wirelessly, and the battery charging device can control the indoor power supply by wire, specifically including:

[0064] When the battery is connected to the battery charging device, it sends a data packet containing identity authentication information to the battery charging device for verifying the authenticity of the battery and managing its charging behavior. The battery chip stores a battery identity ID, which serves as the unique identifier of the battery for identifying and verifying the authenticity of the battery. Since the battery identity ID stored in the chip is tamper-proof, the authenticity of the battery can be verified by reading the battery identity ID in the chip to ensure that the battery used is manufactured and certified through regular channels.

[0065] After receiving the authentication information sent by the battery, the battery charging device compares it with the list of battery IDs stored internally to confirm whether it is a preset authenticated battery ID. If there is data in the ID list of the battery charging device that matches the received battery identity ID, the system will confirm that the battery is a compliant battery and thus allow it to proceed with the subsequent charging process. Otherwise, the charging output is turned off.

[0066] The battery charging device further confirms whether the battery is charging through the indoor power supply by obtaining the battery status. If the battery charging device detects that the battery is charging indoors, it will turn off the charging switch. The battery information obtained by the battery charging device is also used for thermal runaway warning. When the internal air pressure of the battery reaches a threshold and there is a risk of thermal runaway by detecting the internal air pressure, a warning is sent to the user.

[0067] This application is based on battery authentication with an implanted chip and is applied to the battery and the charging device. The charging device only performs charging operations on authenticated batteries. In addition, if the verified battery is in an indoor charging state, the charging device will interrupt its output to prevent the battery from charging indoors.

[0068] In an exemplary embodiment, authenticating the battery identity specifically includes:

[0069] S31: Control the battery charging device to receive the authentication information of the battery and compare the battery identity in the authentication information with the compliant battery identities pre-stored in the battery charging device.

[0070] S32: If there is identity information in the compliant battery identity that matches the battery identity in the authentication information, determine that the battery is a compliant battery; the compliant battery is a battery that has passed the authentication.

[0071] S33: If there is no identity information in the compliant battery identity that matches the battery identity in the authentication information, determine that the battery authentication fails, and turn off the charging output.

[0072] In an exemplary embodiment, before S3, it further includes: measuring the internal air pressure of the battery based on the direct measurement method of a pressure sensor or the indirect inversion method based on the strain of an explosion-proof valve; where, as Figure 5 shown, the direct measurement method based on a pressure sensor obtains the internal air pressure of the battery in real time through a pressure sensor 3 placed at the groove 2 of the battery top cover 1, and the signal of this pressure sensor can be transmitted outside the battery; the indirect inversion method based on the strain of an explosion-proof valve determines the internal air pressure of the battery according to the theoretical correction relationship between the strain signal generated by a strain gauge pasted on the surface of the explosion-proof valve piece and the internal air pressure of the battery.

[0073] As Figure 6 shown, the indirect inversion method based on the strain of an explosion-proof valve is specifically the following method:

[0074] The strain gauge 5 is pasted on the surface of the explosion-proof valve 4, and based on the theoretical correction relationship between the strain signal generated by the strain gauge and the internal air pressure of the battery, the internal air pressure of the battery is determined based on the strain of the explosion-proof valve.

[0075] The stress on the explosion-proof valve is jointly caused by the internal air pressure of the battery and the external air pressure acting on the explosion-proof valve, and it can be obtained that:

[0076] σ valve =P - P0

[0077] where, σ valve is the stress on the explosion-proof valve; P is the internal air pressure of the battery; P0 is the standard atmospheric pressure.

[0078] According to Hooke's law of elasticity:

[0079]

[0080] where, ε1 is the strain in the same direction as the strain gauge, σ1 is the stress in the same direction as the strain gauge; σ2 is the stress perpendicular to the surface of the explosion-proof valve, and it can be considered that σ2 = σ valve ; σ3 is the stress in the direction perpendicular to both σ1 and σ2; E is the elastic modulus of the explosion-proof valve; μ is the Poisson's ratio of the explosion-proof valve.

[0081] In the safe state of the battery, it is approximately considered that the explosion-proof valve has only one direction of principal strain, that is, ε2 = ε3 = 0. Then there is:

[0082]

[0083] Combining the above equations, the theoretical relationship between the strain of the explosion-proof valve and the internal air pressure is obtained:

[0084]

[0085] Considering that the existence of the above assumptions will lead to deviations in the theoretical relationship, a correction coefficient K0 is introduced to obtain the theoretical correction relationship between the strain of the explosion-proof valve and the internal air pressure:

[0086]

[0087] The correction coefficient K0 can be determined through the following experimental methods:

[0088] Paste the metal strain gauge on the explosion-proof valve of a certain type of hard-shell battery, and place a pressure sensor in the groove of the battery top cover at the same time. Conduct charge and discharge tests on the hard-shell battery, and record the internal air pressure sensing data P and the strain data ε1 of the explosion-proof valve of the battery. Calculate the slope through linear fitting, and then calculate the correction coefficient according to the above-mentioned theoretical correction relationship. To make the correction coefficient more in line with the actual battery, through multiple groups of experiments, the average value of the correction coefficients of multiple groups of experiments is obtained to obtain the air pressure correction coefficient K0 of this type of hard-shell battery.

[0089] According to the ideal gas state equation:

[0090] PV = nRT

[0091] Among them, V is the volume of the internal cavity of the battery; n is the amount of substance of the gas; R is the molar gas constant; T is the temperature of the internal cavity of the battery.

[0092] According to the above formula, it can be seen that the internal air pressure P of the battery is affected by the volume V of the internal cavity of the battery, the amount of gas production n of the battery, and the temperature T of the internal cavity of the battery. Among them, the volume of the battery cavity and the amount of gas production are closely related to the health status of the battery. Therefore, the air pressure changes caused by these two factors can be classified into one category, denoted as the air pressure change value P caused by gas production due to battery aging gas .

[0093] Based on the indirect inversion method of the strain of the explosion-proof valve, the theoretical correction relationship between the strain of the explosion-proof valve and the internal air pressure is constructed. The internal air pressure of the battery can be obtained through the strain of the explosion-proof valve, which does not require a large amount of space and has great application value.

[0094] By implanting a highly sensitive pressure sensor through conformal design inside the battery top cover, the internal air pressure change and its change rate (dP / dt) are monitored in real time, and early accurate warning of thermal runaway is realized according to the threshold model established under different working conditions.

[0095] In an exemplary embodiment, S3 can be replaced by the following steps.

[0096] S41: Determine the predicted value of the battery SOH based on the internal air pressure and the internal cavity temperature of the battery.

[0097] S42: Determine the estimated value of the battery SOC based on the current, temperature, and internal air pressure during the operation of the battery.

[0098] In an exemplary embodiment, S41 can be replaced by the following steps.

[0099] S411: Decouple the internal air pressure of the battery according to the temperature-pressure correspondence relationship and the internal cavity temperature to determine the air pressure change value caused by temperature; the temperature-pressure correspondence relationship is a mathematical model of the temperature and pressure change relationship obtained by discharging the battery to 0% SOC, in an open-circuit state, placing it in a thermostat, setting a temperature point every set temperature within the set temperature range, and recording the internal air pressure of the battery after maintaining a constant temperature for a set period of time at each temperature point.

[0100] In practical applications, the temperature-pressure correspondence relationship is as follows:

[0101] Discharge the hard-shell battery to be tested to 0% SOC, place it in a thermostat in an open-circuit state, set a temperature point every 5°C within the temperature range of 15°C to 45°C, maintain a constant temperature for 4 hours at each temperature point, record the internal air pressure at this time, obtain the mathematical model P = f(T) of the temperature and pressure change relationship, and deduce the change trend of the internal air pressure P of the battery from the gas temperature T.

[0102] As Figure 7 shown, by decoupling the internal air pressure P of the battery, the air pressure curves caused by the battery temperature and the air pressure caused by aging are obtained under long-cycle working conditions.

[0103] S412: Determine the air pressure change value caused by battery aging gas production according to the internal air pressure of the battery and the air pressure change value.

[0104] In practical applications, in order to obtain accurate battery health characteristics, after obtaining the internal air pressure of the battery, it is necessary to decouple the obtained internal air pressure P to remove the air pressure change value P T caused by temperature and obtain the air pressure change value P gas caused by battery aging gas production.

[0105] S413: Input the air pressure change value caused by battery aging gas production into the machine learning model to determine the predicted value of the battery SOH.

[0106] In practical applications, the machine learning model is trained in the following way:

[0107] Using P gas and the cycle capacity as the input of the training samples, a battery SOH prediction method based on the Gaussian Process Regression (GPR) model is provided. By taking the air pressure health index as the input, combining the Gaussian kernel function and the linear mean function, a prediction model is constructed, and then the high-precision prediction of the battery SOH is realized. By setting the length scale and variance parameters of the Gaussian kernel, the fitting ability of the model can be flexibly adjusted, so as to optimize the prediction performance. After the prediction is completed, the prediction accuracy of the model is evaluated according to the calculated Root-Mean-Square Error (RMSE), and the prediction effect is intuitively shown by comparing the curves of the actual and predicted SOH.

[0108] As Figure 8 shown, input P gas into the trained machine learning model to obtain the battery SOH estimation curve based on the internal air pressure signal.

[0109] A new health feature for SOH estimation is proposed based on the internal air pressure signal. The SOH estimation accuracy of the aged air pressure after removing the temperature influence is improved by one order of magnitude compared with the previous SOH estimation based on the electrical signal.

[0110] The above battery SOH estimation process based on the internal air pressure is applicable to the health status monitoring and prediction in the battery management system, which helps to extend the battery life, optimize the battery maintenance cycle, and reduce the usage cost.

[0111] In an exemplary embodiment, S42 can be replaced by the following steps.

[0112] S421: Calculate the air pressure change rate according to the time series data of the internal air pressure of the battery.

[0113] S422: Establish a polynomial regression model according to the air pressure change rate.

[0114] S423: Calibrate the polynomial regression model by using the current, temperature, SOC and air pressure change rate under known conditions.

[0115] S424: Use the calibrated polynomial regression model to determine the battery SOC estimation value according to the real-time measured current, temperature and air pressure change rate under unknown SOC conditions.

[0116] In practical applications, such as Figure 9As shown, considering the quadratic non - linear relationships of various variables and the first - order non - linear relationships between variables, using the least - squares method as the solution algorithm to solve the polynomial regression model, the identified parameters include: a(I)= - 0.0041005; b(T)=0.024732; c(SOC)= - 0.30028; d(I 2 )= - 1.228×10 -6 ; e(T 2 )= - 0.00044168; f(SOC 2 )=0.0015877; g(I*T)=0.00012795; h(I*SOC)= - 0.00099728; i(T*SOC)=0.0086595; j(constant term)= - 0.33038; goodness of fit R 2 =0.9322; root mean square error (RMSE)=0.0005.

[0117] S42 includes data acquisition and pre - processing, feature modeling, model parameter calibration and training, and SOC inversion.

[0118] Among them, data acquisition and pre - processing: Conduct experimental tests on single - cell batteries under different SOC, temperature, and current conditions, and record the internal pressure P, current I, temperature T, known SOC, and time t at corresponding moments. Perform pre - processing operations such as filtering, cleaning, and normalization on the collected data.

[0119] Feature modeling: Using the rate of change of pressure dP / dt as the target variable, taking I, T, SOC, and their square terms, cross - terms (I 2 , T 2 , SOC 2 , I*T, I*SOC, T*SOC) and other parameters as independent variables to establish a polynomial regression model. The model form is:

[0120]

[0121] where a, b, c, d, e, f, g, h, i, j are regression coefficients.

[0122] Model parameter calibration and training: Using the experimental data obtained in data acquisition and pre - processing, solve the above - mentioned model parameters (a - j) through linear regression or other optimization methods (such as least - squares, regularization regression, or Bayesian regression) to obtain a set of optimal regression coefficients, so that the model has a high goodness of fit (R 2 ) and a low error (such as root mean square error RMSE) on the sample data set.

[0123] Apply the calibrated model to an independent test dataset, compare the dP / dt output by the model with the actually measured dP / dt, and test the applicability and accuracy of the model. If the error is large, the polynomial order can be increased, the cross terms can be increased or decreased, or nonlinear regression techniques can be used for optimization.

[0124] Inverse calculation of SOC: During actual use, online measure the rate of change of the internal air pressure of the battery dP / dt, current I, and temperature T. Substitute the measured dP / dt into the model. Solve for the SOC variable in this equation (it can be transformed into a standard quadratic or higher-order equation according to the actual situation, and then the root can be obtained by numerical methods) to obtain the SOC estimated value at the current moment. If the equation is relatively complex, an iterative numerical solution algorithm (such as the Newton-Raphson method) can be used to update the SOC value according to the currently known parameters (I, T, dP / dt).

[0125] Online correction and adaptive update: During long-term operation, the model parameters can be recalibrated periodically, or when error accumulation is detected, compare the SOC with the reference SOC (such as the full charge point, full discharge point, or fused with other estimation methods) to adaptively update the regression parameters or the SOC estimated value.

[0126] An SOC estimation device based on the above SOC estimation process includes: a pressure sensor, a temperature sensor, a current sensor, and a control processing unit. The control processing unit is used to receive the sensor signals and run the above polynomial regression model, thereby outputting the real-time SOC estimated value.

[0127] Through the above technical solution, the method described in this application utilizes the new measurement parameter of the internal air pressure of the battery, combines the traditional current and temperature signals, and performs multi-parameter fitting and estimation on the SOC, which is beneficial to improving the SOC estimation accuracy and robustness.

[0128] This application measures the rate of change of the internal air pressure of the battery cell over time, and establishes a polynomial fitting model for the nonlinear relationship between this change and the current I, temperature T, and state of charge SOC, thereby achieving high-precision SOC estimation.

[0129] This application improves the SOC estimation accuracy, supplements the traditional estimation method with the new parameter of the internal air pressure, and can better capture the internal reaction characteristics of the battery; it can also improve the adaptability and reliability. Through the nonlinear polynomial regression model, it can adapt to the SOC estimation of different battery types and different working conditions.

[0130] This application predicts the SOC based on the change of the internal air pressure of the battery, and can complete the estimation using relatively simple measurement signals (pressure, temperature, current), without the need for complex observer design.

[0131] In an exemplary embodiment, in S4, formulating suppression measures according to the actual scenario for early suppression of thermal runaway can be replaced by the following steps.

[0132] S51: When the actual scenario is an energy storage power station scenario, based on the internally monitored battery pressure in real time, locate the position of the abnormal battery PACK according to the warning signal, and determine whether the warning signal is continuously abnormal. If so, execute S52; if not, execute S53.

[0133] S52: Control the inspection robot to move to the position of the abnormal battery PACK, transport the abnormal battery PACK to the designated disposal location, and introduce it into the injection pipeline 6 through the preset pipeline inlet hole 7 of the abnormal battery PACK housing for suppression operation, as Figure 10 shown.

[0134] S53: Continue to monitor the internally battery pressure.

[0135] In practical applications, when the actual scenario is an energy storage power station scenario, as Figure 11 shown, first, monitor the internally battery pressure in real time. When an abnormality occurs, the power station warning system issues a warning through the pressure sensing signal and locates it to the specific battery PACK position. The inspection robot moves to the warning position, removes the PACK connection parts, and extracts the abnormal PACK, then observes whether the signal continues to change abnormally after standing. If there is no abnormality, it is restored to its original position and continues to operate; if the abnormal signal continues to occur, proceed to the next step.

[0136] Secondly, the robot quickly transports the abnormal PACK to the designated disposal location.

[0137] Finally, locate the abnormal battery in the abnormal PACK, introduce the injection pipeline through the preset pipeline inlet hole of the PACK housing, send it to the abnormal battery position, and spray and suppress the cooling medium at the pole column of the abnormal battery cell.

[0138] In an exemplary embodiment, in S4, formulating suppression measures according to the actual scenario for early suppression of thermal runaway can be replaced by the following steps.

[0139] S61: When the actual scenario is an electric vehicle scenario, based on the internally monitored battery pressure in real time, control the in-vehicle warning system to issue a warning according to the pressure sensing signal, and determine whether the pressure sensing signal is continuously abnormal. If so, execute S62; if not, execute S63.

[0140] S62: Turn on all vehicle electrical equipment to consume electrical energy, and expose the abnormal battery and suppress the abnormal battery by means of overall removal of the battery pack cover or destructive removal of a partial cover.

[0141] S63: Continuously monitor the internal air pressure of the battery.

[0142] In practical applications, when the actual scenario is an electric vehicle scenario, as Figure 12 shown, first, continuously monitor the internal air pressure of the battery. When an abnormality occurs, the in-vehicle warning system issues a warning through the air pressure sensing signal.

[0143] Secondly, the driver pulls over to the side of the road to observe whether the observation signal continues to be abnormal. If it continues to be abnormal, turn on all the vehicle's electrical equipment to run, continuously consume electrical energy and stay away from the vehicle, and call the fire department to the scene.

[0144] Finally, when the firefighters arrive at the scene, use a jack to lift the vehicle, and expose the abnormal battery by means of overall removal of the battery pack cover plate or destructive removal of a partial cover plate. Then, complete the early suppression of the thermal runaway of the abnormal battery by spraying the cooling medium at the battery terminal.

[0145] Compared with the traditional warnings based on temperature and voltage, the air pressure warning signal in this application appears earlier and can issue a warning several minutes to dozens of minutes before the thermal runaway occurs.

[0146] The sensitivity of the warning threshold in this application is increased by 2-3 orders of magnitude, which is more conducive to accurately identifying early minute abnormalities.

[0147] This application is applicable to a variety of working conditions (different rates, different operating environments), has universality and reliability, is conducive to integration in the battery system, and realizes real-time online safety monitoring.

[0148] This application is applicable to a variety of application scenarios (energy storage, electric vehicles), has universality and reliability, and is conducive to eliminating the potential safety hazards brought by the battery in the early stage.

[0149] This application can accurately locate the battery according to the sensing signal, avoiding the operation of suppressing the entire battery pack due to the abnormality of one battery, and reducing the cost of battery danger suppression.

[0150] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combinations of these technical features do not conflict, they should all be considered as the scope described in this specification.

[0151] Specific examples are used in this article to elaborate on the principle and implementation manner of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, based on the idea of this application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to this application.

Claims

1. A method for estimating battery SOX based on internal sensing signals, characterized in that, The battery SOX estimation method based on internal sensing signals includes: Conduct abuse experiments on the battery under different rates and environmental conditions to determine the warning threshold and warning lead time; Authenticate the battery through security authentication and obtain the current sensing signals of the authenticated battery; the current sensing information includes the current, temperature, and internal air pressure during the operation of the battery; Determine the battery SOH prediction value and battery SOC estimation value according to the current sensor signals.

2. The battery SOX estimation method based on internal sensing signals according to claim 1, wherein Conduct abuse experiments on the battery under different rates and environmental conditions to determine the warning threshold and warning lead time, specifically including: Conduct abuse experiments on the battery under different rates and environmental conditions to obtain the sensing signals during the operation of the battery; the sensing signals include voltage, temperature, and internal air pressure of the battery; Based on the sensing signals, extract the key features under different rates and environmental conditions; the key features include the inflection point of the air pressure rise in the early stage of thermal runaway and the air pressure change rate; Determine the warning threshold and warning lead time under different rates and environmental conditions according to the key features.

3. The battery SOX estimation method based on internal sensing signals according to claim 1, characterized in that, Authenticate the battery through security authentication and obtain the current sensing signals of the authenticated battery, specifically including: When the battery is connected to the battery charging device, authenticate the battery identity and charge the authenticated battery; the battery identity is the identity identifier stored in the internal chip of the battery; among them, authenticating the battery identity specifically includes: Control the battery charging device to receive the authentication information of the battery and compare the battery identity in the authentication information with the compliant battery identities pre-stored in the battery charging device; If there is an identity information in the compliant battery identities that matches the battery identity in the authentication information, determine that the battery is a compliant battery; the compliant battery is the authenticated battery; If there is no identity information in the compliant battery identities that matches the battery identity in the authentication information, determine that the battery authentication fails and turn off the charging output; When the authenticated battery is in the indoor charging state, control the battery charging device to stop output and prohibit the battery from charging indoors; Control the battery charging device to obtain the current sensing signals of the authenticated battery and determine whether there is a thermal runaway risk for the authenticated battery according to the current sensing signals.

4. The method for estimating the battery SOX based on the internal sensing signal according to claim 1, wherein Before determining the battery SOH prediction value and battery SOC estimation value according to the current sensor signals, it also includes: Measure the internal air pressure of the battery by the direct measurement method based on a pressure sensor or the indirect inversion method based on the strain of the explosion-proof valve. Among them, the direct measurement method based on a pressure sensor obtains the internal air pressure of the battery in real time through a pressure sensor placed in the groove of the battery top cover. The indirect inversion method based on the strain of the explosion-proof valve determines the internal air pressure of the battery according to the theoretical correction relationship between the strain signal generated by the strain gauge pasted on the surface of the explosion-proof valve piece and the internal air pressure of the battery. The theoretical correction relationship is P is the internal air pressure of the battery; ε1 is the measured strain of the strain gauge; E is the elastic modulus of the explosion-proof valve; μ is the Poisson's ratio of the explosion-proof valve; P0 is the standard atmospheric pressure; K0 is the correction coefficient.

5. The method for estimating battery SOX based on internal sensing signals according to claim 1, wherein Determine the battery SOH prediction value and battery SOC estimation value according to the current sensor signals, specifically including: Determine the battery SOH prediction value according to the internal air pressure and internal cavity temperature of the battery; Determine the battery SOC estimation value according to the current, temperature, and internal air pressure during the operation of the battery.

6. The method for estimating battery SOX based on internal sensing signals according to claim 5, characterized in that, Determine the battery SOH prediction value according to the internal air pressure and internal cavity temperature of the battery, specifically including: Decouple the internal air pressure of the battery according to the temperature-pressure correspondence relationship and the internal cavity temperature, and determine the air pressure change value caused by temperature; the temperature-pressure correspondence relationship is a mathematical model of the temperature and pressure change relationship obtained by discharging the battery to 0% SOC, in an open circuit state, placing it in a thermostat, setting a temperature point at each set temperature within the set temperature range, and recording the internal air pressure of the battery after maintaining a constant temperature for a set period of time at each temperature point. Determine the air pressure change value caused by battery aging gas production according to the internal air pressure of the battery and the air pressure change value. Input the air pressure change value caused by battery aging gas production into the machine learning model to determine the battery SOH prediction value.

7. The battery SOX estimation method based on internal sensing signals according to claim 5, wherein Determine the battery SOC estimation value according to the current, temperature and internal air pressure of the battery during operation, specifically including: Calculate the air pressure change rate according to the time series data of the internal air pressure of the battery. Establish a polynomial regression model according to the air pressure change rate. Calibrate the polynomial regression model by using the current, temperature, SOC and air pressure change rate under known conditions. Use the calibrated polynomial regression model to determine the battery SOC estimation value according to the real-time measured current, temperature and air pressure change rate under unknown SOC conditions.

8. The battery SOX estimation method based on internal sensing signals according to claim 1, characterized in that Determine the battery SOH prediction value and the battery SOC estimation value according to the current sensor signal. After that, it further includes: Based on the battery SOH prediction value and the battery SOC estimation value, according to the real-time collected current sensor signal, the warning threshold and the warning lead time, give a warning in the early stage of the battery thermal runaway, and formulate a suppression measure according to the actual scenario to suppress the thermal runaway in the early stage.

9. The battery SOX estimation method based on internal sensing signals according to claim 8, wherein, Formulate a suppression measure according to the actual scenario to suppress the thermal runaway in the early stage, specifically including: When the actual scenario is an energy storage power station scenario, based on the real-time monitored internal air pressure of the battery, locate the position of the abnormal battery PACK according to the warning signal, and judge whether the warning signal is continuously abnormal. If so, control the inspection robot to move to the position of the abnormal battery PACK, transport the abnormal battery PACK to the designated disposal location, and introduce it into the injection pipeline through the preset pipeline inlet hole of the abnormal battery PACK housing for suppression operation. If not, continue to monitor the internal air pressure of the battery.

10. The method for estimating the battery state of health (SOH) based on internal sensing signals according to claim 8, wherein Formulate a suppression measure according to the actual scenario to suppress the thermal runaway in the early stage, specifically including: When the actual scenario is an electric vehicle scenario, based on the real-time monitored internal air pressure of the battery, control the in-vehicle warning system to give a warning according to the air pressure sensor signal, and judge whether the air pressure sensor signal is continuously abnormal. If so, turn on all the vehicle electrical equipment to consume electric energy, and adopt the method of overall removal of the battery pack cover or destructive removal of the local cover to expose the abnormal battery and suppress the abnormal battery. If not, continue to monitor the internal air pressure of the battery.

Images