A lithium-ion battery fault identification method and system

Abstract

The present invention discloses a method and system for identifying lithium-ion battery faults. The method includes subjecting the lithium-ion battery to a preset test experiment to obtain test data. Based on the test data, the lithium-ion battery's equivalent circuit structure is subjected to model parameter fitting and state-of-charge estimation to establish an equivalent circuit digital model. Dual-source data of the lithium-ion battery is generated based on the equivalent circuit digital model and the lithium-ion battery's physical system. The deviation similarity of the dual-source data over a preset operating time of the lithium-ion battery is statistically analyzed to determine the current operating state of the lithium-ion battery. If the lithium-ion battery is abnormal, the parameter change of the equivalent circuit digital model is minimized based on the optimized variables under preset dual-source deviation conditions. Based on the parameter change, the fault type of the lithium-ion battery is determined to obtain a fault identification result. This embodiment effectively identifies the fault type of the lithium-ion battery and improves the accuracy and reliability of abnormality determination and fault type identification of the lithium-ion battery.

Description

Technical Field

[0001] The present invention relates to the field of lithium-ion battery fault identification, and in particular to a lithium-ion battery fault identification method and system. Background Art

[0002] Lithium-ion batteries, as a flexible and convenient energy storage resource, have been applied in numerous fields, including power grids, electric vehicles, and aerospace. However, safety incidents involving lithium-ion batteries are frequent, and their application reliability is a key constraint to their development. Research on lithium-ion battery operating status identification and fault location technologies has become a hot topic in the lithium-ion battery field.

[0003] Currently, methods for detecting abnormal conditions and locating faults in lithium-ion batteries can be roughly divided into two categories: mechanism-based methods and data-based methods. The mechanism-based method analyzes the electrochemical reaction process of lithium-ion batteries to analyze their external characteristics under different operating conditions and faults, establishes the relationship between internal mechanisms and external characteristics, and forms a mapping of internal abnormalities in lithium-ion batteries by analyzing changes in external characteristics, thereby achieving abnormality calibration and fault identification. This type of method is generally complex, and the complex and highly coupled electrochemical reactions within the battery make it difficult to form a concise and effective representation of variable reactions, making it difficult to apply in practice. The data-based method, on the other hand, uses large amounts of historical data from lithium-ion batteries, applies big data and artificial intelligence technologies, mines lithium-ion battery data features, analyzes the causes of faults, and forms a judgment on abnormal lithium-ion operating conditions and fault classification. Although this type of method can ignore the complex internal mechanism of lithium-ion batteries and greatly simplify the abnormality calibration method of lithium-ion batteries, its capability is limited by data quality and is greatly affected by the data. Incomplete and incomplete data often leads to erroneous results, causing adverse effects. Summary of the Invention

[0004] The present invention provides a lithium-ion battery fault identification method and system, which can effectively identify the fault type of the lithium-ion battery and improve the accuracy and reliability of abnormality determination and fault type identification of the lithium-ion battery.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides a lithium-ion battery fault identification method, comprising:

[0006] The lithium-ion battery is subjected to a preset test experiment to obtain test data, and based on the test data, the equivalent circuit structure of the lithium-ion battery is subjected to model parameter fitting and charge state estimation to establish an equivalent circuit digital model;

[0007] Generate dual-source data of the lithium-ion battery based on the equivalent circuit digital model and the lithium-ion battery physical system; wherein the dual-source data includes simulation data and physical measurement data;

[0008] Statistically analyzing the deviation similarity of dual-source data of the lithium-ion battery during the preset operation time, and judging the current operating status of the lithium-ion battery based on the deviation similarity;

[0009] If the current operating state of the lithium-ion battery is abnormal, under the preset dual-source deviation condition, the parameter change of the equivalent circuit digital model is minimized according to the optimization variable, and the fault type of the lithium-ion battery is determined based on the parameter change to obtain the fault identification result.

[0010] In an embodiment of the present invention, a lithium-ion battery is subjected to a preset test experiment to obtain test data. Based on the test data, the equivalent circuit structure of the lithium-ion battery is subjected to model parameter fitting and state of charge estimation to establish an equivalent circuit digital model. Dual-source data of the lithium-ion battery is generated based on the equivalent circuit digital model and the physical system of the lithium-ion battery. The dual-source data includes simulation data and physical measurement data. The deviation similarity of the dual-source data of the lithium-ion battery for a preset operation time is statistically analyzed, and the current operating state of the lithium-ion battery is determined based on the deviation similarity. If the current operating state of the lithium-ion battery is abnormal, the parameter change of the equivalent circuit digital model is minimized based on the optimization variables under the preset dual-source deviation condition, and the fault type of the lithium-ion battery is determined based on the parameter change to obtain a fault identification result. A dual-source data structure for lithium-ion batteries is constructed using simulation data from a digital model of a lithium-ion battery equivalent circuit and physical test data from the battery. By analyzing the deviation in the dual-source data, abnormal operating conditions of the lithium-ion battery can be determined. After determining abnormal lithium-ion battery operation, the simulation model parameters are adjusted by minimizing the error between the simulation model data and the measured data. The model parameter changes are analyzed to determine the lithium-ion battery fault type and accurately locate the fault. Compared with existing data-based lithium-ion battery early warning methods, the present invention significantly reduces the requirements for lithium-ion battery data quality, thereby achieving reliable fault location. Compared with model-based lithium-ion operational risk methods, the present invention reduces the reliance on model accuracy, ensures the accuracy of lithium-ion battery abnormality determination and fault location, and effectively identifies the lithium-ion battery fault type. Furthermore, compared with existing digital-analog-driven lithium-ion battery digital mirroring or digital twin technologies, the present invention can effectively simulate and accurately identify fault conditions by analyzing model parameter changes, helping to improve the operational safety and reliability of lithium-ion battery systems, increase the accuracy and reliability of lithium-ion battery abnormality determination and fault type identification, and reduce the workload of operation and maintenance.

[0011] As a preferred solution, based on the test data, the equivalent circuit structure of the lithium-ion battery is subjected to model parameter fitting and state of charge estimation to establish an equivalent circuit digital model, specifically:

[0012] Select a preset order of undetermined parameter equivalent circuit digital model based on the equivalent circuit structure and fault type of the lithium-ion battery; the fault types include internal short circuit, abnormal growth of the solid electrolyte interface (SEI) film, and lithium dendrites;

[0013] Based on the digital model of the equivalent circuit with undetermined parameters and test data, the undetermined coefficients of the function are adjusted by a parameter optimization method to obtain a functional expression of the model parameters of the digital model of the equivalent circuit with undetermined parameters and the influencing factors; wherein the undetermined coefficients of the function are determined based on the functional relationship between the model parameters of the digital model of the equivalent circuit with undetermined parameters and the influencing factors; the parameter optimization methods include the least squares method, the particle swarm method, and the simulated annealing method;

[0014] The digital model of the equivalent circuit with undetermined parameters and the function expression are used to estimate the state of charge of the lithium ion to obtain a state of charge value of the first lithium ion battery. The digital model of the equivalent circuit is established based on the first state of charge value of the lithium ion battery and the digital model of the equivalent circuit with undetermined parameters.

[0015] As a preferred solution, under the preset dual-source deviation condition, the parameter change of the equivalent circuit digital model is minimized according to the optimization variables, specifically:

[0016] The model parameters and influencing factors of the equivalent circuit digital model are used as optimization variables; the influencing factors include state of charge, temperature, and current charge and discharge rate;

[0017] Construct the optimization objective function and constraint conditions based on the optimization variables and the preset dual-source deviation conditions;

[0018] The optimization variables, objective functions and constraints are subjected to particle swarm optimization calculation to obtain the parameter values ​​of the optimization variables under abnormal conditions;

[0019] The parameter variation is calculated based on the parameter values ​​of the optimized variables and the parameter values ​​of the equivalent circuit digital model under the abnormal state.

[0020] As a preferred solution, the objective function is optimized as follows:

[0021]

[0022] Where J is the optimization objective function; X(t)=[x i (t)|i=1,2,…,n] represents the model parameters of the equivalent circuit digital model of the lithium-ion battery before the lithium-ion battery is judged to be in an abnormal state at time t; X + (t) = [x i + (t)|i=1,2,…,n] represents the parameter variables to be optimized of the equivalent circuit digital model of the lithium-ion battery after the lithium-ion battery is determined to be in an abnormal state at time t; n represents the number of optimization variables;

[0023] Constraints, specifically:

[0024] stα + ≤X + (t)≤β +

[0025]

[0026] Among them, α + and β + are the upper and lower bounds of the optimization variables; f(X + (t)) represents the output voltage function, U M (t) represents the output voltage measurement value of the lithium-ion battery physical system at time t; To preset the dual-source deviation condition, and σ(t) represent the mean and standard deviation of the deviation values ​​of the dual-source data of the lithium-ion battery in the L period before time t, respectively.

[0027] As a preferred solution, dual-source data of lithium-ion batteries is generated based on the equivalent circuit digital model and the physical system of lithium-ion batteries, specifically:

[0028] Collect the output current and output voltage of the lithium-ion battery physical system during actual operation, and calibrate the output voltage as physical measurement data;

[0029] The output current is input into the equivalent circuit digital model, and simulation data of the output voltage of the equivalent circuit digital model is obtained through simulation calculation.

[0030] As a preferred solution, the deviation similarity of the dual-source data of the lithium-ion battery running for a preset time is statistically analyzed, specifically:

[0031] The lithium-ion battery is operated for a preset time, and all dual-source data of the lithium-ion battery under the preset time are counted;

[0032] According to the factors affecting the measurement error, a time window is determined, and based on all the dual-source data of the lithium-ion battery under the preset time, the deviation value and deviation similarity of the dual-source data of the lithium-ion battery within the time window are calculated; wherein the deviation similarity includes the deviation mean and standard deviation.

[0033] As a preferred solution, the current operating state of the lithium-ion battery is determined based on the deviation similarity, specifically:

[0034] According to the deviation mean and standard deviation, the normal range is obtained;

[0035] Calculate the current deviation value of the lithium-ion battery according to the dual-source data of the lithium-ion battery in the current operating state;

[0036] If the current deviation value of the lithium-ion battery is within the normal range, it is determined that the current operating state of the lithium-ion battery is normal;

[0037] If the current deviation value of the lithium-ion battery is not within the normal range, it is determined that the current operating state of the lithium-ion battery is an abnormal state.

[0038] As a preferred solution, the lithium-ion battery is subjected to a preset test experiment to obtain test data, specifically:

[0039] Conduct hybrid power pulse characteristic tests and open circuit voltage tests on lithium-ion batteries under different test environments and conditions to obtain test data;

[0040] The test data includes the output voltage and current data of lithium-ion batteries under different temperatures, different charge and discharge rates, and different states of charge.

[0041] As a preferred solution, after determining the fault type of the lithium-ion battery according to the parameter change and obtaining the fault identification result, the method further includes:

[0042] According to the identification results, the lithium-ion battery is maintained and repaired;

[0043] Re-test the lithium-ion battery after maintenance and overhaul to obtain the latest test data. Based on the latest test data, perform model parameter fitting and charge state estimation on the equivalent circuit structure of the lithium-ion battery to establish the latest equivalent circuit digital model;

[0044] The latest equivalent circuit digital model is used for fault identification of lithium-ion batteries after maintenance and overhaul.

[0045] In order to solve the same technical problem, an embodiment of the present invention further provides a lithium-ion battery fault identification system, comprising: a model building module, a dual-source data module, an abnormality judgment module and a fault identification module;

[0046] The model building module is used to perform a preset test experiment on the lithium-ion battery to obtain test data, and based on the test data, perform model parameter fitting and charge state estimation on the equivalent circuit structure of the lithium-ion battery to establish an equivalent circuit digital model;

[0047] The dual-source data module is used to generate dual-source data of lithium-ion batteries based on an equivalent circuit digital model and a physical system of the lithium-ion battery; wherein the dual-source data includes simulation data and physical measurement data;

[0048] The abnormality judgment module is used to statistically analyze the deviation similarity of the dual-source data of the lithium-ion battery during the preset operation time, and judge the current operating status of the lithium-ion battery based on the deviation similarity;

[0049] The fault identification module is used to minimize the parameter change of the equivalent circuit digital model according to the optimization variables under the preset dual-source deviation conditions when the current operating state of the lithium-ion battery is abnormal, and to determine the fault type of the lithium-ion battery based on the parameter change to obtain the fault identification result. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 : A schematic flow chart of an embodiment of a lithium-ion battery fault identification method provided by the present invention;

[0051] Figure 2 : A schematic structural diagram of a fault identification method for a lithium-ion battery according to an embodiment of the present invention;

[0052] Figure 3 : An equivalent circuit diagram of an embodiment of a lithium-ion battery fault identification method provided by the present invention;

[0053] Figure 4 : A schematic diagram of lithium-ion battery abnormality determination according to an embodiment of a lithium-ion battery fault identification method provided by the present invention;

[0054] Figure 5 : A flow chart of calculating model parameters using a particle swarm optimization method according to an embodiment of a lithium-ion battery fault identification method provided by the present invention;

[0055] Figure 6 : A structural schematic diagram of an embodiment of a lithium-ion battery fault identification system provided by the present invention;

[0056] Figure 7 : A schematic diagram of the execution flow of an embodiment of a lithium-ion battery fault identification system provided by the present invention. DETAILED DESCRIPTION

[0057] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0058] Example 1

[0059] Please refer to Figure 1 , is a flow chart of a fault identification method for a lithium-ion battery provided by an embodiment of the present invention, wherein the structural diagram of the fault identification method is as follows: Figure 2As shown. The fault identification method of this embodiment is applicable to lithium-ion batteries. This embodiment uses the dual-source data structure of lithium-ion batteries to analyze the deviation of the dual-source data to determine the abnormal operating conditions of the lithium-ion batteries. By minimizing the parameter changes of the equivalent circuit digital model, the fault type of the lithium-ion battery is effectively identified, thereby improving the accuracy and reliability of the abnormal determination and fault type identification of the lithium-ion battery. The fault identification method includes steps 101 to 104, each of which is specifically as follows:

[0060] Step 101: Perform a preset test experiment on the lithium-ion battery to obtain test data, and perform model parameter fitting and state of charge estimation on the equivalent circuit structure of the lithium-ion battery based on the test data to establish an equivalent circuit digital model.

[0061] In this embodiment, a digital model of the lithium-ion battery equivalent circuit is established, and hybrid power pulse characteristic tests and open-circuit voltage tests are performed on the lithium-ion battery under different environments and test conditions. An equivalent circuit structure of the lithium-ion battery is assumed, and model parameters are adjusted using test data and parameter fitting methods. A state of charge estimation method based on the lithium-ion battery equivalent circuit model is designed.

[0062] Optionally, the lithium-ion battery is subjected to a preset test experiment to obtain test data, specifically: under different test environments and test conditions, the lithium-ion battery is subjected to a mixed power pulse characteristic test and an open circuit voltage test to obtain test data; wherein the test data is the output voltage data and output current data of the lithium-ion battery under different temperatures, different charge and discharge rates and different states of charge.

[0063] In this embodiment, a hybrid power pulse characteristic test and an open circuit voltage test are performed on a lithium-ion battery under different test environments and test conditions to obtain output voltage data and output current data of the lithium-ion battery under different temperatures, charge and discharge rates, and different states of charge, i.e., test data.

[0064] Optionally, based on the test data, the equivalent circuit structure of the lithium-ion battery is subjected to model parameter fitting and state of charge estimation to establish an equivalent circuit digital model, specifically including steps 11 to 13, each of which is as follows:

[0065] Step 11: Selecting a preset order of undetermined parameter equivalent circuit digital model based on the equivalent circuit structure and fault type of the lithium-ion battery; wherein the fault types include internal short circuit, abnormal growth of solid electrolyte interface SE I film, and lithium dendrite;

[0066] In this embodiment, in combination with the ability of the lithium-ion battery circuit equivalent circuit digital model to reflect common lithium-ion battery faults, and based on the fault identification requirements of specific applications, such as internal short circuits, abnormal growth of solid electrolyte interface SE I films, lithium dendrites, and other fault types (abnormal problems), lithium-ion battery equivalent circuit digital models of different orders are selected to obtain the output voltage of the equivalent circuit digital model. That is, based on the equivalent circuit structure and fault type of the lithium-ion battery, an equivalent circuit digital model with undetermined parameters of a preset order is selected. For example, the Ri nt lithium-ion battery equivalent circuit only represents a voltage source and a series internal resistance, so it can only reflect faults with internal resistance changes, such as short circuits and aging. The Thevenin second-order equivalent circuit, in addition to having the internal resistance of the Ri nt model, can also simulate the electrochemical polarization and concentration polarization processes of the lithium-ion battery, and has the ability to reflect faults such as lithium dendrites in the lithium-ion battery.

[0067] As an example of this embodiment, a second-order Thevenin equivalent circuit digital model is used as an example for explanation. The equivalent circuit structure is as follows: Figure 3 As shown, the output voltage of the second-order Thevenin equivalent circuit digital model can be obtained as:

[0068]

[0069] Among them, U ocv (t) is the open circuit voltage of the lithium ion battery at time t, τ1(t) and τ2(t) represent the delay time constant of the electrochemical polarization process and concentration polarization process of the lithium ion battery at time t, respectively. b (t)×C b (t) and R th (t)×C th (t), R0(t) is the equivalent internal resistance of the lithium-ion battery at time t, R b (t) and C b (t) are the electrochemical polarization internal resistance and capacitance of the lithium-ion battery at time t, R th (t) and C th (t) is the concentration polarization internal resistance / capacitance of the lithium-ion battery at time t, I0(t) is the output current of the lithium-ion battery at time t, U L (t) is the output voltage of the lithium-ion battery at time t.

[0070] Step 12: Based on the digital model of the equivalent circuit with undetermined parameters and the test data, the undetermined coefficients of the function are adjusted by a parameter optimization method to obtain a functional expression of the model parameters of the digital model of the equivalent circuit with undetermined parameters and the influencing factors; wherein the undetermined coefficients of the function are determined based on the functional relationship between the model parameters of the digital model of the equivalent circuit with undetermined parameters and the influencing factors; the parameter optimization methods include the least squares method, the particle swarm optimization method, and the simulated annealing method;

[0071] In this embodiment, the functional relationship between the parameters of the lithium-ion battery equivalent circuit digital model and its influencing factors is the functional relationship between the parameters of the equivalent circuit digital model and the influencing factors such as the state of charge, temperature, and charge and discharge rate, such as a quadratic function, a cubic function, etc., to determine the coefficients to be adjusted in the function of the equivalent circuit model parameters and the influencing factors; wherein the model parameter U ocv (t), R0(t), R b (t), C b (t), R th (t) and C th (t) are the undetermined parameters of the digital model of the lithium-ion battery equivalent circuit at time t. These time-varying parameters are related to the battery state of charge S of the lithium ion at time t. soc (t), temperature T(t) and current charge and discharge rate C r (t) and other factors, namely:

[0072]

[0073] Based on the test data of the lithium-ion battery obtained in step 101 and the lithium-ion battery equivalent circuit model set in step 12, the function coefficients are adjusted using the least squares method or other optimization methods, such as particle swarm optimization, simulated annealing, and other optimization methods to obtain a functional expression of the lithium-ion battery equivalent circuit model parameters and their influencing factors, that is, a functional expression of the model parameters of the digital model of the equivalent circuit with undetermined parameters and the influencing factors.

[0074] Step 13: Using the undetermined parameter equivalent circuit digital model and the function expression to estimate the lithium ion state of charge, obtain a first lithium ion battery state of charge value, and establish an equivalent circuit digital model based on the first lithium ion battery state of charge value and the undetermined parameter equivalent circuit digital model.

[0075] In this embodiment, based on the undetermined digital model parameters, model parameters, and function expressions of the lithium-ion battery equivalent circuit in steps 11 and 12, a lithium-ion state of charge (SOC) value is calculated using a lithium-ion state of charge (SOC) estimation method, thereby completing the establishment of a digital model of the lithium-ion battery equivalent circuit. The lithium-ion SOC estimation method may be based on a Kalman state optimal estimation theory or an ampere-hour integration method.

[0076] It should be noted that, as shown in step 12, the parameters of the equivalent circuit digital model are affected by factors such as the state of charge, temperature, and charge and discharge rate. In step 13, the state of charge value is calculated and applied to the model parameter calculation of the equivalent circuit digital model with undetermined parameters, thereby completing the establishment of the lithium-ion battery equivalent circuit digital model.

[0077] Step 102: Generate dual-source data of the lithium-ion battery based on the equivalent circuit digital model and the physical system of the lithium-ion battery; wherein the dual-source data includes simulation data and physical measurement data.

[0078] In this embodiment, the dual-source data of the lithium-ion battery is generated by inputting the output current of the lithium-ion battery physical system into the digital model of the lithium-ion battery equivalent circuit, calculating the simulation data of the lithium-ion battery equivalent circuit model and the physical measurement data of the physical system (such as the output voltage of the lithium-ion battery), and generating the dual-source data of the lithium-ion battery simulation and actual measurement.

[0079] Optionally, step 102 specifically includes step 1021 to step 1022, and each step is specifically as follows:

[0080] Step 1021: Collect the output current and output voltage of the lithium-ion battery physical system during actual operation, and calibrate the output voltage as physical measurement data.

[0081] In this embodiment, the output current and output voltage of the lithium-ion battery physical system during actual operation are collected, and the collected output voltage is calibrated as the experimental source data of the lithium-ion battery, that is, the physical measurement data, which is recorded as U at time t. M (t).

[0082] Step 1022: Input the output current into the equivalent circuit digital model, and obtain simulation data of the output voltage of the equivalent circuit digital model through simulation calculation.

[0083] In this embodiment, the output current of the lithium-ion battery physical system collected is loaded into the lithium-ion battery equivalent circuit digital model obtained in step 101, and the simulation data of the output voltage of the lithium-ion battery equivalent circuit digital model is obtained through simulation calculation, which is calibrated as the simulation source data of the lithium-ion battery and recorded as U at time t. L (t);

[0084] Step 103: Statistically analyzing the deviation similarity of the dual-source data of the lithium-ion battery during the preset operation time, and judging the current operation state of the lithium-ion battery based on the deviation similarity.

[0085] In this embodiment, the abnormality of the lithium-ion battery is determined by analyzing the similarity of the deviation of the dual-source data when the lithium-ion battery is running for a long time to determine the operating condition of the lithium-ion battery.

[0086] Optionally, step 103 specifically includes step 1031 to step 1032, and each step is specifically as follows:

[0087] Step 1031: Operate the lithium-ion battery for a preset time, and collect all dual-source data of the lithium-ion battery during the preset time; determine a time window based on factors affecting measurement error, and calculate the deviation value and deviation similarity of the dual-source data of the lithium-ion battery within the time window based on all dual-source data of the lithium-ion battery during the preset time; wherein the deviation similarity includes the deviation mean and standard deviation.

[0088] Step 1032: Obtain a normal range based on the deviation mean and standard deviation; calculate a current deviation value of the lithium-ion battery based on the dual-source data of the lithium-ion battery in the current operating state; if the current deviation value of the lithium-ion battery is within the normal range, determine that the current operating state of the lithium-ion battery is normal; if the current deviation value of the lithium-ion battery is not within the normal range, determine that the current operating state of the lithium-ion battery is abnormal.

[0089] In this embodiment, the dual-source data deviation value ΔU(t) of the lithium-ion battery voltage is calculated as |U M (t)-U L (t)| characterizes the difference between the lithium-ion battery physical system and the equivalent circuit digital model. Since the lithium-ion battery equivalent circuit digital model is based on the lithium-ion battery in its normal state, it represents the output voltage of the lithium-ion battery in this normal state, while the state of the lithium-ion battery physical system is unknown. If the physical system is normal, the dual-source data deviation value ΔU(t) remains essentially unchanged and small. If the physical system is abnormal, the dual-source data deviation value ΔU(t) will increase. Therefore, analyzing the changes in the dual-source data deviation value ΔU(t) of the lithium-ion battery can detect the operating status of the lithium-ion battery.

[0090] When judging the operating status of a lithium-ion battery, it is necessary to consider the modeling error of the lithium-ion battery digital model, the lithium-ion battery aging factor, and the influence of the measurement error of signals such as current and voltage. Based on the measurement error influencing factors (such as modeling error, lithium-ion battery aging factor, current measurement error, and voltage measurement error), the time window L is determined. When determining the abnormal state of the lithium-ion battery at time t, the average value (deviation average) of the deviation value ΔU(t) of the lithium-ion battery dual-source data from tL to t is calculated. and standard deviation According to the deviation mean and standard deviation, the normal range is obtained. If the deviation value of the lithium-ion battery dual-source data at time t is within the normal range, such as: If the lithium-ion battery is in a normal operating state, the battery is judged to be in an abnormal operating state. If the lithium-ion battery is in an abnormal operating state, the fault type of the lithium-ion battery is located. The schematic diagram of the abnormal judgment of the lithium-ion battery is as follows: Figure 4 shown.

[0091] Step 104: If the current operating state of the lithium-ion battery is abnormal, under the preset dual-source deviation condition, the parameter change of the equivalent circuit digital model is minimized according to the optimization variable, and the fault type of the lithium-ion battery is determined based on the parameter change to obtain a fault identification result.

[0092] In this embodiment, a method for locating and identifying the type of lithium-ion battery fault in an abnormal state is established. When the lithium-ion battery is determined to be in an abnormal state, the parameters of the lithium-ion battery equivalent circuit digital model are optimized. Under the condition of a conventional deviation level of the lithium-ion battery dual-source data (preset dual-source deviation condition), the change in the parameters of the equivalent circuit digital model is minimized. The lithium-ion battery fault type is determined based on the change in the parameters of the lithium-ion battery equivalent circuit digital model (the parameter change), and the lithium-ion battery fault is located and the lithium-ion battery fault type is determined.

[0093] Optionally, step 104 specifically includes steps 1041 to 1045, each of which is as follows:

[0094] Step 1041: using the model parameters and influencing factors of the equivalent circuit digital model as optimization variables; wherein the influencing factors include state of charge, temperature, and current charge and discharge rate.

[0095] In this embodiment, the lithium-ion battery equivalent model parameters and influencing factors are used as optimization variables. As an example of this embodiment, the second-order Thevenin equivalent circuit digital model is taken as an example, where the model parameter U + ocv (t), R + 0(t), R + b (t), C + b (t), R + th (t), C + th (t) and the state of charge S of the lithium-ion battery + soc (t), temperature T + (t) and current charge and discharge rate C + r (t) and other factors are optimization variables.

[0096] Step 1042: Constructing an optimization objective function and constraint conditions based on the optimization variables and the preset dual-source deviation conditions;

[0097] In this embodiment, minimizing the change in the parameters of the equivalent circuit digital model is used as the optimization objective function, optimization variables, and dual-source data deviation range constraint conditions.

[0098] Optionally, optimize the objective function, specifically:

[0099]

[0100] Where J is the optimization objective function; X(t)=[x i (t)|i=1,2,…,n] represents the model parameters of the equivalent circuit digital model of the lithium-ion battery before the lithium-ion battery is judged to be in an abnormal state at time t; X + (t) = [x i + (t)|i=1,2,…,n] represents the parameter variables to be optimized of the equivalent circuit digital model of the lithium-ion battery after the lithium-ion battery is determined to be in an abnormal state at time t; n represents the number of optimization variables;

[0101] Optional constraints, specifically:

[0102] stα + ≤X + (t)≤β +

[0103]

[0104] Among them, α + and β + are the upper and lower bounds of the optimization variables; f(X + (t)) represents the output voltage function, U M (t) represents the output voltage measurement value of the lithium-ion battery physical system at time t; To preset the dual-source deviation condition, and σ(t) represent the mean and standard deviation of the deviation values ​​of the dual-source data of the lithium-ion battery in the L period before time t, respectively.

[0105] As an example of this embodiment, taking the second-order Thevenin equivalent circuit digital model as an example, the optimization objective function J and the constraints are as follows:

[0106]

[0107] stα + ≤X + (t)≤β +

[0108]

[0109] Among them, ocv (t),R0(t),Rb (t),C b (t),R th (t),C th (t),S soc (t),T(t),C r (t)] represents the digital model parameters of the lithium-ion battery equivalent circuit before the lithium-ion battery is judged to be in an abnormal state at time t; X + (t) = [x + 1(t),x + 2(t),x + 3(t),x + 4(t),x + 5(t),x + 6(t),x + 7(t),x + 8(t),x + 9(t)]=[U + ocv (t),R + 0(t),R + b (t),C + b (t),R + th (t),C + th (t),S + soc (t),T + (t),C + r (t)] represents the parameter variables to be optimized in the digital model of the lithium-ion battery equivalent circuit after the lithium-ion battery is judged to be in an abnormal state at time t; α + and β + Respectively represent the upper and lower bound constraints of the optimization variables; f(X + (t)) represents the output voltage function of the digital model of the lithium-ion battery equivalent circuit.

[0110] Step 1043: performing particle swarm optimization calculation on the optimization variables, the objective function, and the constraints to obtain the parameter values ​​of the optimization variables under abnormal conditions.

[0111] In this embodiment, based on the constructed optimization objective function and constraint conditions, the optimization method is used to calculate the parameter variables to be optimized of the digital model of the lithium-ion battery equivalent circuit after the lithium-ion battery is determined to be in an abnormal state, and the parameter values ​​of the optimized variables under the abnormal state are calculated, that is, the parameter variables X to be optimized of the digital model of the lithium-ion battery equivalent circuit after the lithium-ion battery is determined to be in an abnormal state. +(t), wherein the optimization method includes but is not limited to the particle swarm optimization method, and a flow chart of the particle swarm optimization method for calculating the model parameters after the lithium-ion battery is determined to be operating abnormally, such as Figure 5 As shown. Set the particle swarm algorithm parameters, initialize the particle positions represented by the lithium-ion battery equivalent model parameters, calculate the fitness function (objective function) represented by each particle, reinitialize the particles that do not meet the constraints, sort the fitness function values ​​of this iteration, take the minimum fitness function value, and determine whether it is less than the historical minimum fitness function value. If it is not less than the historical minimum fitness function value, determine whether the cutoff condition is met. When the cutoff condition is met, update the historical minimum fitness function value and the historical minimum value of each particle. If it is less than the historical minimum fitness function value, update the historical minimum fitness function value and the historical minimum value of each particle, and then determine whether the cutoff condition is met. If the cutoff condition is not met, continue to use the particle swarm algorithm update formula to update the particle position, obtain the new particle position, calculate the fitness function (objective function) represented by each particle, reinitialize the particles that do not meet the constraints, sort the fitness function values ​​of this iteration, take the minimum fitness function value, and determine whether it is less than the historical minimum fitness function value until the cutoff condition is met.

[0112] Step 1044: Calculate the parameter change according to the parameter value of the optimized variable and the parameter value of the equivalent circuit digital model under the abnormal state.

[0113] In this embodiment, the deviation rate of the parameters of the lithium-ion battery equivalent circuit digital model before and after the lithium-ion battery is determined to be in an abnormal state, that is, the parameter change δ(t), is calculated. The optimized variable under the abnormal state is the parameter value of the lithium-ion battery equivalent circuit digital model after the lithium-ion battery is determined to be in an abnormal state, and the parameter value of the equivalent circuit digital model is the parameter value of the lithium-ion battery equivalent circuit digital model before the lithium-ion battery is determined to be in an abnormal state, that is, the parameter change δ(t)=|X(t)-X + (t)| / X(t).

[0114] Step 1045: Determine the fault type of the lithium-ion battery according to the parameter change and obtain a fault identification result.

[0115] In this embodiment, the type of lithium-ion battery fault is determined based on the parameter change (deviation rate). For example, if the deviation rate of the lithium-ion battery internal resistance parameter R0(t) is close to 1, that is, the lithium-ion battery internal resistance calculated in step 1044 is close to 0, it indicates that the lithium-ion battery may have an internal short circuit. The larger the deviation rate, the more obvious the fault. If it indicates other faults, such as lithium dendrites in lithium-ion batteries, then the electrochemical polarization parameter R in the equivalent model is required. b The deviation rate of (t) is close to 1. Other fault types can be judged in the same way.

[0116] Optionally, after determining the fault type of the lithium-ion battery based on the parameter change and obtaining the fault identification result, the method further includes: performing maintenance and overhaul on the lithium-ion battery based on the identification result; re-performing a preset test experiment on the lithium-ion battery after maintenance and overhaul to obtain the latest test data, and performing model parameter fitting and charge state estimation on the equivalent circuit structure of the lithium-ion battery based on the latest test data to establish the latest equivalent circuit digital model; and using the latest equivalent circuit digital model for fault identification of the lithium-ion battery after maintenance and overhaul.

[0117] In this embodiment, the lithium-ion battery fault type is determined based on the parameter change (deviation rate), and maintenance and repair of the lithium-ion battery is performed based on the lithium-ion battery fault type identification result. After the maintenance and repair, the lithium-ion battery test data under different environmental and test conditions is re-acquired to establish a new lithium-ion battery digital model module for subsequent lithium-ion battery anomaly determination and fault location.

[0118] It should be noted that in order to integrate the two types of methods (mechanism method and data method), and at the same time make use of the current digital twin concept and process control diagnostic mode, a lithium-ion battery abnormal state detection method based on data and model hybrid drive has been proposed. However, it is still in its infancy and is mostly used to calibrate the abnormal state of lithium-ion batteries, but not for determining the type of battery fault. The development and application of this type of technology is still insufficient. Therefore, the present invention focuses on the problems of abnormal detection and fault location of lithium-ion batteries, starting from the deviation between the physical system of lithium-ion batteries and the digital model, and proposes a lithium-ion battery fault location method based on dual-source data, laying the foundation for the large-scale safe application of lithium-ion batteries.

[0119] In the implementation of the present invention, a dual-source data structure for lithium-ion batteries is constructed using simulation data from a digital model of a lithium-ion battery equivalent circuit and physical test data from the lithium-ion battery. By analyzing the deviation in the dual-source data, abnormal lithium-ion battery operating conditions can be determined. After determining abnormal lithium-ion battery operation, simulation model parameters are adjusted by minimizing the error between the simulation model data and the measured data. Changes in the model parameters are analyzed to determine the type of lithium-ion battery fault and accurately locate the fault. Compared with existing data-based lithium-ion battery early warning methods, the present invention significantly reduces the requirements for lithium-ion battery data quality, thereby achieving reliable fault location. Compared with model-based lithium-ion operational risk methods, the present invention reduces the reliance on model accuracy, ensures the accuracy of lithium-ion battery abnormality determination and fault location, and effectively identifies the type of lithium-ion battery fault. Furthermore, compared with existing digital-analog-driven lithium-ion battery digital mirroring or digital twin technologies, the present invention can effectively simulate and accurately identify fault conditions by analyzing changes in model parameters, helping to improve the operational safety and reliability of lithium-ion battery systems, increase the accuracy and reliability of lithium-ion battery abnormality determination and fault type identification, and reduce the workload of operation and maintenance. By utilizing dual-source data to locate and identify lithium-ion battery fault types in lithium-ion battery application examples, the digital-analog driving capabilities of lithium-ion batteries are fully utilized, and the adaptive judgment of the lithium-ion battery operating safety status and the autonomous identification of fault risk types are effectively and simply realized, thereby improving the reliability and intelligence of lithium-ion battery applications, and bringing new methods and new paths for the safe implementation and convenient maintenance of large-scale, large-capacity lithium-ion batteries.

[0120] Example 2

[0121] Accordingly, see Figure 6 , Figure 6 FIG1 is a schematic diagram of the structure of a second embodiment of a lithium-ion battery fault identification system provided by the present invention. FIG1 is a schematic diagram of the execution flow of the lithium-ion battery fault identification system, as shown in FIG1 Figure 7 As shown. Figure 6 As shown, the fault identification system of the lithium-ion battery includes a model building module 601, a dual-source data module 602, an abnormality judgment module 603 and a fault identification module 604;

[0122] The model building module 601 is used to perform a preset test experiment on the lithium-ion battery to obtain test data, and based on the test data, perform model parameter fitting and charge state estimation on the equivalent circuit structure of the lithium-ion battery to establish an equivalent circuit digital model;

[0123] In this embodiment, the model building module 601 is the lithium-ion battery digital modeling module S1, which performs hybrid power pulse characteristic tests and open circuit voltage tests on lithium-ion batteries under different environments and test conditions. Based on the test data, a parameter fitting method is used to establish a lithium-ion battery equivalent circuit digital model and a state of charge estimation method.

[0124] The dual-source data module 602 is used to generate dual-source data of the lithium-ion battery based on the equivalent circuit digital model and the physical system of the lithium-ion battery; wherein the dual-source data includes simulation data and physical measurement data;

[0125] In this embodiment, the dual-source data module 602 is the dual-source data generation module S2, which calculates the simulation data of the lithium-ion battery equivalent circuit digital model and the measurement data of the physical system based on the output current of the lithium-ion battery physical system to complete the dual-source data generation process.

[0126] The abnormality judgment module 603 is used to statistically analyze the deviation similarity of the dual-source data of the lithium-ion battery during the preset operation time, and judge the current operation status of the lithium-ion battery based on the deviation similarity;

[0127] In this embodiment, the abnormality judgment module 603 is the battery abnormality judgment module S3, which analyzes the similarity of the deviation of the dual-source data under long-term operation of the lithium-ion battery and judges the abnormality of the lithium-ion battery.

[0128] The fault identification module 604 is used to minimize the parameter change of the equivalent circuit digital model according to the optimization variables under the preset dual-source deviation condition when the current operating state of the lithium-ion battery is abnormal, and determine the fault type of the lithium-ion battery based on the parameter change to obtain a fault identification result.

[0129] In this embodiment, the fault identification module 604 is the battery fault location and identification module S4. When a lithium-ion battery is abnormal, the parameters of the lithium-ion battery equivalent circuit digital model are optimized. Under the condition of the normal deviation level of the lithium-ion battery dual-source data, the change in the equivalent circuit digital model parameters is minimized, and the type of lithium-ion battery fault is determined based on the parameter changes.

[0130] By implementing the embodiments of the present invention, it is possible to realize digital twin simulation of lithium-ion battery systems, complete the evaluation of lithium-ion battery operating status, and form an accurate evaluation of lithium-ion battery safety risks and fault types, which brings benefits to the application reliability and maintenance convenience of lithium-ion battery energy storage power stations. The lithium-ion battery dual-source data structure is composed of lithium-ion battery equivalent circuit digital model simulation data and lithium-ion battery physical test data. By analyzing the deviation of the dual-source data, the abnormal operating conditions of the lithium-ion battery can be judged. After determining that the lithium-ion battery is operating abnormally, the simulation model parameters are adjusted by minimizing the error between the simulation model data and the measurement data. The model parameter changes are analyzed to obtain the lithium-ion battery fault type and accurately locate the fault. Compared with the existing data-based lithium-ion battery early warning method, the present invention greatly reduces the quality requirements for lithium-ion battery data, thereby achieving the reliability of fault location. Compared with the model-based lithium-ion operation risk method, the present invention reduces the dependence on model accuracy, ensuring the accuracy of lithium-ion battery abnormality judgment and fault location. In addition, compared with the existing digital mirror or digital twin technology of lithium-ion batteries based on digital-analog drive, the present invention can effectively simulate and accurately identify fault conditions by analyzing changes in model parameters, which helps to improve the operating safety and reliability of lithium-ion battery systems and reduce the workload of operation and maintenance.

[0131] The above-mentioned lithium-ion battery fault identification system can implement the lithium-ion battery fault identification method of the above-mentioned method embodiment. The optional options in the above-mentioned method embodiment also apply to this embodiment and are not described in detail here. The remaining contents of the embodiment of this application can refer to the contents of the above-mentioned method embodiment and are not repeated in this embodiment.

[0132] The above specific embodiments further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for identifying a fault of a lithium-ion battery, characterized in that: include: The lithium-ion battery is subjected to a preset test experiment to obtain test data, and based on the test data, the equivalent circuit structure of the lithium-ion battery is subjected to model parameter fitting and state of charge estimation to establish an equivalent circuit digital model; Generate dual-source data of the lithium-ion battery according to the equivalent circuit digital model and the physical system of the lithium-ion battery; wherein the dual-source data includes simulation data and physical measurement data; Statistically analyzing the deviation similarity of the dual-source data of the lithium-ion battery during the preset operation time, and judging the current operating state of the lithium-ion battery according to the deviation similarity; If the current operating state of the lithium-ion battery is abnormal, under preset dual-source deviation conditions, the parameter change of the equivalent circuit digital model is minimized according to the optimization variables, and the fault type of the lithium-ion battery is determined based on the parameter change to obtain a fault identification result; wherein the optimization variables are model parameters and influencing factors of the equivalent circuit digital model; the model parameters and the influencing factors are obtained by combining the equivalent circuit digital model with the test data; the influencing factors include state of charge, temperature, and current charge and discharge rate.

2. The lithium-ion battery fault identification method according to claim 1, wherein: According to the test data, the equivalent circuit structure of the lithium-ion battery is subjected to model parameter fitting and state of charge estimation to establish an equivalent circuit digital model, specifically: Selecting a preset order of an equivalent circuit digital model of undetermined parameters according to the equivalent circuit structure of the lithium-ion battery and the fault type; wherein the fault type includes internal short circuit, abnormal growth of the solid electrolyte interface SEI film, and lithium dendrites; According to the digital model of the equivalent circuit with undetermined parameters and the test data, the coefficients of the function to be determined are adjusted by a parameter optimization method to obtain a functional expression of the model parameters of the digital model of the equivalent circuit with undetermined parameters and the influencing factors; wherein the coefficients of the function to be determined are determined according to the functional relationship between the model parameters of the digital model of the equivalent circuit with undetermined parameters and the influencing factors; the parameter optimization method includes the least squares method, the particle swarm method and the simulated annealing method; The undetermined parameter equivalent circuit digital model and the function expression are used to estimate the lithium ion state of charge to obtain a first lithium ion battery state of charge value, and the equivalent circuit digital model is established based on the first lithium ion battery state of charge value and the undetermined parameter equivalent circuit digital model.

3. The lithium-ion battery fault identification method according to claim 2, wherein: Under the preset dual-source deviation condition, the parameter change of the equivalent circuit digital model is minimized according to the optimization variable, specifically: Constructing an optimization objective function and constraint conditions according to the optimization variables and the preset dual-source deviation conditions; Performing particle swarm optimization calculation on the optimization variable, the objective function, and the constraint condition to obtain parameter values ​​of the optimization variable under abnormal conditions; The parameter variation is calculated according to the parameter value of the optimization variable in the abnormal state and the parameter value of the equivalent circuit digital model.

4. The lithium-ion battery fault identification method according to claim 3, wherein: The optimization objective function is specifically: Wherein, J is the optimization objective function; X(t)=[x i (t)|i=1, 2, ..., n] represents the model parameters of the equivalent circuit digital model of the lithium-ion battery before the lithium-ion battery is determined to be in an abnormal state at time t; X + (t) = [x i + (t)|i=1, 2, ..., n] represents the parameter variables to be optimized of the equivalent circuit digital model of the lithium-ion battery after the lithium-ion battery is determined to be in an abnormal state at time t; n represents the number of the optimization variables; The constraints are specifically: s.t.α + ≤X + (t)≤β + Among them, α + and β + are the upper and lower bound constraints of the optimization variables respectively; f(X + (t)) represents the output voltage function, U M (t) represents the output voltage measurement value of the lithium-ion battery physical system at time t; is the preset dual-source deviation condition, and σ(t) represent the mean and standard deviation of the deviation values ​​of the dual-source data of the lithium-ion battery within a period L before time t, respectively.

5. The lithium-ion battery fault identification method according to claim 1, wherein: The step of generating the dual-source data of the lithium-ion battery according to the equivalent circuit digital model and the physical system of the lithium-ion battery is specifically as follows: Collecting the output current and output voltage of the lithium-ion battery physical system during actual operation, and calibrating the output voltage as the physical measurement data; The output current is input into the equivalent circuit digital model, and simulation data of the output voltage of the equivalent circuit digital model is obtained through simulation calculation.

6. The lithium-ion battery fault identification method according to claim 1, wherein: The statistical analysis of the deviation similarity of the dual-source data of the lithium-ion battery running for a preset time is specifically as follows: Running the lithium-ion battery for the preset time, and collecting all dual-source data of the lithium-ion battery during the preset time; A time window is determined based on factors affecting measurement errors, and a deviation value and deviation similarity of the dual-source data of the lithium-ion battery within the time window are calculated based on all dual-source data of the lithium-ion battery at the preset time; wherein the deviation similarity includes a deviation mean and a standard deviation.

7. The lithium-ion battery fault identification method according to claim 6, wherein: The current operating state of the lithium-ion battery is determined according to the deviation similarity, specifically: Obtaining a normal range according to the deviation mean and the standard deviation; Calculating a current deviation value of the lithium-ion battery according to the dual-source data of the lithium-ion battery in the current operating state; If the current deviation value of the lithium-ion battery is within the normal range, determining that the current operating state of the lithium-ion battery is a normal state; If the current deviation value of the lithium-ion battery is not within the normal range, it is determined that the current operating state of the lithium-ion battery is the abnormal state.

8. The lithium-ion battery fault identification method according to claim 1, wherein: The lithium-ion battery is subjected to a preset test experiment to obtain test data, specifically: Performing a hybrid power pulse characteristic test and an open circuit voltage test on the lithium-ion battery under different test environments and test conditions to obtain the test data; The test data are output voltage data and output current data of the lithium-ion battery under different temperatures, different charge and discharge rates, and different states of charge.

9. The lithium-ion battery fault identification method according to claim 1, wherein: After determining the fault type of the lithium-ion battery according to the parameter change and obtaining a fault identification result, the method further includes: Performing maintenance and repair on the lithium-ion battery according to the identification result; Re-performing a preset test experiment on the lithium-ion battery after maintenance and overhaul to obtain the latest test data, and then performing model parameter fitting and charge state estimation on the equivalent circuit structure of the lithium-ion battery based on the latest test data to establish the latest equivalent circuit digital model; The latest equivalent circuit digital model is used for fault identification of the lithium-ion battery after maintenance and overhaul.

10. A lithium-ion battery fault identification system, characterized in that: include: Establish model module, dual-source data module, abnormality judgment module and fault identification module; The model building module is used to perform a preset test experiment on the lithium-ion battery to obtain test data, and based on the test data, perform model parameter fitting and charge state estimation on the equivalent circuit structure of the lithium-ion battery to establish an equivalent circuit digital model; The dual-source data module is used to generate dual-source data of the lithium-ion battery according to the equivalent circuit digital model and the physical system of the lithium-ion battery; wherein the dual-source data includes simulation data and physical measurement data; The abnormality judgment module is used to statistically analyze the deviation similarity of the dual-source data of the lithium-ion battery during the preset operation time, and judge the current operation state of the lithium-ion battery according to the deviation similarity; The fault identification module is configured to minimize a parameter change of the equivalent circuit digital model based on optimization variables under preset dual-source deviation conditions when the current operating state of the lithium-ion battery is abnormal, and determine the fault type of the lithium-ion battery based on the parameter change to obtain a fault identification result. The optimization variables are model parameters and influencing factors of the equivalent circuit digital model; the model parameters and influencing factors are obtained by combining the equivalent circuit digital model with the test data; and the influencing factors include state of charge, temperature, and current charge and discharge rate.

Images