Online prediction method for residual life of battery pack

Abstract

The invention discloses a battery pack residual life on-line prediction method. The method comprises the following steps: S1, constructing and optimizing a charging amount calculation model by using battery pack cycle life off-line test data; and S2, predicting the residual life of the battery pack on line by using the charging amount calculation model and the battery pack operation data under the actual working condition. According to the method, laboratory research results are combined with actual operation data of the battery, online modeling and model parameter correction are performed according to historical data of the battery, and the RUL of the battery pack is predicted through a cyclic calculation method, so that continuous improvement of the online prediction effect of the RUL of the battery pack is realized, and on the basis of an existing commercially-applied battery management technology, the reliability of the battery pack is improved. And the residual life of the battery pack can be accurately and quickly predicted online by a machine with general calculation power.

Description

Technical Field

[0001] The present invention relates to the technical field of battery management, and in particular to an online prediction method for the remaining life of a battery pack. Background Art

[0002] Real-time prediction of the remaining useful life (RUL) of lithium batteries is of great significance for the safe operation of lithium battery systems and reducing operation and maintenance costs. At present, methods such as particle filtering, machine learning, and fusion prediction methods for online prediction of the RUL of lithium-ion battery packs have been studied in laboratories. However, the difficulty of predicting the RUL of lithium-ion battery packs lies in:

[0003] 1. There is a big gap between laboratory test conditions and actual working conditions. For example, the laboratory usually tests under a single condition with constant stress, while the actual environment usually has a complex stress under complex dynamic operating conditions. Therefore, the single cell RUL prediction model established using laboratory data lacks accuracy guarantee for RUL prediction in the actual working environment of the battery pack.

[0004] 2. The RUL of lithium-ion batteries not only depends on the minimum life of the single cell, but is also affected by the consistency of the RUL of each cell, which further increases the difficulty of online prediction of the RUL of lithium-ion battery packs. Summary of the invention

[0005] The present invention aims to improve the online prediction accuracy of the RUL of a lithium-ion battery pack and provides an online prediction method for the remaining service life of a battery pack.

[0006] To achieve this object, the present invention adopts the following technical solutions:

[0007] A method for online prediction of remaining life of a battery pack is provided, comprising the steps of:

[0008] S1, build and optimize the charging capacity calculation model using the offline test data of battery pack cycle life;

[0009] S2, using the charge capacity calculation model and the battery pack operation data under actual working conditions to online predict the remaining life of the battery pack.

[0010] Preferably, in step S1, the method for constructing and optimizing the charging capacity calculation model comprises the steps of:

[0011] S11, perform a constant current or constant power charge and discharge cycle test on the battery pack sample to obtain offline test data, including the battery pack single charging time t, charging current, charging power, remaining power increment dsoc, first remaining power at the start of charging, second remaining power at the end of charging, charging capacity, cumulative charging capacity tc, battery pack voltage, and number of cycle tests collected after the battery pack sample undergoes each constant current or constant power charge and discharge cycle;

[0012] S12, constructing the charging capacity calculation model expressed by the following formula (1):

[0013]

[0014] In formula (1), C n Indicates the current charge capacity of the battery pack, C n-1 Indicates the last charge of the battery pack;

[0015] w1, w2, w3 are weights, and w1+w2+w3=1;

[0016] k1 represents the equivalent charging current or charging power during the charging process of the battery pack;

[0017] k2 represents the maximum charge capacity of the battery pack for the first time. When this value cannot be obtained, the rated charge capacity Q of the battery pack is taken. r ;

[0018] k3 represents the attenuation rate of the battery pack's single charge capacity;

[0019] f1(n) represents the product of the single charge capacity decay rates of the battery pack in the first n cycles in step S11, which is obtained by formula (2):

[0020]

[0021] in, Q i and Q i-1 represent the charge capacity of the i-th cycle and the charge capacity of the i-1-th cycle respectively;

[0022] S13, k1, k2, k3 in formula (1) are optimized and inserted into formula (1) to obtain the optimized charging capacity calculation model expressed by the following formula (3):

[0023]

[0024] k1m, k2m, k3m are the optimal solutions corresponding to k1, k2, k3 respectively.

[0025] Preferably, the method for optimizing k1, k2, and k3 in formula (1) comprises the steps of:

[0026] S121, taking a plurality of continuous charge and discharge data segments collected from the battery pack sample, and calculating the charge and discharge data according to the C n , t, dsoc, tc data, and obtain several solution equations about k1, k2, k3;

[0027] S122, determining whether the number of equations to be solved is equal to "4",

[0028] If so, the LU decomposition method is used to obtain the optimal solutions k1m, k2m, and k3m corresponding to k1, k2, and k3 respectively;

[0029] If not, proceed to step S123;

[0030] S123, determining whether the number of equations to be solved is greater than "4",

[0031] If so, the least squares method is applied to obtain the optimal solutions k1m, k2m, and k3m corresponding to k1, k2, and k3 respectively;

[0032] If not, the optimal solution corresponding to at least one of k1, k2, and k3 is solved by the attenuation characteristic analysis method.

[0033] Preferably, under the "if no" judgment condition of step S122, let the optimal solution k1m of k1 be the current average value of the charging current of each of the charging and discharging data segments extracted in step S121; let the optimal solution k2m of k2 be the rated charging capacity Q of the battery pack sample. r The optimal solution k3m of k3 is calculated according to the following formula (4):

[0034]

[0035] Preferably, the method for assigning weights w1, w2, and w3 comprises the steps of:

[0036] A1, randomly select a set of battery pack charge and discharge data collected in the cycle test, and then calculate the k1×t and C in the charge and discharge data n The absolute value of the first difference of the true value, With C n The absolute value of the second difference of the true value, k3×C n-1 With C n The absolute value of the third difference of the true value;

[0037] A2, calculating the sum of the first difference absolute value, the second difference absolute value, and the third difference absolute value;

[0038] A3, calculate the first ratio of the absolute value of the first difference to the sum as the weight value of w1; calculate the second ratio of the absolute value of the second difference to the sum as the weight value of w2; calculate the third ratio of the absolute value of the third difference to the sum as the weight value of w3.

[0039] Preferably, in step S2, the method for predicting the remaining battery life of the battery pack online comprises the steps of:

[0040] S21, assigning an initial value of "0" to the remaining charging times L of the battery pack;

[0041] S22, for the battery pack, calculate the charge capacity C1 of the current charge by formula (3), and the model independent variables for calculating C1 are t=t1, dsoc=dsoc1, tc=tc1, wherein tc1 represents the current cumulative charge capacity of the battery pack; dsoc1 represents the remaining charge increment of the battery pack from the start to the end of the current charge of the battery pack; t1 represents the charging time of the battery pack at the current time;

[0042] S23, calculating the actual maximum charge capacity of the battery pack at present

[0043] S24, determining the actual maximum chargeable capacity value Is it less than or equal to the minimum design capacity of the battery pack?

[0044] If yes, output the accumulated result of the remaining charging times L of the battery pack as the prediction result of the remaining life of the battery pack;

[0045] If not, then "1" is added to the remaining charging times L, and tc1=C1+tc1 is set, and then the process returns to step S22.

[0046] Preferably, dsoc1≥0.8.

[0047] Preferably, t1 is obtained by solving the following formula (5):

[0048]

[0049] In formula (5), t0 represents the single charging time of the battery pack to which dsoc0, which has the value closest to dsoc1, belongs in the historical data of the online test of the battery pack;

[0050] dsoc0 represents the remaining power increment of the battery pack sample during a single charging process in the historical data.

[0051] Preferably, the minimum design power of the battery pack is the product of the rated voltage of each single cell in the battery pack, the initial power of the single cell, the total number of single cells, and a constant p.

[0052] The present invention combines laboratory research results with actual battery operation data, builds online models and corrects model parameters based on historical battery data, and predicts battery pack RUL through a cyclic calculation method, thereby achieving continuous improvement in the online prediction effect of battery pack RUL. This allows machines with average computing power to accurately and quickly predict the remaining life of a battery pack online based on existing commercially applied battery management technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0054] Figure 1 is a diagram of implementation steps of a method for online prediction of remaining life of a battery pack provided by an embodiment of the present invention;

[0055] Figure 2 This is a graph showing the battery pack's room temperature charge capacity attenuation rate versus cycle number. DETAILED DESCRIPTION

[0056] The technical solution of the present invention is further described below with reference to the accompanying drawings and through specific implementation methods.

[0057] Among them, the drawings are only used for illustrative explanations, and they only represent schematic diagrams rather than actual pictures, and should not be understood as limitations on this patent; in order to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0058] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right", "inner", "outer", etc. appear, the orientation or position relationship indicated is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, the terms describing the position relationship in the drawings are only used for illustrative purposes and cannot be understood as a limitation on this patent. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0059] In the description of the present invention, unless otherwise clearly specified and limited, if the term "connection" or the like appears to indicate the connection relationship between components, the term should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two components or the interaction relationship between two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0060] Taking the lithium-ion battery energy storage power station as an example of the application scenario of online prediction of the remaining battery life, assuming that the battery group (i.e., cluster) has historical charging and discharging data (for example, charging and discharging historical data of more than 1 month), the remaining battery life of the battery group can be continuously predicted online according to the following method. Taking the lithium iron phosphate energy storage battery group as an example, an online prediction method for the remaining battery life of a battery group provided by an embodiment of the present invention includes the following three technical parts:

[0061] 1. Battery pack cycle life offline test

[0062] 1. Take a battery pack sample and perform constant current or constant power charge and discharge cycle test at different charge and discharge depths (such as 20% dsoc, 60% dsoc, 100% dsoc) and different currents (such as 0.3C, 0.5C, 1C, 2C) at room temperature to ensure that the operating voltage range of the single cells in the battery pack is within the limit range of the lower and upper voltages (such as the limit range of 2.5V-3.65V), until the charge or discharge amount of the battery pack reaches the set limit (such as 80% of the initial charge amount). The charge and discharge depth dsoc is the difference between the soc (ratio of the remaining power to the total power) of the battery pack at the start of charge or discharge and the soc when the charge or discharge is terminated.

[0063] After a specific cycle (for example, after every 10 constant current charge and discharge cycles), the battery pack sample is tested at rated current by first charging and then discharging to determine the actual maximum charge capacity of the battery pack. This cycle test is repeated until the battery pack charge capacity reaches the preset end-of-life value (such as 80% of the initial charge capacity).

[0064] Finally, the battery pack voltage, charging current, charging power, number of cycle tests, the first remaining power at the start of charging, the second remaining power at the end of charging, the single charging time t, the remaining power increment dsoc, the charging amount, and the cumulative charging amount collected during each cycle test are obtained.

[0065] For example, assuming that the time interval for collecting offline test data for the battery pack sample is every 10 constant current charge and discharge cycles, when 10 constant current charge and discharge cycles are completed, the battery pack sample is tested at rated current first full and then empty. During the full charge process, the charging current of the battery pack sample, the first remaining capacity at the beginning of charging, the battery charging time t, and the second remaining capacity at the end of charging are collected, and then the difference between the second remaining capacity and the first remaining capacity is calculated as the remaining capacity increment dsoc, and the charging capacity of the battery pack sample is calculated based on parameters such as the charging current, charging power, and charging time t. The cumulative charging capacity is the sum of each historical charging capacity of the battery pack sample. The number of cycle tests refers to the number of constant current charge and discharge cycles that the battery pack sample has undergone up to the time of offline test data collection.

[0066] 2. Battery Pack Charge Modeling

[0067] The chargeable capacity C of the battery pack is related to the single charging time t of the battery pack, the soc increment dsoc (i.e., the difference between the soc at the end of charging and the soc at the beginning of charging), and the cumulative charge capacity tc. As tc increases, the charging capacity of the battery pack gradually decreases. This embodiment constructs an online calculation model of the charge capacity of C, t, dsoc, and tc through the following formula (1).

[0068]

[0069] In formula (1), C n Indicates the current charge capacity of the battery pack, C n-1 Indicates the last charge of the battery pack;

[0070] w1, w2, w3 are weights, and w1+w2+w3=1;

[0071] k1 represents the equivalent charging current or charging power during the charging process of the battery pack;

[0072] k2 represents the maximum charge capacity of the battery pack for the first time. When this value cannot be obtained, the rated charge capacity Q of the battery pack is taken. r ;

[0073] k3 represents the attenuation rate of the battery pack's single charge capacity;

[0074] f1(n) represents the product of the single charge capacity decay rates of the battery pack in the first n cycles in step S11, which is obtained by formula (2):

[0075]

[0076] in, Q i and Q i-1represent the charge capacity of the i-th cycle and the charge capacity of the i-1-th cycle respectively;

[0077] In the above offline test data, C n , t, dsoc, tc are known values. Therefore, formula (1) is converted into a multiple regression equation of k1, k2, k3 for solution. Through multiple charging, such as charging data under offline conditions for one month, the equation group about k1, k2, k3 expressed in the above formula (1) is solved. After solving the values ​​of k1, k2, k3, specify t, dsoc, tc, and the charge capacity C of the battery pack can be calculated. n The value of .

[0078] The specific method for solving k1, k2, and k3 in formula (1) is:

[0079] (1) Assign specific weight values ​​to w1, w2, and w3. The assignment method is:

[0080] Take any set of battery pack charge and discharge data collected in the cycle test, and then calculate the k1×t and C in the charge and discharge data n The absolute value of the first difference between the true value, k2×dsoc and C n The absolute value of the second difference between the true value, k3 and C n The absolute value of the third difference of the true value; then the sum of the absolute value of the first difference, the absolute value of the second difference, and the absolute value of the third difference is calculated;

[0081] Then, the first ratio of the absolute value of the first difference to the sum is calculated as the weight value of w1; the ratio of the absolute value of the second difference to the sum is calculated as the weight value of w2; and the ratio of the absolute value of the third difference to the sum is calculated as the weight value of w3.

[0082] For example, the values ​​assigned to w1, w2, and w3 are 0.64, 0.21, and 0.15 respectively.

[0083] (2) Take a number of consecutive charge and discharge data segments from the battery pack sample, and then calculate the C n , t, dsoc, tc data, and obtain several equations for k1, k2, and k3. The rated capacity of the battery pack is 240Ah, the rated voltage of the battery pack is 768V, the rated voltage of the single cell is 3.2V, the initial capacity of the single cell is 240Ah, and the total number of single cells is 240 strings.

[0084] The continuous charge and discharge data segments referred to here are: for example, assuming that the battery pack sample is tested with full charge and then empty under rated current when completing 10 constant current charge and discharge cycles (for example, recorded as test a), then the completed full charge and then empty test is a charge and discharge data segment, and offline test data is collected once for the battery pack sample in this charge and discharge data segment; then, after performing 10 constant current charge and discharge cycles, the next full charge and then empty test is performed under rated current (for example, recorded as test b), and this full charge and then empty test b is a charge and discharge test segment continuous with test a.

[0085] The number of equations to be solved regarding k1, k2, and k3 is related to the number of continuous charge and discharge data segments extracted, and the number of continuous charge and discharge data segments affects the performance of the constructed online prediction model for the remaining life of the battery pack.

[0086] After obtaining several solution equations for k1, k2, and k3, determine whether the number of solution equations is equal to "4".

[0087] If so, the LU decomposition method is used to obtain the optimal solutions k1m, k2m, and k3m corresponding to k1, k2, and k3 respectively;

[0088] If not, go to the next step to continue to determine whether the number of equations to be solved is greater than "4";

[0089] The following are examples of the LU decomposition method for solving the optimal solutions k1m, k2m, and k3m corresponding to k1, k2, and k3 respectively:

[0090] 1) According to formula (1) and formula (2), the four charging data are sorted into a matrix form about k1, k2, and k3:

[0091]

[0092] 2) Perform LU decomposition on the coefficient matrix on the left side of the matrix equation and solve it by computer in the MATLAB environment to obtain

[0093]

[0094] 3) Solve the a priori equations and obtain the intermediate solution, that is,

[0095]

[0096] 4) Solve the a posteriori equations and obtain the final solution, that is,

[0097]

[0098] Under the above "if not" judgment condition, continue to judge whether the number of equations to be solved is greater than "4".

[0099] If so, the least squares method is applied to obtain the optimal solutions k1m, k2m, and k3m corresponding to k1, k2, and k3 respectively;

[0100] If not, then the optimal solution corresponding to at least one of k1, k2, and k3 is solved by the attenuation characteristic analysis method. In this embodiment, it is preferred to solve the optimal solution k3m corresponding to k3; let the optimal solution k1m of k1 be the current average value of the charging current of each charging and discharging data segment extracted; let the optimal solution k2m of k2 be the rated charge capacity Q of the battery pack sample r The optimal solution k3m of k3 is calculated according to formula (4):

[0101]

[0102] The following are examples of the method for solving the optimal solutions k1m, k2m, and k3m corresponding to k1, k2, and k3 respectively using the least squares method:

[0103] 1) According to formula (1) and formula (2), the five charging data are sorted into a matrix form about k1, k2, and k3:

[0104]

[0105] 2) Using the lsqnonneg function in the MATLAB environment, we can obtain the final solution:

[0106]

[0107] The following combination Figure 2 The solution method of k3m is specifically explained using formula (4):

[0108] First, obtain an array consisting of the number of cycles and the corresponding charge capacity in the offline test data, that is,

[0109] Apply formula (4) to obtain the array The first column is the number of cycles, and the second column is the k3m corresponding to the number of cycles.

[0110] 3. Prediction of remaining battery life

[0111] The remaining life of a battery pack is a decreasing value. If the remaining life of a battery pack needs to be calculated at a certain point in its operation, the historical operation data of the battery pack during the historical period is first obtained to complete the remaining life prediction modeling. Then, based on the future operation data of the future period, the battery cycle process is simulated to iterate the model, and then the remaining life is predicted. The specific steps include the following:

[0112] (1) The remaining number of charging times L is set to an initial value of "0";

[0113] (2) For the battery pack, the next charge amount C1 that is continuous with the current charge is calculated by formula (3), and the model independent variables for calculating C1 are t=t1, dsoc=dsoc1, tc=tc1, where tc1 represents the current cumulative charge amount of the battery pack; dsoc1 represents the remaining power increment of the battery pack from the start to the end of the current charge of the battery pack; t1 represents the charging time of the battery pack at the current time;

[0114] t1 is calculated by the following formula (4):

[0115]

[0116] In formula (4), t0 represents the single charging time of the battery pack carried in the data to which dsoc0, which is closest to the value of dsoc1, belongs in the historical data of the online test of the battery pack;

[0117] dsoc0 represents the remaining power increment of the battery pack sample carried in the offline data.

[0118] For example, suppose that when the battery pack sample is tested offline, three dsoc0 values ​​are obtained, which are 0.78, 0.81, and 0.82 respectively, and dsoc1=0.8. Then dsoc0=0.81, which is closest to the dsoc1 value, is obtained. Then, the single charging time of the battery pack sample carried in the offline data collected in the offline test to which dsoc0 belongs is obtained and recorded as t0.

[0119] (3) Calculate the current charging rate of the battery pack

[0120] (4) Determine the charging rate Is it less than or equal to the minimum design capacity of the battery pack?

[0121] If yes, then the accumulated result of the remaining charging times L of the battery pack is output as the remaining life prediction result of the battery pack;

[0122] If not, then the remaining charging times L of the battery pack are accumulated by "1", and tc1=C1+tc1 is set, and then the process returns to step S22.

[0123] The minimum design capacity of the battery pack Cmin = total number of single cells × single cell rated voltage × single cell initial capacity × constant p. The constant p is set by the user. Usually, the battery pack ends its life when p = 0.8.

[0124] In summary, the present invention combines laboratory research results with actual battery operation data, builds online model and corrects model parameters based on historical battery data, and predicts battery pack RUL through a cyclic calculation method, thereby achieving continuous improvement in the online prediction effect of battery pack RUL. This allows machines with average computing power to accurately and quickly predict the remaining life of a battery pack online based on existing commercial battery management technology.

[0125] It should be noted that the above specific implementations are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art should understand that various modifications, equivalent substitutions, changes, etc. can be made to the present invention. However, as long as these changes do not deviate from the spirit of the present invention, they should be within the scope of protection of the present invention. In addition, some terms used in the specification and claims of this application are not restrictive, but are only for the convenience of description.

Claims

1. A method for online prediction of remaining battery life, characterized in that: Includes steps: S1, build and optimize the charging capacity calculation model using the offline test data of battery pack cycle life; S2, using the charge capacity calculation model and the battery pack operation data under actual working conditions to online predict the remaining life of the battery pack.

2. The method for online prediction of remaining battery life according to claim 1, characterized in that: In step S1, the method for constructing and optimizing the charging capacity calculation model includes the following steps: S11, perform a constant current or constant power charge and discharge cycle test on the battery pack sample to obtain offline test data, including the battery pack single charging time t, charging current, charging power, remaining power increment dsoc, the first remaining power at the start of charging, the second remaining power at the end of charging, the charge capacity, the cumulative charge capacity tc, the battery pack voltage, and the number of cycle tests collected after the battery pack sample undergoes each constant current charge and discharge cycle; S12, constructing the charging capacity calculation model expressed by the following formula (1): In formula (1), C n Indicates the current charge capacity of the battery pack, C n-1 Indicates the last charge of the battery pack; w1, w2, w3 are weights, and w1+w2+w3=1; k1 represents the equivalent charging current or charging power during the charging process of the battery pack; k2 represents the maximum charge capacity of the battery pack for the first time. When this value cannot be obtained, the rated charge capacity Q of the battery pack is taken. r ; k3 represents the attenuation rate of the battery pack's single charge capacity; f1(n) represents the product of the single charge capacity decay rates of the battery pack in the first n cycles in step S11, which is obtained by formula (2): in, Q i and Q i-1 represent the charge capacity of the i-th cycle and the charge capacity of the i-1-th cycle respectively; S13, k1, k2, k3 in formula (1) are optimized and inserted into formula (1) to obtain the optimized charging capacity calculation model expressed by the following formula (3): k1m, k2m, k3m are the optimal solutions corresponding to k1, k2, k3 respectively.

3. The method for online prediction of remaining battery life according to claim 2, characterized in that: The method for optimizing k1, k2, and k3 in formula (1) comprises the following steps: S121, taking a plurality of continuous charge and discharge data segments collected from the battery pack sample, and calculating the charge and discharge data according to the C n , t, dsoc, tc data, and obtain several solution equations about k1, k2, k3; S122, determining whether the number of equations to be solved is equal to "4", If so, the LU decomposition method is used to obtain the optimal solutions k1m, k2m, and k3m corresponding to k1, k2, and k3 respectively; If not, proceed to step S123; S123, determining whether the number of equations to be solved is greater than "4", If so, the least squares method is applied to obtain the optimal solutions k1m, k2m, and k3m corresponding to k1, k2, and k3 respectively; If not, the optimal solution corresponding to at least one of k1, k2, and k3 is solved by the attenuation characteristic analysis method.

4. The method for online prediction of remaining battery life according to claim 3, characterized in that: Under the "if no" judgment condition of step S122, let the optimal solution k1m of k1 be the current average value of the charging current of each of the charging and discharging data segments extracted in step S121; let the optimal solution k2m of k2 be the rated charging capacity Q of the battery pack sample. r The optimal solution k3m of k3 is calculated according to the following formula (4):

5. The method for online prediction of remaining battery life according to claim 2, characterized in that: The method for assigning weights w1, w2, and w3 includes the following steps: A1, randomly select a set of battery pack charge and discharge data collected in the cycle test, and then calculate the k1×t and C in the charge and discharge data n The absolute value of the first difference of the true value, With C n The absolute value of the second difference of the true value, k3×C n-1 With C n The absolute value of the third difference of the true value; A2, calculating the sum of the first difference absolute value, the second difference absolute value, and the third difference absolute value; A3, calculating a first ratio of the absolute value of the first difference to the sum as a weight value of w1; Calculate a second ratio of the absolute value of the second difference to the sum as a weight value of w2; A third ratio of the third absolute value of the difference to the sum is calculated as a weight value of w3.

6. The method for online prediction of remaining battery life according to any one of claims 2 to 5, characterized in that: In step S2, the method for predicting the remaining battery life of the battery pack online comprises the following steps: S21, assigning an initial value of "0" to the remaining charging times L of the battery pack; S22, for the battery pack, calculate the charge capacity C1 of the current charge by formula (3), and the model independent variables for calculating C1 are t=t1, dsoc=dsoc1, tc=tc1, wherein tc1 represents the current cumulative charge capacity of the battery pack; dsoc1 represents the remaining charge increment of the battery pack from the start to the end of the current charge of the battery pack; t1 represents the charging time of the battery pack at the current time; S23, calculating the actual maximum charge capacity of the battery pack at present S24, determining the actual maximum chargeable capacity value Is it less than or equal to the minimum design capacity of the battery pack? If yes, output the accumulated result of the remaining charging times L of the battery pack as the prediction result of the remaining life of the battery pack; If not, then add "1" to the remaining charging times L, set tc1=C1+tc1, and then return to step S22.

7. The method for online prediction of remaining battery life according to claim 6, characterized in that: dsoc1≥0.

8.

8. The method for online prediction of remaining battery life according to claim 6, characterized in that: t1 is obtained by solving the following formula (5): In formula (5), t0 represents the single charging time of the battery pack to which dsoc0, which has the value closest to dsoc1, belongs in the historical data of the online test of the battery pack; dsoc0 represents the remaining power increment of the battery pack sample during a single charging process in the historical data.

9. The method for online prediction of remaining battery life according to claim 6, characterized in that: The minimum design power of the battery pack is the product of the rated voltage of each single cell in the battery pack, the initial power of the single cell, the total number of single cells, and a constant p.

Images