A lithium iron phosphate battery thermal runaway trend prediction method and system

Abstract

The application provides a lithium iron phosphate battery thermal runaway trend prediction method and system. The method comprises the following steps: acquiring real-time operation data of a lithium iron phosphate battery to be predicted, initial estimated values of parameters of an equivalent circuit model, a basic process noise covariance matrix and a basic measurement noise covariance matrix within a preset time length; determining an adjustment factor, and determining an adjusted process noise covariance matrix and an adjusted basic measurement noise covariance matrix based on the adjustment factor; optimizing the initial estimated values of the parameters of the equivalent circuit model corresponding to the lithium iron phosphate battery by using an extended Kalman filtering algorithm, and obtaining an optimized ohmic resistance, a first-stage polarization resistance and a second-stage polarization resistance; determining a temperature rise rate of the lithium iron phosphate battery to be predicted at a current time t and a predicted temperature at a time t+1; and performing thermal runaway trend prediction on the lithium iron phosphate battery. The technical scheme provided by the application improves the reliability, sensitivity and timeliness of the prediction result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery management, and particularly relates to a lithium iron phosphate battery thermal runaway trend prediction method and system. BACKGROUND

[0002] Lithium iron phosphate (LiFePO4) batteries have been widely used in electric vehicles, energy storage systems and other fields due to their high safety performance, long cycle life and low cost. However, lithium iron phosphate batteries may still experience thermal runaway under overcharge, overdischarge, internal short circuit or external high temperature, leading to serious safety accidents such as combustion and explosion. Therefore, accurate prediction of the thermal runaway trend of lithium iron phosphate batteries is crucial for ensuring the safe operation of battery systems.

[0003] The method of predicting the thermal runaway trend of lithium iron phosphate batteries by using the extended Kalman filter algorithm to identify the parameters of the equivalent circuit model online has inherent defects when the battery undergoes specific and common working condition transitions. Specifically, when the lithium iron phosphate battery is subjected to a dynamic operating condition of a sustained current fluctuation (frequent acceleration and deceleration and start-stop of a vehicle under urban traffic congestion, or rapid power throughput of a storage system in response to grid instructions), and then transitions to a low-power output condition of small current continuous operation or near static (the vehicle enters a stable low-speed cruising state, or the storage system enters standby or small power maintenance mode), the standard extended Kalman filter algorithm cannot adapt to the changes in model prediction uncertainty and voltage measurement signal-to-noise ratio caused by the transition from intense to gentle working condition characteristics, due to its usual setting mechanism of static process noise covariance matrix Q and measurement noise covariance matrix R. This inadaptability directly leads to a decrease in the identification accuracy and a slow convergence speed of the battery internal resistance parameter in the initial stage of the working condition transition and the subsequent low-power operation stage. The internal resistance identification deviation introduced by model simplification and excitation changes in the previous dynamic working condition cannot be quickly and effectively corrected in this stage. Since the internal resistance is a key physical quantity for evaluating the thermal stability and early thermal runaway risk of the battery, this internal resistance identification deviation problem that occurs under common working condition transitions significantly weakens the accuracy and timeliness of the thermal runaway early warning system that relies on the internal resistance, leading to a deviation in the evaluation of the potential thermal runaway risk of the battery, increasing the probability of missed or delayed reporting of thermal runaway events, and threatening the safe operation of the battery system. Therefore, there is an urgent need to propose a scheme for accurately predicting the thermal runaway of the battery. SUMMARY

[0004] The present application provides a lithium iron phosphate battery thermal runaway trend prediction method and system to at least solve the technical problem of deviation in the evaluation of the potential thermal runaway risk of the battery, increasing the probability of missed or delayed reporting of thermal runaway events, and threatening the safe operation of the battery system.

[0005] The first aspect embodiment of the present application provides a lithium iron phosphate battery thermal runaway trend prediction method, the method comprises:

[0006] Obtain real-time operation data of a lithium iron phosphate battery to be predicted within a preset time period, initial estimated values of parameters of an equivalent circuit model corresponding to the lithium iron phosphate battery, a basic process noise covariance matrix and a basic measurement noise covariance matrix of the lithium iron phosphate battery, wherein the real-time operation data comprises working current and key point temperature of the battery, and the initial estimated values of the parameters comprise initial estimated values of ohmic resistance, first-stage polarization resistance, second-stage polarization resistance and state of charge;

[0007] Determine an adjustment factor of the basic process noise covariance matrix and a voltage noise adjustment factor of the basic measurement noise covariance matrix according to the working current of the lithium iron phosphate battery to be predicted within the preset time period, and determine an adjusted process noise covariance matrix based on the adjustment factor of the basic process noise covariance matrix, and determine an adjusted basic measurement noise covariance matrix based on the voltage noise adjustment factor of the basic measurement noise covariance matrix;

[0008] Optimize the initial estimated values of the parameters of the equivalent circuit model corresponding to the lithium iron phosphate battery based on the adjusted process noise covariance matrix and the adjusted measurement noise covariance matrix, and utilize an extended Kalman filtering algorithm to obtain optimized ohmic resistance, first-stage polarization resistance, second-stage polarization resistance and state of charge;

[0009] Determine the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t and the predicted temperature at the time t+1 according to the working current, the key point temperature, the optimized ohmic resistance, the first-stage polarization resistance, the second-stage polarization resistance and the state of charge;

[0010] Perform thermal runaway trend prediction on the lithium iron phosphate battery based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature at the time t+1 and the key point temperature.

[0011] Preferably, the method further comprises:

[0012] Determine the average working current absolute value of the lithium iron phosphate battery to be predicted within the preset time period based on the working current of the lithium iron phosphate battery to be predicted within the preset time period, and judge whether the basic measurement noise covariance matrix needs to be adjusted based on the average working current absolute value;

[0013] If the base measurement noise covariance matrix needs to be adjusted, a voltage noise adjustment factor of the base measurement noise covariance matrix is determined, otherwise the base measurement noise covariance matrix is taken as the adjusted measurement noise covariance matrix.

[0014] Further, the equivalent circuit model corresponding to the lithium iron phosphate battery is a second-order model.

[0015] Further, the adjustment factor of the base process noise covariance matrix is determined according to the working current of the lithium iron phosphate battery to be predicted within a preset time length, comprising:

[0016] determining the current standard deviation and the average of the current change rate of the lithium iron phosphate battery to be predicted within a preset time length based on the working current of the lithium iron phosphate battery to be predicted within the preset time length;

[0017] determining a comprehensive dynamic index within the preset time length according to the current standard deviation and the current change rate of the lithium iron phosphate battery to be predicted within the preset time length, and then determining the adjustment factor of the base process noise covariance matrix according to the comprehensive dynamic index.

[0018] Further, the calculation formula of the comprehensive dynamic index is as follows:

[0019]

[0020] In the formula, is the comprehensive dynamic index, is the current standard deviation, is the average of the current change rate, is the rated current of the lithium iron phosphate battery, is the sampling time interval of the extended Kalman filtering algorithm;

[0021] The calculation formula of the adjustment factor of the base process noise covariance matrix is as follows:

[0022]

[0023] In the formula, is the adjustment factor of the base process noise covariance matrix, is the sensitivity parameter, is the working condition dynamic saturation reference point.

[0024] Further, the calculation formula of the voltage noise adjustment factor of the base measurement noise covariance matrix is as follows:

[0025]

[0026] In the formula, a voltage noise adjustment factor for the base measurement noise covariance matrix, a preset current judgment threshold, an average working current absolute value, an adjustment intensity factor for the base measurement noise covariance matrix, a function, when the condition is met, the value is , otherwise .

[0027] Further, the calculation formula of the adjusted process noise covariance matrix is as follows:

[0028]

[0029] In the formula, is the adjusted process noise covariance matrix, is the base process noise covariance matrix;

[0030] The calculation formula of the adjusted base measurement noise covariance matrix is as follows:

[0031]

[0032] In the formula, is the adjusted base measurement noise covariance matrix, is the base measurement noise variance of the terminal voltage, is the base measurement noise variance of the terminal voltage.

[0033] Further, the determination of the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t and the predicted temperature at the time t+1 according to the working current, the key point temperature, the optimized ohmic resistance, the first level polarization resistance, the second level polarization resistance and the state of charge comprises:

[0034] determining the total heat generation rate at the current time t according to the working current, the key point temperature, the optimized ohmic resistance, the first level polarization resistance, the second level polarization resistance and the state of charge;

[0035] inputting the total heat generation rate at the current time t into a first-order lumped parameter thermal model, and solving the first-order lumped parameter thermal model to obtain the temperature rise rate at the current time t and the predicted temperature at the time t+1.

[0036] Preferably, the thermal runaway trend prediction of the lithium iron phosphate battery based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature at the time t+1 and the key point temperature comprises:

[0037] ​determining whether the temperature rise rate at the current time t is greater than a preset temperature rise rate threshold, and / or whether the predicted temperature at the t+1 time is greater than or equal to a preset thermal runaway critical point threshold, and / or whether the key point temperature is greater than a preset temperature threshold, if yes, determining that the lithium iron phosphate battery has a thermal runaway trend, otherwise, the lithium iron phosphate battery does not have a thermal runaway trend.

[0038] The second aspect embodiment of the present application provides a lithium iron phosphate battery thermal runaway trend prediction system, comprising:

[0039] An acquisition module is configured to acquire real-time operation data of a lithium iron phosphate battery to be predicted, initial estimated values of parameters of an equivalent circuit model corresponding to the lithium iron phosphate battery, a basic process noise covariance matrix and a basic measurement noise covariance matrix of the lithium iron phosphate battery within a preset time length, wherein the real-time operation data comprises a working current and a key point temperature of the battery, and the initial estimated values of the parameters comprise an initial estimated value of an ohmic resistance, an initial estimated value of a first-stage polarization resistance, an initial estimated value of a second-stage polarization resistance, and an initial estimated value of a state of charge.

[0040] A first determination module is configured to determine an adjustment factor of the basic process noise covariance matrix and a voltage noise adjustment factor of the basic measurement noise covariance matrix according to the working current of the lithium iron phosphate battery to be predicted within the preset time length, and determine an adjusted process noise covariance matrix based on the adjustment factor of the basic process noise covariance matrix and determine an adjusted basic measurement noise covariance matrix based on the voltage noise adjustment factor of the basic measurement noise covariance matrix.

[0041] An optimization module is configured to optimize the initial estimated values of the parameters of the equivalent circuit model corresponding to the lithium iron phosphate battery based on the adjusted process noise covariance matrix and the adjusted measurement noise covariance matrix, and obtain optimized ohmic resistance, first-stage polarization resistance, second-stage polarization resistance, and state of charge by using an extended Kalman filtering algorithm.

[0042] A second determination module is configured to determine a temperature rise rate at a current time t and a predicted temperature at a t+1 time of the lithium iron phosphate battery to be predicted according to the working current, the key point temperature, the optimized ohmic resistance, first-stage polarization resistance, second-stage polarization resistance, and state of charge.

[0043] A prediction module is configured to perform thermal runaway trend prediction on the lithium iron phosphate battery based on the temperature rise rate at the current time t, the predicted temperature at the t+1 time, and the key point temperature of the lithium iron phosphate battery to be predicted.

[0044] The technical scheme provided by the embodiments of the present application at least brings the following beneficial effects:

[0045] The application provides a lithium iron phosphate battery thermal runaway trend prediction method and system. The method comprises the following steps: acquiring real-time operation data of a lithium iron phosphate battery to be predicted within a preset time period, initial estimated values of parameters of an equivalent circuit model corresponding to the lithium iron phosphate battery, a basic process noise covariance matrix of the lithium iron phosphate battery and a basic measurement noise covariance matrix of the lithium iron phosphate battery, wherein the real-time operation data comprises working current and key point temperature of the battery; the initial estimated values of the parameters comprise initial estimated values of ohmic resistance, first-stage polarization resistance and second-stage polarization resistance; determining an adjustment factor of the basic process noise covariance matrix and a voltage noise adjustment factor of the basic measurement noise covariance matrix according to the working current of the lithium iron phosphate battery to be predicted within the preset time period, and determining an adjusted process noise covariance matrix based on the adjustment factor of the basic process noise covariance matrix, and determining an adjusted basic measurement noise covariance matrix based on the voltage noise adjustment factor of the basic measurement noise covariance matrix; optimizing the initial estimated values of the parameters of the equivalent circuit model corresponding to the lithium iron phosphate battery based on the adjusted process noise covariance matrix and the adjusted measurement noise covariance matrix, and obtaining optimized ohmic resistance, first-stage polarization resistance, second-stage polarization resistance and state of charge by using an extended Kalman filtering algorithm; determining a temperature rise rate of the lithium iron phosphate battery to be predicted at a current time t and a predicted temperature of the lithium iron phosphate battery to be predicted at a time t+1 according to the working current, the key point temperature, the optimized ohmic resistance, the first-stage polarization resistance, the second-stage polarization resistance and the state of charge; and performing thermal runaway trend prediction on the lithium iron phosphate battery based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature of the lithium iron phosphate battery to be predicted at the time t+1 and the key point temperature. The technical scheme provided by the application improves the accuracy and real-time performance of battery resistance identification, and further improves the reliability and precision of thermal runaway prediction.

[0046] Additional aspects and advantages of the application will be made apparent by the following description and the appended claims. BRIEF DESCRIPTION OF DRAWINGS

[0047] The above and / or additional aspects and advantages of the application will become apparent and be readily appreciated from the following description, including the accompanying drawings, wherein:

[0048] Figure 1 A flowchart of a lithium iron phosphate battery thermal runaway trend prediction method according to an embodiment of the application is provided.

[0049] Figure 2 A structural diagram of a lithium iron phosphate battery thermal runaway trend prediction system according to an embodiment of the application is provided. DETAILED DESCRIPTION

[0050] Embodiments of the present application are described below in detail, examples of which are shown in the drawings, wherein the same or similar notations represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by reference to the drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0051] The present application provides a lithium iron phosphate battery thermal runaway trend prediction method and system, the method comprising: obtaining real-time operation data of a lithium iron phosphate battery to be predicted within a preset time period, initial estimated values of each parameter of an equivalent circuit model corresponding to the lithium iron phosphate battery, a basic process noise covariance matrix and a basic measurement noise covariance matrix of the lithium iron phosphate battery, wherein the real-time operation data includes working current and key point temperature of the battery; the initial estimated values of each parameter include initial estimated values of ohmic resistance, first-stage polarization resistance and second-stage polarization resistance; determining an adjustment factor of the basic process noise covariance matrix and a voltage noise adjustment factor of the basic measurement noise covariance matrix based on the working current of the lithium iron phosphate battery to be predicted within the preset time period, and determining an adjusted process noise covariance matrix based on the adjustment factor of the basic process noise covariance matrix, and determining an adjusted basic measurement noise covariance matrix based on the voltage noise adjustment factor of the basic measurement noise covariance matrix; optimizing the initial estimated values of each parameter of the equivalent circuit model corresponding to the lithium iron phosphate battery based on the adjusted process noise covariance matrix and the adjusted measurement noise covariance matrix, and using an extended Kalman filtering algorithm to obtain optimized ohmic resistance, first-stage polarization resistance and second-stage polarization resistance; determining the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t and the predicted temperature at the time t+1 based on the working current, the key point temperature, the optimized ohmic resistance, the first-stage polarization resistance, the second-stage polarization resistance and the state of charge; and performing thermal runaway trend prediction on the lithium iron phosphate battery based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature at the time t+1 and the key point temperature. The technical solution provided by the present application improves the accuracy and real-time performance of battery resistance identification, and further improves the reliability and precision of thermal runaway prediction.

[0052] A lithium iron phosphate battery thermal runaway trend prediction method and system according to an embodiment of the present application are described below with reference to the drawings.

[0053] Embodiment one

[0054] Figure 1 A flowchart of a lithium iron phosphate battery thermal runaway trend prediction method according to an embodiment of the present application is shown in FIG. 1, the method comprising: Figure 1 ​

[0055] Step 1: Obtain the real-time operating data of the lithium iron phosphate battery to be predicted within a preset time period, the initial estimated values ​​of each parameter of the equivalent circuit model corresponding to the lithium iron phosphate battery, the basic process noise covariance matrix and the basic measurement noise covariance matrix of the lithium iron phosphate battery. The real-time operating data includes: operating current and key temperature of the battery. The initial estimated values ​​of each parameter include: initial estimated value of ohmic internal resistance, initial estimated value of first-stage polarization internal resistance, initial estimated value of second-stage polarization internal resistance, and initial estimated value of state of charge.

[0056] It should be noted that the equivalent circuit model corresponding to the lithium iron phosphate battery is second-order. Model.

[0057] Specifically, through sensor units within the battery management system (BMS) that are tightly integrated with the lithium iron phosphate battery, at preset system sampling time intervals... (In this embodiment, the interval can be 0.1 seconds, corresponding to a sampling frequency of 10Hz.) The battery terminal voltage is continuously and in real time acquired. Operating current (The current is positive during charging and negative during discharging) and the temperature of one or more key points that can reflect the internal thermal state of the battery. Simultaneously, acquire or pre-store the inherent specifications of the lithium iron phosphate battery, particularly its 1C rated current value. This embodiment refers to an IFR18650 battery, wherein... It is 2A, this This will serve as the benchmark for normalizing the dynamic index of the operating condition in subsequent steps.

[0058] Meanwhile, subsequent parameter identification and optimization are based on an online parameter identification method using an equivalent circuit model (second-order RC model) and an extended Kalman filter (EKF) algorithm. Therefore, the second-order RC equivalent circuit model, which strikes a good balance between model accuracy and computational complexity in this field, was chosen as the basic model to describe the battery's dynamic characteristics. Furthermore, the standard EKF algorithm has been routinely initialized, including setting initial estimates for various parameters of the selected second-order RC model (such as ohmic internal resistance, resistance and capacitance of the two-stage RC network, and initial state of charge (SOC), and setting the basic process noise covariance moment of the lithium iron phosphate battery required by the EKF algorithm. and the covariance matrix of the basic measurement noise This includes the variance of the base voltage measurement noise. .

[0059] In this embodiment, the state vector of the second-order RC model Defined as (representing the states of charge respectively) First RC network voltage Second RC network voltage ), and the current As the system input, the terminal voltage As the output of the system, the basic process noise covariance matrix is: It is a diagonal matrix, and its expression is: , Representing a diagonal matrix, in this embodiment... The setting considers the selected second-order RC model (state vector is The engineering practice of ) is explained in detail below:

[0060]

[0061] The methods for obtaining each component are as follows:

[0062] This value reflects the... Uncertainty arises from the use of the Coulomb integration method for prediction. This uncertainty primarily stems from the cumulative effect of measurement noise from the current sensor and small deviations in the Coulomb efficiency. By analyzing the noise specifications of the current sensor and the Coulomb efficiency characteristics of the battery, the uncertainty is estimated. The variance of errors that may occur within a sampling step.

[0063] It is set to a very small value because it acknowledges that the Coulomb integral is highly accurate in the short term. It is a typical order of magnitude.

[0064] and This value reflects the second order. Uncertainties exist in the model's description of the complex electrochemical polarization and concentration polarization dynamics of the battery. Due to... A network is a simplification of a real electrochemical process, and its parameters (resistance, capacitance) will change with... The temperature and current change nonlinearly, and the model cannot fully capture these changes. In existing technologies, offline testing (…) (Test) Obtain battery response data under different states, and statistically estimate the error variance of polarization voltage prediction by analyzing the difference (residual) between model prediction and actual response.

[0065] and The ratio is usually set. Several orders of magnitude larger, because it acknowledges that modeling the polarization process has greater uncertainty than modeling the charge integral. A typical value representing moderate model uncertainty, which allows the extended Kalman filter algorithm to have enough adjustment space for the estimation of the polarization voltage under dynamic conditions.

[0066] The basic measurement noise covariance matrix Under this input-output definition, it is a scalar, and its expression is:

[0067]

[0068] Wherein, is the terminal voltage The basic measurement noise variance is mainly set according to the accuracy characteristics of the voltage sensor.

[0069] Step 2: Determine the adjustment factor of the basic process noise covariance matrix and the voltage noise adjustment factor of the basic measurement noise covariance matrix according to the working current of the lithium iron phosphate battery to be predicted within the preset time length, and determine the adjusted process noise covariance matrix based on the adjustment factor of the basic process noise covariance matrix, and determine the adjusted basic measurement noise covariance matrix based on the voltage noise adjustment factor of the basic measurement noise covariance matrix;

[0070] Further, the adjustment factor of the basic process noise covariance matrix is determined according to the working current of the lithium iron phosphate battery to be predicted within the preset time length, comprising:

[0071] Determine the current standard deviation and the average value of the current change rate of the lithium iron phosphate battery to be predicted within the preset time length based on the working current of the lithium iron phosphate battery to be predicted within the preset time length.

[0072] Determine the comprehensive dynamic index within the preset time length according to the current standard deviation and the current change rate of the lithium iron phosphate battery to be predicted within the preset time length, and then determine the adjustment factor of the basic process noise covariance matrix according to the comprehensive dynamic index.

[0073] Wherein, the calculation formula of the comprehensive dynamic index is as follows:

[0074]

[0075] In the formula, is the comprehensive dynamic index, is the current standard deviation, is the average value of the current change rate, is the rated current of the lithium iron phosphate battery, which can be 2A in this embodiment, is the sampling time interval of the extended Kalman filter algorithm, which is 0.1s in this embodiment, that is, 10Hz;

[0076] The adjustment factor of the base process noise covariance matrix is calculated as follows:

[0077]

[0078] In the formula, is the adjustment factor of the base process noise covariance matrix, and its value range is , is a sensitivity parameter, which is used to control the steepness of the response curve when transitioning from near (steady state) to near (dynamic state). The greater the value, the more dramatic the transition. In the present embodiment, the value of is set to . This value is selected to enable the adjustment factor to produce an effective and smooth response within the critical change interval of the integrated dynamicity index . Specifically, the steady state dynamicity saturation reference point of the present embodiment is set to . It is expected that will be able to complete its main transition process (from near to near ) when changes within a reasonable range near , for example , i.e., the interval from to is the steady state dynamicity saturation reference point, which defines the value of when the integrated dynamicity index is equal to this value. The value of is , i.e., the adjustment effect reaches half the level. According to the type battery used in the present embodiment, the selected value is

[0079] .

[0080]

[0081] In the formula, is the adjusted process noise covariance matrix, is the base process noise covariance matrix;

[0082] It should be noted that, in order to achieve dynamic adjustment of the process noise covariance matrix to adapt to the transition of the battery operating condition from high dynamicity to steady state, the present step constructs the matrix adjustment factor The factor directly acts on the base process noise covariance matrix which is calculated based on the online acquired battery operation data and two core regulation parameters, and ensures the adjustment factor varying in the interval . The calculation formula is as follows:

[0083]

[0084] Firstly, the core part of the formula calculates a comprehensive dynamicity index. This index normalizes both the current standard deviation and the average current variation amount (denoted by ) in a sampling period relative to the rated current of the battery, and fuses them into a dimensionless comprehensive measure in the form of Euclidean norm. The larger the value, the more the current working condition deviates from the absolute stable state, and the stronger the dynamicity.

[0085] Secondly, the exponential term in the denominator behaves as follows:

[0086] When the comprehensive dynamicity is much smaller than the saturation reference point (that is, the working condition is very stable), is a larger positive number, resulting in a very large term, so the value of the entire (that is, 1 / (1+large number)) tends to .

[0087] When the comprehensive dynamicity is much larger than the saturation reference point (that is, the working condition is very dynamic), is a larger negative number, resulting in a very small term close to , so the value of the entire (that is, 1 / (1+small number)) tends to .

[0088] When the comprehensive dynamicity is exactly equal to the saturation reference point , is , the term is , and the value of the entire is .

[0089] The parameter controls the steepness of the transition curve near the center point . The larger, the steeper the curve, The closer to the transition from to is, the more sensitive the response to deviations is.

[0090] Therefore, the value of smoothly increases from as the operating condition transitions from to

[0091] . This means that: When the operating condition is very smooth will tend to its lower limit (a very small positive number, but not 0). At this point, the adjusted process noise covariance matrix will be significantly reduced, which is consistent with the logic that model uncertainty decreases under a smooth operating condition and that more trust should be placed in the model's predictions.

[0092] When the operating condition is very dynamic the value of will tend to its upper limit . At this point, is close to or equal to the underlying , which allows a higher estimate of model uncertainty to be maintained under dynamic operating conditions.

[0093] The final adjusted matrix:

[0094]

[0095] The calculated matrix adjustment factor is multiplied by the underlying process noise covariance matrix , resulting in the process noise covariance matrix that has been optimized by this step, which will be used in the subsequent iteration of the algorithm.

[0096] In the embodiments of the present disclosure, the method further comprises:

[0097] determining an average working current absolute value of the lithium iron phosphate battery to be predicted within the preset time length based on the working current of the lithium iron phosphate battery to be predicted within the preset time length, and determining whether the underlying measurement noise covariance matrix needs to be adjusted based on the average working current absolute value;

[0098] ​If the basic measurement noise covariance matrix needs to be adjusted, the voltage noise adjustment factor of the basic measurement noise covariance matrix is ​​determined; otherwise, the basic measurement noise covariance matrix is ​​used as the adjusted measurement noise covariance matrix.

[0099] It should be noted that the determination is based on whether the battery is operating at a low current. If so, the measurement noise covariance matrix is ​​dynamically increased according to the actual magnitude of the current. This corresponds to the noise variance term in voltage measurement. Specifically, when the monitored average current amplitude is below a preset low current threshold, adjustment of the voltage measurement noise variance is initiated. The smaller the current amplitude (i.e., the worse the signal-to-noise ratio), the greater the increase in the voltage measurement noise variance. In this way, Under conditions of low current and low signal-to-noise ratio, the algorithm can automatically reduce its reliance on voltage measurement data and depend more on the data obtained through the steps described. Optimized model predictions are protected from noise contamination and help avoid biases introduced by previous operating conditions.

[0100] Furthermore, the voltage noise adjustment factor of the basic measurement noise covariance matrix is ​​calculated as follows:

[0101]

[0102] In the formula, The voltage noise adjustment factor is used as the basis for measuring the noise covariance matrix. The preset current judgment threshold, also known as the low current judgment threshold, is set when the absolute value of the battery's average current... When the current falls below this threshold, adjustment of the voltage measurement noise variance is initiated. This embodiment uses the small current judgment threshold. Set as This value was selected specifically for the lithium iron phosphate battery used in this embodiment (a battery with a nominal capacity of...). New battery status AC internal resistance is approximately of This assessment considers both the type of lithium iron phosphate battery cell and the voltage measurement accuracy of a typical battery management system. The typical noise standard deviation of the voltage measurement system referenced in this embodiment is approximately... To ensure that the voltage response signal generated by the battery can be effectively identified and distinguished from measurement noise under low current excitation, this response signal is typically required to be at least several times the noise standard deviation. Considering the above... The type of battery can stably generate the equivalent internal resistance component of the response under low current excitation. To generate at least voltage response (i.e.) times more If the noise standard deviation is 0, then the required excitation current is approximately 0.5%. Further considering that the battery internal resistance increases due to aging, low temperature and other factors in practical application, and to provide more sufficient excitation for the algorithm to ensure the effectiveness of parameter identification, the is set to , which can effectively define the small current working area where the signal-to-noise ratio of the voltage signal is significantly reduced when the working current is lower than this value, and the measurement noise covariance adjustment needs to be started, so as to guarantee the parameter identification in this area; , which can effectively define the small current working area where the signal-to-noise ratio of the voltage signal is significantly reduced when the working current is lower than this value, and the measurement noise covariance adjustment needs to be started, so as to guarantee the parameter identification in this area; is the absolute value of the average working current, is the adjustment intensity factor of the basic measurement noise covariance matrix, which controls the maximum relative intensity of the voltage measurement noise variance that can be amplified under small current conditions. In this embodiment, is set to , when the signal-to-noise ratio of the voltage measurement is significantly deteriorated due to too small current, the measurement noise covariance can be sufficiently but not excessively compensated to guarantee the performance of the algorithm; is a function, when the condition is met, its value is , otherwise it is .

[0103] The calculation formula of the adjusted basic measurement noise covariance matrix is as follows:

[0104]

[0105] In the formula, is the adjusted basic measurement noise covariance matrix, is the basic measurement noise variance of the terminal voltage after adjustment, , is the basic measurement noise variance of the terminal voltage.

[0106] It should be noted that, in order to realize the dynamic adjustment of the voltage measurement noise item in the measurement noise covariance matrix , this step constructs the embodiment form of the optimization factor , that is, the matrix voltage noise adjustment factor , which directly acts on the basic voltage measurement noise variance . The calculation formula of the matrix voltage noise adjustment factor is as follows:

[0107]

[0108] R matrix voltage noise adjustment factor It dynamically adjusts the reliability of voltage measurement based on whether the battery is operating at low current and the specific magnitude of the current, thereby solving the problem of fixed voltage measurement reliability under low current and low signal-to-noise ratio conditions. Matrix leads to The problem of over-reliance on voltage measurement data contaminated by noise.

[0109] Only when the absolute value of the average current obtained in real time is... Less than the core manually set threshold The value of the indicator function is only at that time. Only then will subsequent adjustments take effect; otherwise, the indicator function value will be... This makes the entire product term... ,at this time The value will be equal to 1, meaning no adjustment will be made, and the voltage measurement noise variance will remain at its baseline value. This design ensures that the adjustment is only activated when there is a genuine need to focus on small current effects.

[0110] When the adjustment is activated This item serves as a gradual adjustment. Its value changes with... From near Change to near And from close Increased linearly to near This means that when the current is just below the threshold... When, the value of this term is smaller, for Its contribution is also relatively small, with only slight adjustment; and when the current is very small, far below the threshold, the value of this term is close to... ,right The contribution is the greatest. This linearly gradual design makes the intensity of the adjustment inversely proportional to the current magnitude (i.e., the degree of signal-to-noise ratio degradation), which conforms to the actual physical situation.

[0111] parameter This determines the maximum strength of the adjustment. The entire product term... The maximum value is .therefore, The maximum value is This means that the voltage measurement noise variance can be amplified up to its base value. Times. In this embodiment, Set as This value is chosen to account for the average current. When the voltage signal approaches zero, the signal-to-noise ratio is extremely low (for the reference in this embodiment). Battery, in The amplitude of the voltage signal generated under current and The measurement noise standard deviation is comparable, and the signal-to-noise ratio is only [missing information]. about 2.5 times (left) or 3.5 times (right), which can improve the equivalent noise standard deviation of voltage measurement by about This degree of noise amplification is sufficient to make the algorithm significantly reduce the trust in such low-quality measurement data, and thus preferentially rely on model prediction, effectively suppressing the interference of measurement noise on the identification results of parameters such as internal resistance. Meanwhile, when the current is slightly lower than but still has a certain amplitude, the value of smoothly transitions from to , with the adjustment amplitude being adapted to the degree of deterioration of the signal-to-noise ratio.

[0112] The final adjusted voltage measurement noise variance in the matrix:

[0113]

[0114] The calculated matrix voltage noise adjustment factor is multiplied by the basic voltage measurement noise variance , and the voltage measurement noise variance optimized by this step is obtained.

[0115] The measurement noise covariance matrix is updated as:

[0116] The final measurement noise covariance matrix is obtained by updating the basic matrix (which is usually a diagonal matrix containing the basic noise variances of all measurements such as voltage and current). The element on the diagonal corresponding to the voltage measurement is replaced by the newly calculated . The noise variance terms of other measurement quantities remain unchanged in this step. This will be used together with in the subsequent algorithm iteration calculation.

[0117] Step 3: Based on the adjusted process noise covariance matrix, the adjusted measurement noise covariance matrix, and using the extended Kalman filter algorithm, the initial estimates of each parameter of the equivalent circuit model corresponding to the lithium iron phosphate battery are optimized to obtain the optimized ohmic internal resistance, first-order polarization resistance, second-order polarization resistance, and state of charge.

[0118] It should be noted that after obtaining the dynamically adjusted process noise covariance matrix and measurement noise covariance matrix , this step applies these two optimized covariance matrices to the extended Kalman filter algorithm to replace the fixed basic covariance matrix used in the standard application of the algorithm.​ and (are the parts related to voltage measurement are updated by the corresponding terms in The algorithm itself is based on the selected second-order equivalent circuit model and its pre-set initial parameter state to perform iterative operations. In each calculation period, the algorithm first performs a prediction step, using the battery model and the state estimation value at the last time to predict the state at the current time; then, in the update step, using the residual between the actual measurement value at the current time and the predicted value, and combining the Kalman gain, the state estimation value is corrected. Through this series of standard iterative process, the algorithm can identify and update each parameter in the equivalent circuit model online, especially the resistive parameters closely related to the internal energy loss and heat generation of the battery. The final output of this step is the battery internal resistance parameter that can more accurately reflect the real state of the battery when switching from high dynamics to small current stable working conditions after the optimization method of the application. The parameters include the optimized ohmic internal resistance and the first-order polarization resistance , the second-order polarization resistance , etc. These optimized internal resistance parameters will be used as key input data for subsequent thermal runaway trend prediction.

[0119] wherein, in a calculation period, the iterative calculation process mainly includes the following steps:

[0120] 1. Prediction step:

[0121] State prediction: using the state optimal estimation value at the last time and the state equation of the second-order model, to predict the state prior estimation value at the current time t.

[0122] Error covariance prediction: using the error covariance matrix at the last time , the Jacobian matrix of the state transition matrix , and the optimized process noise covariance matrix , to predict the error covariance prior estimation at the current time t.

[0123] Here, instead of the fixed . The value of

[0124] 2. Update step:

[0125] Kalman gain calculation: using the predicted error covariance , Jacobian matrix of the observation matrix , optimized measurement noise covariance matrix to calculate the Kalman gain at the current time .

[0126] Here, (where the adjusted voltage noise variance is substituted for the fixed . . The value of is determined by the real-time average current size.

[0127] State update: using the residual between the actual voltage measurement value at the current time and the model predicted voltage value, and combining the Kalman gain , the prior estimate value of the state is corrected to obtain the optimal estimation value of the state at the current time (where the optimized internal resistance and state of charge are included).

[0128] Error covariance update: using the Kalman gain and the predicted error covariance , the error covariance matrix is updated to obtain , which is prepared for the next iteration.

[0129] Step 4: determining the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t and the predicted temperature at time t+1 according to the working current, the key point temperature, the optimized ohmic internal resistance, the first-stage polarization resistance, the second-stage polarization resistance, and the state of charge.

[0130] In the embodiments of the present disclosure, the step 4 specifically includes:

[0131] determining the total heat generation rate at the current time t according to the working current, the key point temperature, the optimized ohmic internal resistance, the first-stage polarization resistance, the second-stage polarization resistance, and the state of charge.

[0132] inputting the total heat generation rate at the current time t into a first-order lumped parameter thermal model, and solving the first-order lumped parameter thermal model to obtain the temperature rise rate at the current time t and the predicted temperature at time t+1.

[0133] It should be noted that the total heat generation rate of the battery at the current time t is calculated first.

[0134] Subsequently, the calculated total heat generation rate is taken as heat input, and a first-order lumped parameter thermal model is used for processing. By discretizing and solving the first-order lumped parameter thermal model through the forward Euler method, the current temperature rise rate of the battery can be estimated in real time, and the temperature change trajectory of the battery in a future short period of time can be predicted.

[0135] wherein the total heat generation rate is superimposed by Joule heat , polarization heat and reaction heat .

[0136]

[0137] wherein the Joule heat is calculated by the formula: .

[0138] The polarization heat is calculated by the formula: .

[0139] The reaction heat is calculated by the formula: .

[0140] wherein is the open-circuit voltage temperature coefficient.

[0141] The mathematical expression of the first-order lumped parameter thermal model is as follows:

[0142]

[0143] wherein is the mass of the battery, is the specific heat capacity of the battery, is the comprehensive convective heat transfer coefficient between the surface of the battery and the environment, is the surface heat transfer area of the battery, is the real-time measured battery environment temperature, is the temperature rise rate of the battery.

[0144] Step 5: based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature at t+1 and the key point temperature, the thermal runaway trend of the lithium iron phosphate battery is predicted.

[0145] In the embodiments of the present disclosure, the step 5 specifically comprises:

[0146] determining whether the temperature rise rate at the current time t is greater than a preset temperature rise rate threshold, and / or whether the predicted temperature at the t+1 time is greater than or equal to a preset thermal runaway critical point threshold, and / or whether the key point temperature is greater than a preset temperature threshold, if yes, determining that the lithium iron phosphate battery has a thermal runaway trend, otherwise, the lithium iron phosphate battery does not have a thermal runaway trend.

[0147] It should be noted that the currently calculated battery temperature, temperature rise rate and predicted future temperature are compared with a set of preset multi-dimensional thermal runaway criteria formulated for the safe operation characteristics of the lithium iron phosphate battery. These criteria generally include but are not limited to: whether the battery temperature exceeds the allowed maximum safe working temperature, whether the temperature rise rate exceeds the set danger threshold, and whether the predicted future temperature will reach the known thermal runaway critical point. Once any or a combination of criteria is triggered, the system determines that the battery has a thermal runaway risk.

[0148] In the embodiments of the present disclosure, when the lithium iron phosphate battery has a thermal runaway trend, a warning is given.

[0149] In summary, the lithium iron phosphate battery thermal runaway trend prediction method proposed in the embodiment significantly improves the accuracy of internal resistance identification under critical working conditions, so that the heat generation rate calculated based on the optimized internal resistance can more truly reflect the internal thermal state of the battery, thereby ensuring the quality of the input data of the entire thermal runaway trend prediction process, and ultimately improving the reliability, sensitivity and timeliness of the prediction results, effectively enhancing the operation safety of the lithium iron phosphate battery system.

[0150] Embodiment two

[0151] Figure 2 The structure diagram of a lithium iron phosphate battery thermal runaway trend prediction system provided according to an embodiment of the present application is shown in Figure 2 As shown in the figure, the system comprises:

[0152] The acquisition module 100 is configured to acquire real-time operation data of a lithium iron phosphate battery to be predicted within a preset time period, initial estimated values of parameters of an equivalent circuit model corresponding to the lithium iron phosphate battery, a basic process noise covariance matrix and a basic measurement noise covariance matrix of the lithium iron phosphate battery, wherein the real-time operation data comprises working current and key point temperature of the battery, and the initial estimated values of the parameters comprise initial estimated value of ohmic resistance, initial estimated value of first-stage polarization resistance, initial estimated value of second-stage polarization resistance, and initial estimated value of state of charge.

[0153] The equivalent circuit model corresponding to the lithium iron phosphate battery is a second-order model.

[0154] The first determination module 200 is configured to determine an adjustment factor of the basic process noise covariance matrix and a voltage noise adjustment factor of the basic measurement noise covariance matrix according to the working current of the lithium iron phosphate battery to be predicted within a preset time length, and determine an adjusted process noise covariance matrix based on the adjustment factor of the basic process noise covariance matrix and an adjusted basic measurement noise covariance matrix based on the voltage noise adjustment factor of the basic measurement noise covariance matrix.

[0155] The optimization module 300 is configured to optimize each parameter initial estimation value of the equivalent circuit model corresponding to the lithium iron phosphate battery based on the adjusted process noise covariance matrix and the adjusted measurement noise covariance matrix and by using an extended Kalman filtering algorithm to obtain an optimized ohmic resistance, a first-stage polarization resistance, a second-stage polarization resistance and a state of charge.

[0156] The second determination module 400 is configured to determine a temperature rise rate of the lithium iron phosphate battery to be predicted at a current time t and a predicted temperature at a time t+1 according to the working current, the key point temperature, the optimized ohmic resistance, the first-stage polarization resistance, the second-stage polarization resistance and the state of charge.

[0157] The prediction module 500 is configured to perform thermal runaway trend prediction on the lithium iron phosphate battery based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature at the time t+1 and the key point temperature.

[0158] In the embodiments of the present disclosure, the first determination module 200 is further configured to:

[0159] The average working current absolute value of the lithium iron phosphate battery to be predicted within the preset time length is determined based on the working current of the lithium iron phosphate battery to be predicted within the preset time length, and whether the basic measurement noise covariance matrix needs to be adjusted is determined based on the average working current absolute value.

[0160] If the basic measurement noise covariance matrix needs to be adjusted, the voltage noise adjustment factor of the basic measurement noise covariance matrix is determined, otherwise, the basic measurement noise covariance matrix is taken as the adjusted measurement noise covariance matrix.

[0161] In the embodiments of the present disclosure, the first determination module 200 is further configured to:

[0162] The current standard deviation and the average value of the current change rate of the lithium iron phosphate battery to be predicted within the preset time length are determined based on the working current of the lithium iron phosphate battery to be predicted within the preset time length.

[0163] The comprehensive dynamic degree index in the preset time length is determined according to the current standard deviation and the current change rate of the lithium iron phosphate battery to be predicted in the preset time length, and then the adjustment factor of the basic process noise covariance matrix is determined according to the comprehensive dynamic degree index.

[0164] The calculation formula of the comprehensive dynamic degree index is as follows:

[0165]

[0166] In the formula, is the comprehensive dynamic degree index, is the current standard deviation, is the average value of the current change rate, is the rated current of the lithium iron phosphate battery, is the sampling time interval of the extended Kalman filtering algorithm;

[0167] The calculation formula of the adjustment factor of the basic process noise covariance matrix is as follows:

[0168]

[0169] In the formula, is the adjustment factor of the basic process noise covariance matrix, is the sensitivity parameter, is the working condition dynamic degree saturation reference point.

[0170] The calculation formula of the voltage noise adjustment factor of the basic measurement noise covariance matrix is as follows:

[0171]

[0172] In the formula, is the voltage noise adjustment factor of the basic measurement noise covariance matrix, is the preset current judgment threshold, is the average working current absolute value, is the adjustment intensity factor of the basic measurement noise covariance matrix, is a function, when the condition is met, the value is , otherwise .

[0173] The calculation formula of the adjusted process noise covariance matrix is as follows:

[0174]

[0175] In the formula, is the adjusted process noise covariance matrix, is the basic process noise covariance matrix.

[0176] The calculation formula of the adjusted base measurement noise covariance matrix is as follows:

[0177]

[0178] In the formula, is the adjusted base measurement noise covariance matrix, is the adjusted base measurement noise variance of the terminal voltage, , is the base measurement noise variance of the terminal voltage.

[0179] In the embodiments of the present disclosure, the second determination module 400 is further configured to:

[0180] determine a total heat generation rate at the current time t according to the working current, the key point temperature, the optimized ohmic resistance, the first-stage polarization resistance, the second-stage polarization resistance, and the state of charge;

[0181] input the total heat generation rate at the current time t into a first-order lumped parameter thermal model, and solve the first-order lumped parameter thermal model to obtain a temperature rise rate at the current time t and a predicted temperature at the time t+1.

[0182] In the embodiments of the present disclosure, the prediction module 500 is further configured to:

[0183] determine whether the temperature rise rate at the current time t is greater than a preset temperature rise rate threshold, and / or whether the predicted temperature at the time t+1 is greater than or equal to a preset thermal runaway critical point threshold, and / or whether the key point temperature is greater than a preset temperature threshold, and if so, determine that the lithium iron phosphate battery has a thermal runaway trend, otherwise, the lithium iron phosphate battery does not have a thermal runaway trend.

[0184] In summary, the lithium iron phosphate battery thermal runaway trend prediction system proposed in the embodiments significantly improves the accuracy of internal resistance identification under critical working conditions, so that the heat generation rate calculated based on the optimized internal resistance can more truly reflect the internal thermal state of the battery, thereby ensuring the input data quality of the entire thermal runaway trend prediction process, and ultimately improving the reliability, sensitivity and timeliness of the prediction results, effectively enhancing the operation safety of the lithium iron phosphate battery system.

[0185] In the description of the application, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the application. The illustrative description of the above terms in the specification does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, different embodiments or examples described in the specification and the features of different embodiments or examples can be combined and combined by those skilled in the art without contradiction, if any.

[0186] Any process or method descriptions or descriptions of the flow diagrams in the flow charts described herein or otherwise described in this specification can be understood as representing the steps of a method or process, including one or more steps for implementing custom logic functions or processes, and the scope of the preferred embodiments of the present application includes additional implementation involving other steps, which can be performed at substantially the same time or in reverse order or in other order, and the inclusion of additional or alternative steps can be understood as a preferred embodiment of the application.

[0187] Although the embodiments of the application have been shown and described above, it should be understood that the above-described embodiments are exemplary and are not to be construed as limiting the application, and those skilled in the art can make changes, modifications, replacements and variations to the above-described embodiments within the scope of the application.

Claims

1. A method for predicting thermal runaway tendency of a lithium iron phosphate battery, characterized in that, The method comprises: acquiring real-time operation data of a lithium iron phosphate battery to be predicted within a preset time length, initial estimated values of each parameter of an equivalent circuit model corresponding to the lithium iron phosphate battery, a basic process noise covariance matrix and a basic measurement noise covariance matrix of the lithium iron phosphate battery, wherein the real-time operation data comprises a working current and a key point temperature of the battery, and the initial estimated values of each parameter comprise an initial estimated value of an ohmic resistance, an initial estimated value of a first-stage polarization resistance, an initial estimated value of a second-stage polarization resistance, and an initial estimated value of a state of charge; determining an adjustment factor of the basic process noise covariance matrix and a voltage noise adjustment factor of the basic measurement noise covariance matrix according to the working current of the lithium iron phosphate battery to be predicted within the preset time length, and determining an adjusted basic process noise covariance matrix based on the adjustment factor of the basic process noise covariance matrix, and determining an adjusted basic measurement noise covariance matrix based on the voltage noise adjustment factor of the basic measurement noise covariance matrix; optimizing the initial estimated values of each parameter of the equivalent circuit model corresponding to the lithium iron phosphate battery based on the adjusted basic process noise covariance matrix and the adjusted basic measurement noise covariance matrix, and using an extended Kalman filtering algorithm to obtain optimized ohmic resistance, first-stage polarization resistance, second-stage polarization resistance, and state of charge; determining a temperature rise rate of the lithium iron phosphate battery to be predicted at a current time t and a predicted temperature at a time t+1 according to the working current, the key point temperature, the optimized ohmic resistance, first-stage polarization resistance, second-stage polarization resistance, and state of charge; performing thermal runaway trend prediction on the lithium iron phosphate battery based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature at the time t+1, and the key point temperature.

2. The method of claim 1, wherein, The method further comprises: determining an average absolute value of the working current of the lithium iron phosphate battery to be predicted within the preset time length based on the working current of the lithium iron phosphate battery to be predicted within the preset time length, and judging whether the basic measurement noise covariance matrix needs to be adjusted based on the average absolute value of the working current; if the basic measurement noise covariance matrix needs to be adjusted, determining a voltage noise adjustment factor of the basic measurement noise covariance matrix, otherwise, taking the basic measurement noise covariance matrix as an adjusted basic measurement noise covariance matrix.

3. The method of claim 2, wherein, The equivalent circuit model corresponding to the lithium iron phosphate battery is a second-order model. model.

4. The method of claim 3, wherein, The determination of the adjustment factor of the basic process noise covariance matrix according to the working current of the lithium iron phosphate battery to be predicted within the preset time length comprises: determining a current standard deviation and an average value of a current change rate of the lithium iron phosphate battery to be predicted within the preset time length based on the working current of the lithium iron phosphate battery to be predicted within the preset time length; determining a comprehensive dynamic index within the preset time length according to the current standard deviation and the current change rate of the lithium iron phosphate battery to be predicted within the preset time length, and then determining the adjustment factor of the basic process noise covariance matrix according to the comprehensive dynamic index.

5. The method of claim 4, wherein, The calculation formula of the comprehensive dynamic index is as follows: In the formula, is a comprehensive dynamic index, is a current standard deviation, is an average value of the current change rate, is a rated current of the lithium iron phosphate battery, is a sampling time interval of the extended Kalman filtering algorithm; The calculation formula of the adjustment factor of the basic process noise covariance matrix is as follows: wherein is an adjustment factor for the base process noise covariance matrix, is a sensitivity parameter, is a working condition dynamic degree saturation reference point.

6. The method of claim 5, wherein, A calculation formula of a voltage noise adjustment factor of the basic measurement noise covariance matrix is as follows: wherein is a voltage noise adjustment factor for the base measurement noise covariance matrix, is a preset current decision threshold, is an average operating current absolute value, is an adjustment strength factor for the base measurement noise covariance matrix, is a function that has a value of when the condition is met, and otherwise.

7. The method of claim 6, wherein, A calculation formula of the adjusted basic process noise covariance matrix is as follows: wherein is the adjusted base process noise covariance matrix, is the base process noise covariance matrix; A calculation formula of the adjusted basic measurement noise covariance matrix is as follows: wherein is the adjusted base measurement noise covariance matrix, is the adjusted base measurement noise variance of the terminal voltage, , is the base measurement noise variance of the terminal voltage.

8. The method of claim 7, wherein, The determination of the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t and the predicted temperature at the time t+1 according to the working current, the key point temperature, the optimized ohmic internal resistance, the first-stage polarization internal resistance, the second-stage polarization internal resistance and the state of charge comprises: The determination of the total heat generation rate at the current time t according to the working current, the key point temperature, the optimized ohmic internal resistance, the first-stage polarization internal resistance, the second-stage polarization internal resistance and the state of charge; The total heat generation rate at the current time t is input into a first-order lumped parameter thermal model, and the first-order lumped parameter thermal model is solved to obtain the temperature rise rate at the current time t and the predicted temperature at the time t+1.

9. The method of claim 1, wherein, The thermal runaway trend prediction of the lithium iron phosphate battery based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature at the time t+1 and the key point temperature comprises: It is judged whether the temperature rise rate at the current time t is greater than a preset temperature rise rate threshold, and / or whether the predicted temperature at the time t+1 is greater than or equal to a preset thermal runaway critical point threshold, and / or whether the key point temperature is greater than a preset temperature threshold, if yes, it is determined that the lithium iron phosphate battery has a thermal runaway trend, otherwise, the lithium iron phosphate battery does not have a thermal runaway trend.

10. A lithium iron phosphate battery thermal runaway trend prediction system, characterized in that, The system comprises: An acquisition module is configured to acquire real-time running data of a lithium iron phosphate battery to be predicted, initial estimated values of parameters of an equivalent circuit model corresponding to the lithium iron phosphate battery, a basic process noise covariance matrix and a basic measurement noise covariance matrix of the lithium iron phosphate battery within a preset time length, wherein the real-time running data comprises a working current and a key point temperature of the battery, and the initial estimated values of the parameters comprise an ohmic internal resistance initial estimated value, a first-stage polarization internal resistance initial estimated value, a second-stage polarization internal resistance initial estimated value and a state of charge initial estimated value; A first determination module is configured to determine an adjustment factor of the basic process noise covariance matrix and a voltage noise adjustment factor of the basic measurement noise covariance matrix according to the working current of the lithium iron phosphate battery to be predicted within the preset time length, and determine an adjusted basic process noise covariance matrix based on the adjustment factor of the basic process noise covariance matrix and determine an adjusted basic measurement noise covariance matrix based on the voltage noise adjustment factor of the basic measurement noise covariance matrix; An optimization module is configured to optimize the initial estimated values of the parameters of the equivalent circuit model corresponding to the lithium iron phosphate battery based on the adjusted basic process noise covariance matrix and the adjusted basic measurement noise covariance matrix, and obtain an optimized ohmic internal resistance, a first-stage polarization internal resistance and a second-stage polarization internal resistance by using an extended Kalman filtering algorithm. A second determination module is configured to determine a temperature rise rate of the lithium iron phosphate battery to be predicted at a current time t and a predicted temperature at a time t+1 according to the working current, the key point temperature, the optimized ohmic resistance, the first-stage polarization resistance, the second-stage polarization resistance, and the state of charge; A prediction module is configured to predict a thermal runaway trend of the lithium iron phosphate battery based on the temperature rise rate of the lithium iron phosphate battery to be predicted at the current time t, the predicted temperature at the time t+1, and the key point temperature.

Images