A method for estimating the health status of lithium-ion batteries under different ambient temperatures

Abstract

The present invention belongs to the technical field of battery health state estimation, and in particular relates to a method for estimating the health state of a lithium-ion battery at different ambient temperatures, comprising the following steps: step 1, performing a charging experiment on a lithium-ion battery at different ambient temperatures, step 2, calculating the SOH of the lithium-ion battery at different ambient temperatures, and the ratio of the SOH at different ambient temperatures to the SOH at room temperature, step 3, establishing a Thevenin equivalent circuit model, an open circuit voltage U oc The present invention establishes a relationship model between the battery's SOH at different temperatures and the SOH at room temperature, obtains the room temperature SOH value corresponding to the battery at different ambient temperatures, and determines whether the battery needs to be replaced.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery health state prediction, and in particular relates to a method for predicting the health state of a lithium-ion battery under different ambient temperatures. Background Art

[0002] At present, lithium-ion batteries are gradually being used in aerospace, defense, military, transportation, and electronic mobile equipment due to their advantages such as low self-discharge rate, long cycle life, no memory, and environmental friendliness. However, in order to ensure the safety and efficiency of lithium-ion batteries during use, it is necessary to use a battery management system to monitor the state of lithium-ion batteries, such as state of charge, state of health, and power state. Among them, the health state, as a quantitative indicator for evaluating the aging degree of lithium-ion batteries, reflects the aging of lithium-ion batteries. The most intuitive manifestation is the reduction in the maximum available capacity of lithium-ion batteries and the decrease in power level, which is reflected internally as capacity decay and increased internal resistance. Therefore, accurately estimating the health state is a key issue in the application of lithium-ion batteries.

[0003] None of the current state of health SOH estimation techniques have studied the relationship between the battery's SOH at different temperatures and the SOH at room temperature. At different ambient temperatures, the chemical reaction speed inside the battery and the speed of lithium ions are different, resulting in changes in the actual capacity and aging rate of the battery. Too high or too low a temperature will affect the performance of the battery. In particular, the capacity of the battery at low temperature is significantly different from the capacity of the battery at room temperature. Taking electric vehicles as an example, in the industry standard QC / T743-2006 for lithium-ion batteries for electric vehicles, Section 6.2.11 stipulates the judgment standard for the battery's health state, that is, at 20℃±2℃, the battery is subjected to a cycle life test. When the battery capacity drops to 80% of its rated capacity, the lithium-ion battery is judged to be failed. Here, the standard temperature of 20℃±2℃ is collectively referred to as room temperature. There is a situation where the battery's SOH is already lower than 80% of the initial SOH at low temperature. Moreover, the battery's SOH value at low temperature may be different from its SOH value at 20℃±2℃. Therefore, according to the standard, it is still impossible to determine whether the battery has failed when the ambient temperature is different. It is necessary to obtain the SOH of the lithium-ion battery at the current temperature corresponding to 20℃±2℃ to determine whether the battery needs to be replaced. Summary of the invention

[0004] The main purpose of the present invention is to solve the problems existing in the prior art. The method for predicting the health status of a lithium-ion battery of the present invention establishes a mapping relationship between the SOH of a lithium-ion battery at different temperatures and the SOH at room temperature, thereby obtaining the room temperature SOH value of the lithium-ion battery corresponding to different ambient temperatures, so that when the ambient temperature is different, it can be accurately determined whether the battery needs to be replaced.

[0005] The technical problem solved by the present invention is achieved by adopting the following technical solution: A method for estimating the health status of a lithium-ion battery under different ambient temperatures comprises the following steps:

[0006] Step 1, conduct a charging experiment on the lithium-ion battery at different ambient temperatures, record the current ambient temperature and collect the voltage, current and time during the charging process of the lithium-ion battery, and multiply the current and sampling time during the charging process to obtain the current capacity of the lithium-ion battery;

[0007] Step 2, according to the SOH capacity definition, calculate the SOH of the lithium-ion battery at different ambient temperatures to obtain the actual SOH values ​​at different ambient temperatures, and calculate the ratio of the SOH of the lithium-ion battery at different ambient temperatures to the SOH at room temperature to obtain the actual room temperature SOH values ​​corresponding to the different ambient temperatures;

[0008] Step 3: Establish the Thevenin equivalent circuit model and use polynomial fitting to establish the open circuit voltage U oc The relationship model between the open circuit voltage U and the state of charge SOC is based on the voltage and current data collected during the charging process of the lithium-ion battery, and the recursive least squares method is used to calculate the open circuit voltage U oc The parameters of the relationship model with the state of charge SOC are identified, and the ohmic internal resistance R of the battery at different ambient temperatures is obtained when the state of charge SOC is specified. 0 , Polarization internal resistance R p , voltage value V;

[0009] Step 4: Set the ohmic internal resistance R at a specific state of charge SOC 0 , Polarization internal resistance R p , voltage value V, SOC value, temperature T are normalized, and then the processed ohmic internal resistance R 0 , Polarization internal resistance R p , voltage value V, SOC value, temperature T are the inputs of Gaussian process regression GPR algorithm, SOH at different ambient temperatures is the output of Gaussian process regression GPR algorithm, a model for estimating SOH at different ambient temperatures is constructed, and the model is used to estimate SOH at different ambient temperatures to obtain the estimated SOH values ​​at different ambient temperatures;

[0010] Step 5: Set the ohmic internal resistance R 0 , Polarization internal resistance R p, voltage value V, SOC value, temperature T, and SOH at different ambient temperatures are used as inputs of the least squares support vector regression LSSVR algorithm, and the particle swarm optimization PSO algorithm is used to optimize the parameters of the least squares support vector regression LSSVR algorithm, and a model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is constructed. The model is used to estimate the ratio of SOH at different temperatures to SOH at room temperature of lithium-ion batteries, and the estimated room temperature SOH values ​​corresponding to different ambient temperatures are obtained;

[0011] Step 6. Calculate the maximum absolute error MAX Absolute error, the mean absolute error MAE, and the root mean square error RMSE based on the estimated SOH values ​​at different ambient temperatures and the actual SOH values ​​at different ambient temperatures, as well as the estimated room temperature SOH values ​​corresponding to different ambient temperatures and the actual room temperature SOH values ​​corresponding to different ambient temperatures. The performance of the estimated SOH values ​​at different ambient temperatures and the estimated room temperature SOH values ​​corresponding to different ambient temperatures are evaluated by the maximum absolute error MAXAbsolute error, the mean absolute error MAE, and the root mean square error RMSE.

[0012] Furthermore, step 1, performing charging experiments on the lithium-ion battery at different ambient temperatures, further includes performing charging experiments on the lithium-ion battery at different ambient temperatures using a standard constant current and constant voltage charging method.

[0013] Further, step 2, according to the SOH capacity definition, calculates the SOH of the lithium-ion battery at different ambient temperatures to obtain the actual SOH values ​​at different ambient temperatures, and calculates the ratio of the SOH of the lithium-ion battery at different ambient temperatures to the SOH at room temperature to obtain the actual room temperature SOH values ​​corresponding to the different ambient temperatures, further including,

[0014] According to the definition of SOH capacity, calculate the SOH of lithium-ion batteries at different ambient temperatures.

[0015]

[0016] In the above formula, Q rated Q is the rated capacity of a new battery when it leaves the factory. aged It is the actual maximum available capacity of the battery after it is put into use;

[0017] Calculate the ratio of SOH of lithium-ion batteries at different ambient temperatures to SOH at room temperature.

[0018]

[0019] In the above formula, SOH is the health status value of the battery at different ambient temperatures. r It is the healthy state value at room temperature.

[0020] Furthermore, in step 3, the Thevenin equivalent circuit model is established, and the open circuit voltage U is established by polynomial fitting. oc The relationship model with the state of charge SOC further includes:

[0021] The Thevenin equivalent circuit model is established as

[0022]

[0023] Where U t Indicates terminal voltage, U oc Represents the open circuit voltage of the battery, R 0 represents the ohmic internal resistance, I represents the current, and U P represents the polarization voltage, C P Represents polarized capacitance, R P represents polarization resistance;

[0024] The relationship model between open circuit voltage Uoc and state of charge SOC is established as

[0025] U oc =K 0 +K 1 SOC+K 2 SOC 2 +K 3 SOC 3 +K 4 SOC 4

[0026] In the formula, K 0 , K 1 , K 2 , K 3 , K 4 is the coefficient.

[0027] Further, in step 3, when the state of charge SOC is specified, the ohmic internal resistance R of the battery at different ambient temperatures is obtained. 0 , Polarization internal resistance R p , voltage value V, further including,

[0028] The state of charge at the end of constant current charging is recorded as the maxSOC value. When the specific state of charge SOC is between 20% and the maxSOC value and the value increases by 5%, the ohmic internal resistance R of the battery at different ambient temperatures is obtained. 0 , polarization internal resistance R p , voltage value V.

[0029] Further, in step 4, the ohmic internal resistance R at a specific state of charge SOC is 0 , Polarization internal resistance R p, voltage value V, SOC value, temperature T are normalized, further including:

[0030] The ohmic internal resistance R at a specific state of charge SOC 0 , Polarization internal resistance R p , voltage value V, SOC value, and temperature T are used as characteristic parameters for estimating the SOH model under different ambient temperatures, and the mapminmax function of MATLAB is used to normalize the parameters to [1,5].

[0031] Further, in step 4, the processed ohmic internal resistance R 0 , Polarization internal resistance R p , voltage value V, SOC value, temperature T are inputs of Gaussian process regression GPR algorithm, SOH at different ambient temperatures is the output of Gaussian process regression GPR algorithm, and a model for estimating SOH at different ambient temperatures is constructed, further including:

[0032] The estimated SOH model at different ambient temperatures is:

[0033] f(x)~gp(m(x),k f (x,x′))

[0034] Among them, m(x), k f (x,x') is the mean and covariance function, m(x), k f (x,x') is defined as

[0035] m(x)=E(f(x))

[0036] k f (x,x')=E[(m(x)-f(x))(m(x')-f(x'))]

[0037] Set the mean function m(x) to zero and the kernel function of the covariance function to the square exponential covariance SE function. The square exponential covariance SE function is

[0038]

[0039] Where S is the dimension of the input variable x, x i s , x j s are the values ​​of the sth input vector, l s Represents the length of each vector in the input variable.

[0040] Further, in step 5, the parameters of the least squares support vector regression (LSSVR) algorithm are optimized using a particle swarm optimization (PSO) algorithm, further including:

[0041] The particle swarm optimization PSO algorithm is used to optimize the regularization parameter C and kernel parameter γ of the least squares support vector regression LSSVR algorithm. The flight speed and position of particle i are adjusted according to the following formula:

[0042]

[0043]

[0044] Where, the subscript i represents the number of particles, d represents the dimension, k represents the number of iterations, and c represents the number of 1 and c 2 are learning factors for adjusting the global extremum and the maximum step length of the local extremum flight direction, respectively. The inertia weight w balances the global search capability and the local search capability. rand() is used to get a random number between 0 and 1.

[0045] Furthermore, in step 5, a model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is constructed, further comprising:

[0046] The model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is:

[0047]

[0048] Among them, α i is the Lagrange multiplier, k(x,x i ) is the kernel function, b is the deviation, and the kernel function is set to the radial kernel function RBF. The radial kernel function RBF is

[0049]

[0050] Here, γ is a constant related to the width of the function.

[0051] Further, in step 6, the maximum absolute error MAX Absolute error, the mean absolute error MAE, and the root mean square error RMSE are calculated, further including:

[0052] The formula for the maximum absolute error MAX Absolute error is

[0053] MAX Absolute error=max|SOH estimation -SOH real |

[0054] Among them, SOH estimation is the estimated SOH value, SOH real is the actual SOH value;

[0055] The formulas for mean absolute error MAE and root mean square error RMSE are:

[0056]

[0057]

[0058] Among them, f(x i ) is the estimated SOH value, y i is the actual SOH value, and N is the number of test samples.

[0059] The beneficial effects of the present invention are:

[0060] 1. The present invention establishes a Thevenin equivalent circuit model and an open circuit voltage U oc The relationship model with the state of charge SOC is based on the voltage and current data collected during the charging process of the lithium-ion battery, and the battery data at different ambient temperatures are obtained to construct a model for estimating the SOH at different ambient temperatures and a model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature. Then, a mapping relationship between the SOH of the lithium-ion battery at different temperatures and the SOH at room temperature is established, and the room temperature SOH value corresponding to the lithium-ion battery at different ambient temperatures is estimated. When the ambient temperature is different, it can accurately determine whether the battery needs to be replaced, thereby improving the effectiveness of predicting the battery health status at different ambient temperatures.

[0061] 2. The present invention calculates the maximum absolute error MAX Absolute error, the mean absolute error MAE, and the root mean square error RMSE based on the estimated SOH values ​​at different ambient temperatures and the actual SOH values ​​at different ambient temperatures, as well as the estimated room temperature SOH values ​​corresponding to different ambient temperatures and the actual room temperature SOH values ​​corresponding to different ambient temperatures. The maximum absolute error MAX Absolute error, the mean absolute error MAE, and the root mean square error RMSE are used to evaluate the performance of the estimated SOH values ​​at different ambient temperatures and the estimated room temperature SOH values ​​corresponding to different ambient temperatures, thereby ensuring the effectiveness and accuracy of the battery health status estimation of the method of the present invention at different ambient temperatures. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 The present invention is a flow chart of a method for estimating the health status of a lithium-ion battery under different ambient temperatures.

[0063] Figure 2 The present invention is a flow chart of an embodiment of a method for estimating the health status of a lithium-ion battery under different ambient temperatures.

[0064] Figure 3 It is a current curve diagram of the lithium-ion battery charging process in an embodiment of the present invention.

[0065] Figure 4It is a terminal voltage curve diagram of the lithium-ion battery during charging in an embodiment of the present invention.

[0066] Figure 5 It is a current diagram under DST condition in an embodiment of the present invention.

[0067] Figure 6 It is a current diagram under US06 working condition in an embodiment of the present invention.

[0068] Figure 7 Schematic diagram of the Thevenin equivalent circuit model in an embodiment of the present invention.

[0069] Figure 8(a)-Figure 8(c) This is a graph of SOH estimation results at the current temperature of the test set using data of battery #8 at 0°C, 24°C, and 30°C in an embodiment of the present invention.

[0070] Figure 9(a)-Figure 9(c) This is a SOH estimation error diagram at the current temperature of the test set using the data of #8 battery at 0°C, 24°C, and 30°C in an embodiment of the present invention.

[0071] Figure 10(a)-Figure 10(g) This is a diagram of the SOH estimation result at the current temperature using battery #7 as the test set in an embodiment of the present invention.

[0072] Figure 11(a)-Figure 11(g) FIG. 4 is a diagram of the SOH estimation error at the current temperature of the test set using battery #7 in an embodiment of the present invention.

[0073] Figure 12(a)-Figure 12(f) This is a diagram of the SOH estimation results at the current temperature using the DST condition as the test set in an embodiment of the present invention.

[0074] Figure 13(a)-Figure 13(f) This is a SOH estimation error diagram at the current temperature using the DST condition as the test set in an embodiment of the present invention.

[0075] Figure 14(a)-Figure 14(e) This is a diagram of the SOH estimation results at the current temperature using the US06 operating condition as the test set in an embodiment of the present invention.

[0076] Figure 15(a)-Figure 15(e) This is a SOH estimation error diagram at the current temperature using the US06 operating condition as the test set in an embodiment of the present invention.

[0077] Figure 16(a)-Figure 16(c) This is a diagram of the room temperature SOH estimation results using the data of #8 battery at 0°C, 24°C, and 30°C as the test set in an embodiment of the present invention.

[0078] Figure 17(a)-Figure 17(c) : This is a room temperature SOH estimation error diagram of the test set using the data of #8 battery at 0°C, 24°C, and 30°C in an embodiment of the present invention.

[0079] Figure 18(a)-Figure 18(g) Graph showing the room temperature SOH estimation results using battery #7 as the test set in an embodiment of the present invention.

[0080] Figure 19(a)-Figure 19(g) Graph showing the room temperature SOH estimation error using battery #7 as the test set in an embodiment of the present invention.

[0081] Figure 20(a)-Figure 20(f) This is a diagram of the room temperature SOH estimation results using the DST condition as the test set in an embodiment of the present invention.

[0082] Figure 21(a)-Figure 21(f) This is a room temperature SOH estimation error diagram using the DST condition as the test set in an embodiment of the present invention.

[0083] Figure 22(a)-Figure 22(f) This is a diagram of the room temperature SOH estimation results using the US06 operating condition as the test set in an embodiment of the present invention.

[0084] Figure 23(a)-Figure 23(f) This is a room temperature SOH estimation error diagram using the US06 operating condition as the test set in an embodiment of the present invention. DETAILED DESCRIPTION

[0085] The technical solution of the present invention will be described clearly and completely below 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. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0086] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are 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, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0087] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" 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 it can be indirectly connected through an intermediate medium, or it can be the internal communication of 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.

[0088] like Figure 1 As shown, the present invention provides a method for estimating the health status of a lithium-ion battery under different ambient temperatures, comprising the following steps:

[0089] Step 1, conduct a charging experiment on the lithium-ion battery at different ambient temperatures, record the current ambient temperature and collect the voltage, current and time during the charging process of the lithium-ion battery, and multiply the current and sampling time during the charging process to obtain the current capacity of the lithium-ion battery;

[0090] Step 2, according to the SOH capacity definition, calculate the SOH of the lithium-ion battery at different ambient temperatures to obtain the actual SOH values ​​at different ambient temperatures, and calculate the ratio of the SOH of the lithium-ion battery at different ambient temperatures to the SOH at room temperature to obtain the actual room temperature SOH values ​​corresponding to the different ambient temperatures;

[0091] Step 3: Establish the Thevenin equivalent circuit model and use polynomial fitting to establish the open circuit voltage U oc The relationship model between the open circuit voltage U and the state of charge SOC is based on the voltage and current data collected during the charging process of the lithium-ion battery, and the recursive least squares method is used to calculate the open circuit voltage U oc The parameters of the relationship model with the state of charge SOC are identified, and the ohmic internal resistance R of the battery at different ambient temperatures is obtained when the state of charge SOC is specified. 0 , Polarization internal resistance R p , voltage value V;

[0092] Step 4: Set the ohmic internal resistance R at a specific state of charge SOC 0 , Polarization internal resistance R p , voltage value V, SOC value, temperature T are normalized, and then the processed ohmic internal resistance R 0 , Polarization internal resistance R p , voltage value V, SOC value, temperature T are the inputs of Gaussian process regression GPR algorithm, SOH at different ambient temperatures is the output of Gaussian process regression GPR algorithm, a model for estimating SOH at different ambient temperatures is constructed, and the model is used to estimate SOH at different ambient temperatures to obtain the estimated SOH values ​​at different ambient temperatures;

[0093] Step 5: Set the ohmic internal resistance R 0 , Polarization internal resistance R p, voltage value V, SOC value, temperature T, and SOH at different ambient temperatures are used as inputs of the least squares support vector regression LSSVR algorithm, and the particle swarm optimization PSO algorithm is used to optimize the parameters of the least squares support vector regression LSSVR algorithm, and a model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is constructed. The model is used to estimate the ratio of SOH at different temperatures to SOH at room temperature of lithium-ion batteries, and the estimated room temperature SOH values ​​corresponding to different ambient temperatures are obtained;

[0094] Step 6. Calculate the maximum absolute error MAX Absolute error, the mean absolute error MAE, and the root mean square error RMSE based on the estimated SOH values ​​at different ambient temperatures and the actual SOH values ​​at different ambient temperatures, as well as the estimated room temperature SOH values ​​corresponding to different ambient temperatures and the actual room temperature SOH values ​​corresponding to different ambient temperatures. The performance of the estimated SOH values ​​at different ambient temperatures and the estimated room temperature SOH values ​​corresponding to different ambient temperatures are evaluated by the maximum absolute error MAXAbsolute error, the mean absolute error MAE, and the root mean square error RMSE.

[0095] Furthermore, step 1, performing charging experiments on the lithium-ion battery at different ambient temperatures, further includes performing charging experiments on the lithium-ion battery at different ambient temperatures using a standard constant current and constant voltage charging method.

[0096] Further, step 2, according to the SOH capacity definition, calculates the SOH of the lithium-ion battery at different ambient temperatures to obtain the actual SOH values ​​at different ambient temperatures, and calculates the ratio of the SOH of the lithium-ion battery at different ambient temperatures to the SOH at room temperature to obtain the actual room temperature SOH values ​​corresponding to the different ambient temperatures, further including,

[0097] According to the definition of SOH capacity, calculate the SOH of lithium-ion batteries at different ambient temperatures.

[0098]

[0099] In the above formula, Q rated Q is the rated capacity of a new battery when it leaves the factory. aged It is the actual maximum available capacity of the battery after it is put into use;

[0100] Calculate the ratio of SOH of lithium-ion batteries at different ambient temperatures to SOH at room temperature.

[0101]

[0102] In the above formula, SOH is the health status value of the battery at different ambient temperatures. r It is the healthy state value at room temperature.

[0103] Furthermore, in step 3, the Thevenin equivalent circuit model is established, and the open circuit voltage U is established by polynomial fitting. oc The relationship model with the state of charge SOC further includes:

[0104] The Thevenin equivalent circuit model is established as

[0105]

[0106] Where U t Indicates terminal voltage, U oc Represents the open circuit voltage of the battery, R 0 represents the ohmic internal resistance, I represents the current, and U P represents the polarization voltage, C P Represents polarized capacitance, R P represents polarization resistance;

[0107] The relationship model between open circuit voltage Uoc and state of charge SOC is established as

[0108] U oc =K 0 +K 1 SOC+K 2 SOC 2 +K 3 SOC 3 +K 4 SOC 4

[0109] In the formula, K 0 , K 1 , K 2 , K 3 , K 4 is the coefficient.

[0110] Further, in step 3, when the state of charge SOC is specified, the ohmic internal resistance R of the battery at different ambient temperatures is obtained. 0 , Polarization internal resistance R p , voltage value V, further including,

[0111] The state of charge at the end of constant current charging is recorded as the maxSOC value. When the specific state of charge SOC is between 20% and the maxSOC value and the value increases by 5%, the ohmic internal resistance R of the battery at different ambient temperatures is obtained. 0 , polarization internal resistance R p , voltage value V.

[0112] Further, in step 4, the ohmic internal resistance R at a specific state of charge SOC is 0 , Polarization internal resistance R p, voltage value V, SOC value, temperature T are normalized, further including:

[0113] The ohmic internal resistance R at a specific state of charge SOC 0 , Polarization internal resistance R p , voltage value V, SOC value, and temperature T are used as characteristic parameters for estimating the SOH model under different ambient temperatures, and the mapminmax function of MATLAB is used to normalize the parameters to [1,5].

[0114] Further, in step 4, the processed ohmic internal resistance R 0 , Polarization internal resistance R p , voltage value V, SOC value, temperature T are inputs of Gaussian process regression GPR algorithm, SOH at different ambient temperatures is the output of Gaussian process regression GPR algorithm, and a model for estimating SOH at different ambient temperatures is constructed, further including:

[0115] The estimated SOH model at different ambient temperatures is:

[0116] f(x)~gp(m(x),k f (x,x′))

[0117] Among them, m(x), k f (x,x') is the mean and covariance function, m(x), k f (x,x') is defined as

[0118] m(x)=E(f(x))

[0119] k f (x,x')=E[(m(x)-f(x))(m(x')-f(x'))]

[0120] Set the mean function m(x) to zero and the kernel function of the covariance function to the square exponential covariance SE function. The square exponential covariance SE function is

[0121]

[0122] Where S is the dimension of the input variable x, x i s , x j s are the values ​​of the sth input vector, l s Represents the length of each vector in the input variable.

[0123] Further, in step 5, the parameters of the least squares support vector regression (LSSVR) algorithm are optimized using a particle swarm optimization (PSO) algorithm, further including:

[0124] The particle swarm optimization PSO algorithm is used to optimize the regularization parameter C and kernel parameter γ of the least squares support vector regression LSSVR algorithm. The flight speed and position of particle i are adjusted according to the following formula:

[0125]

[0126]

[0127] Where, the subscript i represents the number of particles, d represents the dimension, k represents the number of iterations, and c represents the number of 1 and c 2 are learning factors for adjusting the global extremum and the maximum step length of the local extremum flight direction, respectively. The inertia weight w balances the global search capability and the local search capability. rand() is used to get a random number between 0 and 1.

[0128] Furthermore, in step 5, a model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is constructed, further comprising:

[0129] The model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is:

[0130]

[0131] Among them, α i is the Lagrange multiplier, k(x,x i ) is the kernel function, b is the deviation, and the kernel function is set to the radial kernel function RBF. The radial kernel function RBF is

[0132]

[0133] Here, γ is a constant related to the width of the function.

[0134] Further, in step 6, the maximum absolute error MAX Absolute error, the mean absolute error MAE, and the root mean square error RMSE are calculated, further including:

[0135] The formula for the maximum absolute error MAX Absolute error is

[0136] MAX Absolute error=max|SOH estimation -SOH real |

[0137] Among them, SOH estimation is the estimated SOH value, SOH real is the actual SOH value;

[0138] The formulas for mean absolute error MAE and root mean square error RMSE are:

[0139]

[0140]

[0141] Among them, f(x i ) is the estimated SOH value, y i is the actual SOH value, and N is the number of test samples.

[0142] Example

[0143] The flowchart of the embodiment of the method for estimating the health status of a lithium-ion battery when the ambient temperature is different in this example is as follows: Figure 2 As shown, the lithium-ion battery testing equipment is the Xinwei battery testing system CT-4008, the battery used in the test is the cylindrical lithium-ion battery LR18650SZ, and the battery parameters are shown in Table 1.

[0144] Table 1 Battery parameters

[0145]

[0146] The battery was charged and discharged at -10℃, 0℃, 10℃, 20℃, 30℃, 40℃ and room temperature. The specific experimental steps are as follows:

[0147] (1) Use the standard constant current and constant voltage charging method to fully charge the battery and let it sit for one hour;

[0148] (2) Discharge the battery at a constant current of 0.5C, discharging 10% of the battery capacity based on the rated capacity;

[0149] (3) Let the battery stand for one hour, and record the stable terminal voltage value after standing as the discharge open circuit voltage of the battery at that moment;

[0150] (4) Cycle steps (2) and (3) until the battery is discharged to 3.0 V, and the discharge OCV-SOC experiment is completed;

[0151] (5) Perform a constant current charging experiment with a current of 0.5C, and also charge the battery to 10% of its rated capacity;

[0152] (6) Let the battery stand for one hour, and record the stable terminal voltage value after standing as the battery's charging open circuit voltage at that moment;

[0153] (7) Steps (5) and (6) are repeated until the battery is charged to 4.2 V. The charging OCV-SOC experiment is completed.

[0154] Since the battery discharge process depends on the use of the battery system and is difficult to predict, the charging process of the lithium-ion battery is selected for the experiment. The current value and terminal voltage value changes during the charging process are as follows: Figure 3 , Figure 4 shown.

[0155] In addition, the dynamic stress test (DST) condition and US06 condition were selected to simulate the actual working environment of the battery. At 0℃, 10℃, 20℃, 30℃, 40℃ and room temperature, the battery was discharged with these two current conditions. First, the battery was fully charged with constant current and constant voltage, and the current data was imported into the BTS software. The battery was discharged to 3.0V, and the sampling time interval was set to 1s to obtain the battery charge and discharge experimental data at various temperatures. The current under the DST condition is as follows: Figure 5 As shown, the current under US06 condition is as follows Figure 6 shown.

[0156] Establish the Thevenin equivalent circuit model, such as Figure 7 As shown. Using polynomial fitting, a relationship model between the open circuit voltage and state of charge of the #8 battery at room temperature is constructed. Based on the current and voltage data of the constant current charging stage, the recursive least squares method is used to identify the parameters of the model. The state of charge at the end of constant current charging is recorded as the maxSOC value. When the specific state of charge SOC is a period value from 20% to the maxSOC value and the period value increases by 5% (SOC is 20%, 25%, 30%, 35%,..., maxSOC value), the ohmic internal resistance R of the battery at different ambient temperatures is obtained. 0 , polarization internal resistance R p , voltage value V. The characteristic parameters are obtained: ohmic internal resistance R0, polarization internal resistance Rp, voltage value V and corresponding SOC value and temperature T under a specific state of charge SOC.

[0157] Based on the Gaussian process regression algorithm, the data of #8 battery at -10℃, 10℃, 20℃, and 40℃ are used as the training set to train the model, and the data of #8 battery at 0℃, 24℃, and 30℃ are used as the test set. The SOH estimation results at the current temperature are as follows: Figure 8(a)-Figure 8(c) As shown, the SOH estimation error of battery #8 at 0℃, 24℃, and 30℃ is the current temperature of the test set. Figure 9(a)-Figure 9(c) shown.

[0158] Based on the Gaussian process regression algorithm, the model is trained with the data of battery #8 at -10℃, 10℃, 20℃, and 40℃ as the training set, and the SOH estimation results at the current temperature of battery #7 as the test set are as follows: Figure 10(a)-Figure 10(g) As shown, Figure 10(a)-Figure 10(g)They are the SOH estimation result diagrams of battery #7 at -10℃, 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃ respectively. The data is the current temperature of the test set. The SOH estimation error of battery #7 at the current temperature of the test set is as follows Figure 11(a)-Figure 11(g) As shown, Figure 11(a)-Figure 11(g) These are the SOH estimation error diagrams for battery #7 at the current temperature of the test set at -10℃, 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃ respectively.

[0159] Based on the Gaussian process regression algorithm, the data of #8 battery at -10℃, 10℃, 20℃, and 40℃ are used as the training set to train the model. The SOH estimation results at the current temperature of the DST condition as the test set are as follows: Figure 12(a)-Figure 12(f) As shown, Figure 12(a)-Figure 12(f) They are the SOH estimation result diagrams of the DST working condition at 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃, respectively, and the data is the current temperature of the test set. The SOH estimation error of the DST working condition at the current temperature of the test set is as follows Figure 13(a)-Figure 13(f) As shown, Figure 13(a)-Figure 13(f) They are the SOH estimation error diagrams for DST conditions at the current temperature of the test set when the data are at 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃.

[0160] Based on the Gaussian process regression algorithm, the data of #8 battery at 10℃, 10℃, 20℃, and 40℃ are used as the training set to train the model, and the SOH estimation results at the current temperature of the US06 condition as the test set are as follows: Figure 14(a)-Figure 14(e) As shown, Figure 14(a)-Figure 14(e) They are respectively the SOH estimation result diagrams of the US06 working condition at 0℃, 10℃, 20℃, 30℃, and 40℃, where the data is the current temperature of the test set. The SOH estimation error of the US06 working condition at the current temperature of the test set is as follows Figure 15(a)-Figure 15(e) As shown, Figure 15(a)-Figure 15(e) They are the SOH estimation error diagrams for the US06 operating condition at the current temperature of the test set when the data are at 0℃, 10℃, 20℃, 30℃ and 40℃.

[0161] The new feature vector is based on the state of charge SOC, ohmic internal resistance R0, polarization internal resistance Rp, voltage value V, temperature T and SOH at each temperature.

[0162] The least squares support vector regression algorithm based on particle swarm optimization is used to train the model with the data of #8 battery at -10℃, 10℃, 20℃, and 40℃ as the training set, and the data of #8 battery at 0℃, 24℃, and 30℃ as the test set. The room temperature SOH estimation results are as follows: Figure 16(a)-Figure 16(c) As shown, the room temperature SOH estimation error of the #8 battery at 0℃, 24℃, and 30℃ is the test set data. Figure 17(a)-Figure 17(c)shown.

[0163] The least squares support vector regression algorithm based on particle swarm optimization is used to train the model with the data of #8 battery at -10℃, 10℃, 20℃, and 40℃ as the training set, and the room temperature SOH estimation results of #7 battery as the test set are as follows: Figure 18(a)-Figure 18(g) As shown, Figure 18(a)-Figure 18(g) The room temperature SOH estimation result diagrams of the test set are respectively #7 battery at -10℃, 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃. The room temperature SOH estimation error of #7 battery is the test set. Figure 19(a)-Figure 19(g) As shown, Figure 19(a)-Figure 19(g) These are the room temperature SOH estimation error diagrams of battery #7 at -10℃, 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃ temperatures, with the data being the test set.

[0164] The least squares support vector regression algorithm based on particle swarm optimization is used to train the model with the data of #8 battery at -10℃, 10℃, 20℃, and 40℃ as the training set, and the room temperature SOH estimation results of the DST condition as the test set are as follows: Figure 20(a)-Figure 20(f) As shown, Figure 20(a)-Figure 20(f) They are the room temperature SOH estimation result diagrams of the DST working condition at 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃, respectively, and the room temperature SOH estimation error of the DST working condition as the test set is as follows: Figure 21(a)-Figure 21(f) As shown, Figure 21(a)-Figure 21(f) These are the room temperature SOH estimation error diagrams for the DST condition at 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃ temperatures with the data being the test set.

[0165] The least squares support vector regression algorithm based on particle swarm optimization is used to train the model with the data of #8 battery at 10℃, 10℃, 20℃, and 40℃ as the training set, and the room temperature SOH estimation results of US06 working condition as the test set are as follows: Figure 22(a)-Figure 22(f) As shown, Figure 22(a)-Figure 22(f) They are respectively the room temperature SOH estimation result diagrams of the US06 working condition at 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃ data as the test set. The room temperature SOH estimation error of the US06 working condition as the test set is as follows Figure 23(a)-Figure 23(f) As shown, Figure 23(a)-Figure 23(f) They are the room temperature SOH estimation error diagrams of the US06 working condition at 0℃, 10℃, 20℃, 24℃, 30℃, and 40℃ temperatures, with the data being the test set.

[0166] The battery health status prediction method of the present invention is evaluated by calculating the three performance indicators of maximum absolute error MAX Absolute error, mean absolute error MAE, and root mean square error RMSE. The statistical results are shown in Table 2 and Table 3.

[0167] Table 2 SOH estimation results at current temperature

[0168]

[0169]

[0170] Table 3 Room temperature SOH estimation results

[0171]

[0172]

[0173] The above results show that the method proposed by the present invention can effectively estimate the corresponding room temperature SOH value at different temperatures, and the average absolute error does not exceed 1.57%, with high estimation accuracy. It can also accurately estimate the battery SOH value at the current temperature, and the maximum absolute error MAX Absolute error is controlled within 3.88%. It also shows that the method proposed by the present invention is applicable to different working conditions and has universal applicability.

[0174] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for estimating the health status of a lithium-ion battery under different ambient temperatures, characterized in that: The following steps are involved: Step 1, conduct a charging experiment on the lithium-ion battery at different ambient temperatures, record the current ambient temperature and collect the voltage, current and time during the charging process of the lithium-ion battery, and multiply the current and sampling time during the charging process to obtain the current capacity of the lithium-ion battery; Step 2, according to the SOH capacity definition, calculate the SOH of the lithium-ion battery at different ambient temperatures to obtain the actual SOH values ​​at different ambient temperatures, and calculate the ratio of the SOH of the lithium-ion battery at different ambient temperatures to the SOH at room temperature to obtain the actual room temperature SOH values ​​corresponding to the different ambient temperatures; Step 3: Establish the Thevenin equivalent circuit model and use polynomial fitting to establish the open circuit voltage U oc The relationship model between the open circuit voltage U and the state of charge SOC is based on the voltage and current data collected during the charging process of the lithium-ion battery, and the recursive least squares method is used to calculate the open circuit voltage U oc The parameters of the relationship model with the state of charge SOC are identified, and the ohmic internal resistance R0 and polarization internal resistance R of the battery at different ambient temperatures are obtained when the state of charge SOC is specified. p , voltage value V; Step 4: The ohmic internal resistance R0 and the polarization internal resistance R p , voltage value V, SOC value, temperature T are normalized, and then the processed ohmic internal resistance R0 and polarization internal resistance R p , voltage value V, SOC value, temperature T are the inputs of Gaussian process regression GPR algorithm, SOH at different ambient temperatures is the output of Gaussian process regression GPR algorithm, a model for estimating SOH at different ambient temperatures is constructed, and the model is used to estimate SOH at different ambient temperatures to obtain the estimated SOH values ​​at different ambient temperatures; Step 5: Change the ohmic internal resistance R0 and polarization internal resistance R p , voltage value V, SOC value, temperature T, and SOH at different ambient temperatures are used as inputs of the least squares support vector regression LSSVR algorithm, and the particle swarm optimization PSO algorithm is used to optimize the parameters of the least squares support vector regression LSSVR algorithm, and a model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is constructed. The model is used to estimate the ratio of SOH at different temperatures to SOH at room temperature of lithium-ion batteries, and the estimated room temperature SOH values ​​corresponding to different ambient temperatures are obtained; Step 6. Calculate the maximum absolute error MAX Absolute error, the mean absolute error MAE, and the root mean square error RMSE based on the estimated SOH values ​​at different ambient temperatures and the actual SOH values ​​at different ambient temperatures, as well as the estimated room temperature SOH values ​​corresponding to different ambient temperatures and the actual room temperature SOH values ​​corresponding to different ambient temperatures. The performance of the estimated SOH values ​​at different ambient temperatures and the estimated room temperature SOH values ​​corresponding to different ambient temperatures are evaluated by the maximum absolute error MAXAbsolute error, the mean absolute error MAE, and the root mean square error RMSE.

2. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 1, characterized in that: Step 1, performing charging experiments on lithium-ion batteries at different ambient temperatures, further comprising, performing charging experiments on lithium-ion batteries at different ambient temperatures using a standard constant current and constant voltage charging method.

3. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 1, characterized in that: Step 2, according to the SOH capacity definition, calculate the SOH of the lithium-ion battery at different ambient temperatures to obtain the actual SOH values ​​at different ambient temperatures, and calculate the ratio of the SOH of the lithium-ion battery at different ambient temperatures to the SOH at room temperature to obtain the actual room temperature SOH values ​​corresponding to the different ambient temperatures, further comprising, According to the definition of SOH capacity, calculate the SOH of lithium-ion batteries at different ambient temperatures. In the above formula, Q rated Q is the rated capacity of a new battery when it leaves the factory. aged It is the actual maximum available capacity of the battery after it is put into use; Calculate the ratio of SOH of lithium-ion batteries at different ambient temperatures to SOH at room temperature. In the above formula, SOH is the health status value of the battery at different ambient temperatures. r It is the healthy state value at room temperature.

4. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 1, characterized in that: In step 3, the Thevenin equivalent circuit model is established, and the open circuit voltage U is established by polynomial fitting. oc The relationship model with the state of charge SOC further includes: The Thevenin equivalent circuit model is established as Where U t Indicates terminal voltage, U oc represents the open circuit voltage of the battery, R0 represents the ohmic internal resistance, I represents the current, and U P represents the polarization voltage, C P Represents polarized capacitance, R P represents polarization resistance; The relationship model between open circuit voltage Uoc and state of charge SOC is established as U oc =K0+K1SOC+K2SOC 2 +K3SOC 3 +K4SOC 4 In the formula, K0, K1, K2, K3, and K4 are coefficients.

5. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 2, characterized in that: In step 3, when the state of charge SOC is specified, the ohmic internal resistance R0 and polarization internal resistance R0 of the battery at different ambient temperatures are obtained. p , voltage value V, further including, The state of charge at the end of constant current charging is recorded as the maxSOC value. When the specific state of charge SOC is between 20% and the maxSOC value and the value increases by 5%, the ohmic internal resistance R0 and polarization internal resistance R0 of the battery at different ambient temperatures are obtained. p , voltage value V.

6. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 1, characterized in that: In step 4, the ohmic internal resistance R0 and the polarization internal resistance R p , voltage value V, SOC value, temperature T are normalized, further including: The ohmic internal resistance R0 and polarization internal resistance R p , voltage value V, SOC value, and temperature T are used as characteristic parameters for estimating the SOH model under different ambient temperatures, and the mapminmax function of MATLAB is used to normalize the parameters to [1,5].

7. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 1, characterized in that: In step 4, the processed ohmic internal resistance R0 and polarization internal resistance R p , voltage value V, SOC value, temperature T are inputs of Gaussian process regression GPR algorithm, SOH at different ambient temperatures is the output of Gaussian process regression GPR algorithm, and a model for estimating SOH at different ambient temperatures is constructed, further including: The estimated SOH model at different ambient temperatures is: f(x)~gp(m(x),k f (x,x′)) Among them, m(x), k f (x,x') is the mean and covariance function, m(x), k f (x,x') is defined as m(x)=E(f(x)) k f (x,x')=E[(m(x)-f(x))(m(x')-f(x'))] Set the mean function m(x) to zero and the kernel function of the covariance function to the square exponential covariance SE function. The square exponential covariance SE function is Where S is the dimension of the input variable x, x i s , x j s are the values ​​of the sth input vector, l s Represents the length of each vector in the input variable.

8. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 1, characterized in that: In step 5, the parameters of the least squares support vector regression (LSSVR) algorithm are optimized using a particle swarm optimization (PSO) algorithm, further including: The particle swarm optimization PSO algorithm is used to optimize the regularization parameter C and kernel parameter γ of the least squares support vector regression LSSVR algorithm. The flight speed and position of particle i are adjusted according to the following formula: Wherein, the subscript i represents the number of particles, d represents the dimension, k represents the number of iterations, c1 and c2 are learning factors for adjusting the maximum step length of the global extremum and the local extremum flight direction, respectively, the inertia weight w balances the global search capability and the local search capability, and rand() is used to obtain a random number between 0 and 1.

9. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 1, characterized in that: In step 5, a model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is constructed, further comprising: The model for estimating the ratio of SOH at different ambient temperatures to SOH at room temperature is: Among them, α i is the Lagrange multiplier, k(x,x i ) is the kernel function, b is the deviation, and the kernel function is set to the radial kernel function RBF. The radial kernel function RBF is Here, γ is a constant related to the width of the function.

10. The method for estimating the health status of a lithium-ion battery under different ambient temperatures according to claim 1, characterized in that: In step 6, the maximum absolute error MAX Absolute error, mean absolute error MAE, and root mean square error RMSE are calculated, further including: The formula for the maximum absolute error MAX Absolute error is MAX Absolute error=max|SOH estimation -SOH real | Among them, SOH estimation is the estimated SOH value, SOH real is the actual SOH value; The formulas for mean absolute error MAE and root mean square error RMSE are: Among them, f(x i ) is the estimated SOH value, y i is the actual SOH value, and N is the number of test samples.

Images