A battery state of charge adaptive estimation method and device, a vehicle

Abstract

The application provides a battery state of charge adaptive estimation method and device and a vehicle, the method comprising: calculating an open circuit voltage original estimation value through an equivalent circuit model of a to-be-tested battery; when a filtered open circuit voltage estimation value is less than a first preset threshold value, the calculated open circuit voltage original estimation value is valid, and a corresponding state of charge estimation value is queried from an open circuit voltage-state of charge curve, which is recorded as a first state of charge estimation value; and the first state of charge estimation value is used to correct a second state of charge estimation value, thereby improving the estimation accuracy of the state of charge under dynamic driving conditions, preventing the battery from entering a low state of charge interval, and preventing a high state of charge estimation value from causing the battery to have the risk of under-voltage and over-discharge. The application can correct the state of charge estimation error existing in the second state of charge estimation value obtained by the ampere-hour integral method in real time by using the first state of charge estimation value obtained by the multiple forgetting factor recursive least square algorithm.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and in particular to an adaptive estimation method and apparatus for battery state of charge, and a vehicle. Background Technology

[0002] For new energy vehicles such as pure electric vehicles and hybrid electric vehicles, the state of charge (SOC) of the on-board power battery is an important function of the electric vehicle battery management system (BMS). The accuracy of its estimation is directly related to the energy management of the whole vehicle and driving safety.

[0003] Currently, some new energy vehicles may use lithium iron phosphate (LFP) batteries as their onboard power batteries. According to the open circuit voltage-state of charge (OCV-SOC) curve of LFP batteries, the curve is relatively flat and exhibits a double plateau period, resulting in limited opportunities for SOC correction and posing a challenge to the accuracy of SOC estimation. Furthermore, for LFP batteries, commonly used SOC estimation methods in vehicle BMS include the ampere-hour integration method and the open circuit voltage method. However, the ampere-hour integration method has the problem that the accuracy of SOC estimation is strictly dependent on the accuracy of the initial SOC and the current sensor. If there is a large error in the initial SOC or the current measurement is inaccurate, the error will accumulate during the calculation process, leading to an increasingly larger error in the final SOC estimation. The open circuit voltage method, on the other hand, requires the onboard power battery to be stationary for a long time to reach an equilibrium state. This method is only suitable for parked vehicles and is not suitable for online real-time estimation. Therefore, how to accurately estimate the SOC of onboard power batteries is a problem that urgently needs to be solved. Summary of the Invention

[0004] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide a battery state of charge adaptive estimation method and apparatus, and a vehicle, to solve the technical problems existing in the prior art.

[0005] To achieve the above and other related objectives, the present invention provides an adaptive estimation method for battery state of charge, comprising the following steps: obtaining an equivalent circuit model generated in advance or in real time based on the battery under test, and determining the battery current and battery terminal voltage corresponding to the battery under test;

[0006] The equivalent circuit model is calculated based on the battery current and the battery terminal voltage to obtain the corresponding parameter identification results;

[0007] The validity of the parameter identification results is identified, and when the parameter identification results are valid, the original estimate of the open circuit voltage in the parameter identification results is filtered to obtain the filtered estimate of the open circuit voltage.

[0008] From the pre-calibrated or real-time open-circuit voltage-state-of-charge curves, retrieve the state-of-charge estimate corresponding to the filtered open-circuit voltage estimate, and denote it as the first state-of-charge estimate; and,

[0009] When the first state of charge estimate is less than the first preset threshold, the first state of charge estimate is used to correct the pre-generated second state of charge estimate in order to estimate the state of charge of the battery under test in real time.

[0010] In one embodiment of the present invention, the first state of charge estimate and the second state of charge estimate are generated in different ways;

[0011] Alternatively, the first state of charge estimate and the second state of charge estimate can be generated in the same way.

[0012] In one embodiment of the present invention, the first state of charge estimate is generated by a recursive least squares method with multiple forgetting factors, and the second state of charge estimate is generated by an ampere-hour integral method.

[0013] In one embodiment of the present invention, the process of calculating the parameters to be identified by using the recursive least squares method with multiple forgetting factors to calculate the equivalent circuit model based on the battery current and the battery terminal voltage includes:

[0014]

[0015] In the formula, U t (k) represents the battery terminal voltage at the current moment; I L (k) represents the battery current at the current moment; τ p R represents the time constant of the RC element in the equivalent circuit model; s R represents the ohmic internal resistance in the equivalent circuit model. p U represents the polarization resistance in the equivalent circuit model; oc (k) represents the open-circuit voltage at the current moment; where τ p R s R p U oc (k) represents the system state variables in the equivalent circuit model, denoted as the parameters to be identified.

[0016] In one embodiment of the present invention, when calculating the equivalent circuit model using the recursive least squares method with multiple forgetting factors, the method further includes:

[0017] The formulas for calculating the parameters to be identified are vectorized and adjusted, and the results of the vectorization adjustment are used to calculate the parameters to be identified.

[0018] U t (k)=Φ T (k)·θ(k);

[0019]

[0020] θ(k)=[R s +R p R s *τ p τ p U oc (k)] T .

[0021] In one embodiment of the present invention, the process of calculating the equivalent circuit model further includes:

[0022] The state of charge (SOC) and parameter vectors of the battery under test are initialized, and the data covariance matrix P(0) = [P1(0) P2(0) P3(0) P4(0)] is determined. T ;

[0023] Based on the vectorization adjustment results and the data covariance matrix, calculate the update gain L for each single-parameter vector component. i (k), we have:

[0024]

[0025] The update gain L is obtained through each single-parameter vector component. i (k) Covariance P for each single-parameter vector component i (k) is updated, and:

[0026]

[0027] Based on the covariance P of each single-parameter vector component i (k) Calculate the update gain L(k) for all parameter vectors, then:

[0028]

[0029] The updated parameter values ​​are calculated based on the update gain L(k) of all parameter vectors, yielding the parameter identification results:

[0030] θ(k)=θ(k-1)+L(k)*[U t (k)-Φ T (k)·θ(k-1)].

[0031] In one embodiment of the present invention, the process of validating the parameter identification result includes:

[0032] Determine whether the values ​​of the parameters to be identified in the parameter identification results are respectively within the corresponding preset range;

[0033] In addition, the difference between the measured battery voltage and the battery terminal voltage in the parameter identification result is calculated, and it is determined whether the difference is less than a second preset threshold.

[0034] In addition, it determines whether the data covariance is less than a third preset threshold;

[0035] In addition, it is determined whether the fluctuation change value of the original estimated value of the open circuit voltage in the parameter identification result is less than the fourth preset threshold.

[0036] If the value of the parameter to be identified is within the corresponding preset range, the difference is less than the second preset threshold, the data covariance is less than the third preset threshold, and the fluctuation change value of the original estimated value of the open circuit voltage is less than the fourth preset threshold, then the parameter identification result is determined to be valid; otherwise, the parameter identification result is determined to be invalid.

[0037] In one embodiment of the present invention, if the battery under test is installed in a target vehicle, the method further includes the following steps before determining the corresponding battery current and battery terminal voltage:

[0038] When the temperature at the current moment exceeds the preset temperature, the frequency and magnitude of changes in battery terminal voltage and battery current are used to determine whether the target vehicle is in dynamic driving conditions at the current moment.

[0039] If the target vehicle is in dynamic driving conditions at the current moment, the battery current and battery terminal voltage are collected.

[0040] If the target vehicle is not in dynamic driving condition at the current moment, then a dynamic driving condition comparison is performed again based on the frequency and magnitude of changes in battery terminal voltage and battery current.

[0041] The present invention also provides a battery state of charge adaptive estimation device, the device comprising:

[0042] The equivalent circuit module acquires the equivalent circuit model generated in advance or in real time based on the battery under test, and determines the battery current and battery terminal voltage corresponding to the battery under test.

[0043] The calculation module is used to calculate the equivalent circuit model based on the battery current and the battery terminal voltage to obtain the corresponding parameter identification results;

[0044] The open-circuit voltage estimation module is used to identify the validity of the parameter identification result, and when the parameter identification result is valid, to filter the original open-circuit voltage estimate in the parameter identification result to obtain the filtered open-circuit voltage estimate.

[0045] The state of charge estimation module is used to query the filtered open-circuit voltage estimate from a pre-calibrated or real-time calibrated open-circuit voltage-state of charge curve, and record it as the first state of charge estimate; and when the first state of charge estimate is less than a first preset threshold, the first state of charge estimate is used to correct the pre-generated second state of charge estimate, so as to estimate the state of charge of the battery under test in real time.

[0046] The present invention also provides a vehicle equipped with a battery state of charge adaptive estimation device as described above.

[0047] As described above, the present invention provides a battery state-of-charge adaptive estimation method and apparatus, and a vehicle, having the following...

[0048] Beneficial effects:

[0049] This invention first obtains an equivalent circuit model generated in advance or in real time based on the battery under test, and determines the battery current and battery terminal voltage corresponding to the battery under test. Then, based on the battery current and battery terminal voltage, the original estimate of the open-circuit voltage is calculated through the equivalent circuit model of the battery under test. Then, the original estimate of the open-circuit voltage is filtered. When the filtered estimate of the open-circuit voltage is less than a first preset threshold, the open-circuit voltage of the battery under test is considered to be below the inflection point of the open-circuit voltage-state of charge curve and located in the slope segment of the open-circuit voltage-state of charge curve. At this time, the calculated original estimate of the open-circuit voltage is considered to be valid. Then, based on this, the corresponding state of charge estimate is retrieved from the curve using the correspondence of the open-circuit voltage-state of charge curve, and recorded as the first state of charge estimate. The first state of charge estimate is then used to correct the second state of charge estimate, thereby improving the estimation accuracy of the state of charge under dynamic driving conditions, preventing the battery from entering the low state of charge range and resulting in an overestimated state of charge estimate, which could lead to the risk of undervoltage and over-discharge of the battery. If the first state of charge estimate is obtained by the recursive least squares algorithm with multiple forgetting factors, and the second state of charge estimate is obtained by the ampere-hour integration method, then this invention can use the state of charge estimate obtained by the recursive least squares algorithm with multiple forgetting factors to correct the state of charge estimation error that exists when the state of charge estimate is obtained by the ampere-hour integration method in real time.

[0050] Furthermore, when the open-circuit voltage of the battery under test is below the inflection point of the open-circuit voltage-state-of-charge curve, the present invention can simultaneously activate the state-of-charge estimation method based on the equivalent circuit model, thereby avoiding the accidental activation of the state-of-charge estimation method based on the equivalent circuit model during the voltage plateau segment, which could lead to incorrect correction of the state of charge. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating an adaptive estimation method for battery state of charge provided in an embodiment of the present invention.

[0052] Figure 2 A circuit connection diagram of a first-order RC equivalent circuit model provided in an embodiment of the present invention;

[0053] Figure 3 This is a schematic diagram of the open-circuit voltage-state-of-charge curve provided in one embodiment of the present invention;

[0054] Figure 4 This is a flowchart illustrating an adaptive estimation method for battery state of charge provided in another embodiment of the present invention.

[0055] Figure 5 This is a schematic diagram of the hardware structure of a battery state-of-charge adaptive estimation device provided in one embodiment of the present invention. Detailed Implementation

[0056] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0057] It should be noted that the illustrations provided in this embodiment are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0058] Figure 1 A schematic flowchart of an adaptive battery state-of-charge estimation method according to an embodiment of the present invention is shown. Specifically, in an exemplary embodiment, as follows... Figure 1 As shown in the figure, this embodiment provides an adaptive estimation method for battery state of charge, which includes the following steps:

[0059] S110: Obtain the equivalent circuit model generated in advance or in real time based on the battery under test, and determine the battery current and battery terminal voltage corresponding to the battery under test. As an example, in this embodiment or other embodiments, the battery under test includes, but is not limited to, lithium iron phosphate batteries.

[0060] S120: Based on the battery current and battery terminal voltage, the equivalent circuit model is calculated to obtain the corresponding parameter identification results. As an example, in this embodiment or other embodiments, when calculating the equivalent circuit model, the equivalent circuit model can be calculated online recursively or iteratively based on the Multiple Forgetting Factors Recursive Least Squares (MFF-RLS) algorithm to obtain the parameter identification results corresponding to the equivalent circuit model.

[0061] S130: The validity of the parameter identification result is verified. If the parameter identification result is valid, the original estimate of the open-circuit voltage in the parameter identification result is filtered to obtain the filtered estimate of the open-circuit voltage. As an example, the filtering method in this embodiment includes, but is not limited to, low-pass filtering.

[0062] S140: From the pre-calibrated or real-time open-circuit voltage-state-of-charge curve, retrieve the state-of-charge estimate corresponding to the filtered open-circuit voltage estimate, denoted as the first state-of-charge estimate; and, when the first state-of-charge estimate is less than a first preset threshold, use the first state-of-charge estimate to correct the pre-generated second state-of-charge estimate, so as to estimate the state of charge of the battery under test in real time. In this embodiment, the first preset threshold can be set numerically according to the actual application scenario, and this embodiment does not limit the specific application scenario or specific value. When the first state-of-charge estimate is less than the first preset threshold, it is considered that the battery under test is in the state-of-charge interval below the inflection point of the open-circuit voltage-state-of-charge curve, which can be used as a prerequisite activation condition for the SOC estimation method based on the equivalent circuit model. If the open-circuit voltage-state-of-charge (OCC) curve is divided into a slope segment and a plateau segment based on the inflection point of the open-circuit voltage-OCC curve, then the battery under test can be considered to be in the slope segment of the open-circuit voltage-OCC curve. This avoids the erroneous activation of the SOC estimation method based on the equivalent circuit model when the battery under test is in the "plateau segment," which could lead to incorrect SOC correction. Furthermore, in this embodiment or other embodiments, the first and second OCC estimates are generated in different ways; or, they are generated in the same way. As an example, in this embodiment, the first OCC estimate is generated using the multi-forgetting-factor recursive least squares method, and the second OCC estimate is generated using the ampere-hour integration method. Therefore, this embodiment uses the OCC estimate obtained by the MFF-RLS algorithm to correct the OCC estimation error of the ampere-hour integration method in real time.

[0063] Therefore, this embodiment first obtains the equivalent circuit model generated in advance or in real time based on the battery under test, and determines the battery current and battery terminal voltage corresponding to the battery under test; then, based on the battery current and battery terminal voltage, the original estimate of the open circuit voltage is calculated through the equivalent circuit model of the battery under test; then, the original estimate of the open circuit voltage is filtered; when the filtered estimate of the open circuit voltage is less than the first preset threshold, it is considered that the open circuit voltage of the battery under test is below the inflection point of the open circuit voltage-state of charge curve and is located in the slope segment of the open circuit voltage-state of charge curve. At this time, the calculated original estimate of the open circuit voltage is considered valid. Then, based on this, the corresponding relationship of the open circuit voltage-state of charge curve is used to find the state of charge estimate corresponding to the original estimate of the open circuit voltage from the curve, which is recorded as the first state of charge estimate. The first state of charge estimate is then used to correct the second state of charge estimate, thereby improving the estimation accuracy of the state of charge under dynamic driving conditions, preventing the battery from entering the low state of charge range and having an excessively high state of charge estimate, which could lead to the risk of undervoltage and over-discharge of the battery. Furthermore, when the open-circuit voltage of the battery under test is below the inflection point of the open-circuit voltage-state-of-charge curve, this embodiment can simultaneously activate the state-of-charge estimation method based on the equivalent circuit model. This avoids mistakenly activating the state-of-charge estimation method based on the equivalent circuit model during the voltage plateau segment, which could lead to incorrect state-of-charge correction. In addition, if the first state-of-charge estimate is obtained using a multi-forgetting-factor recursive least squares algorithm, and the second state-of-charge estimate is obtained using the ampere-hour integration method, this embodiment can use the state-of-charge estimate obtained by the multi-forgetting-factor recursive least squares algorithm to correct the state-of-charge estimation error existing when using the ampere-hour integration method in real time.

[0064] According to the above description, in an exemplary embodiment, the process of calculating the parameters to be identified by using the recursive least squares method with multiple forgetting factors to calculate the equivalent circuit model based on the battery current and battery terminal voltage includes:

[0065]

[0066] In the formula, U t (k) represents the battery terminal voltage at the current moment; I L (k) represents the battery current at the current moment; τ p R represents the time constant of the RC element in the equivalent circuit model; s R represents the ohmic internal resistance in the equivalent circuit model. p U represents the polarization resistance in the equivalent circuit model; oc (k) represents the open-circuit voltage at the current moment; where τ p R s R p U oc(k) represents the system state variables in the equivalent circuit model, denoted as the parameters to be identified. Therefore, this embodiment uses the collected battery current and battery terminal voltage, employs the MFF-RLS algorithm to recursively derive the current open-circuit voltage (OCV) of the battery online, and obtains an estimated value of the current SOC by looking up a table online based on the offline calibrated OCV-SOC curve. This can, to some extent, solve the problem of accumulated SOC error in the ampere-hour integration method.

[0067] According to the above description, in an exemplary embodiment, when calculating the equivalent circuit model using the recursive least squares method with multiple forgetting factors, this embodiment may further include: vectorizing the calculation formula of the parameters to be identified, that is, adjusting the calculation formula... After vectorization adjustment, we have:

[0068] U t (k)=Φ T (k)·θ(k);

[0069]

[0070] θ(k)=[R s +R p R s *τ p τ p U oc (k)] T .

[0071] According to the above description, in an exemplary embodiment, the process of calculating the equivalent circuit model based on the MFF-RLS algorithm further includes:

[0072] The state of charge (SOC) and parameter vectors of the battery under test are initialized, and the data covariance matrix P(0) = [P1(0) P2(0) P3(0) P4(0)] is determined. T ;

[0073] Based on the vectorization adjustment results and the data covariance matrix, calculate the update gain L for each single-parameter vector component. i (k), we have:

[0074]

[0075] The update gain L is obtained through each single-parameter vector component. i (k) Covariance P for each single-parameter vector component i (k) is updated, and:

[0076]

[0077] Based on the covariance P of each single-parameter vector component i(k) Calculate the update gain L(k) for all parameter vectors, then:

[0078]

[0079] The updated parameter values ​​are calculated based on the update gain L(k) of all parameter vectors, yielding the parameter identification results:

[0080] θ(k)=θ(k-1)+L(k)*[U t (k)-Φ T (k)·θ(k-1)].

[0081] According to the above description, in an exemplary embodiment, the process of validating the parameter identification result includes: determining whether the value of the parameter to be identified in the parameter identification result is within its corresponding preset range, that is, determining whether the time constant of the RC element in the equivalent circuit model is within its corresponding preset value range, whether the ohmic internal resistance in the equivalent circuit model is within its corresponding preset value range, whether the polarization internal resistance in the equivalent circuit model is within its corresponding preset value range, and whether the open-circuit voltage at the current moment is within its corresponding preset value range; calculating the difference between the measured battery voltage and the battery terminal voltage in the parameter identification result, and determining whether the difference is less than a second preset threshold; determining whether the data covariance is less than a third preset threshold; and determining whether the parameter... The parameter identification result is determined to be valid if the fluctuation change of the original estimated value of the open-circuit voltage in the parameter identification result is less than the fourth preset threshold. If the parameter value to be identified is within the corresponding preset range, that is, the time constant of the RC element in the equivalent circuit model is within its corresponding preset value range, the ohmic internal resistance in the equivalent circuit model is within its corresponding preset value range, the polarization internal resistance in the equivalent circuit model is within its corresponding preset value range, the open-circuit voltage at the current moment is within its corresponding preset value range, and the difference between the measured battery voltage and the battery terminal voltage in the parameter identification result is less than the second preset threshold, the data covariance is less than the third preset threshold, and the fluctuation change of the original estimated value of the open-circuit voltage is less than the fourth preset threshold, then the parameter identification result is determined to be valid; otherwise, the parameter identification result is determined to be invalid.

[0082] According to the above description, in an exemplary embodiment, if the battery under test is installed in the target vehicle, before determining the corresponding battery current and battery terminal voltage, this embodiment may further include: when the temperature at the current moment exceeds a preset temperature, determining whether the target vehicle is in dynamic driving condition at the current moment based on the frequency of change of the terminal voltage of the battery under test, the amplitude of change of the terminal voltage, the frequency of change of the current, and the amplitude of change of the current; if the target vehicle is in dynamic driving condition at the current moment, collecting the battery current and battery terminal voltage; if the target vehicle is not in dynamic driving condition at the current moment, re-comparing the dynamic driving condition based on the frequency and amplitude of change of the battery terminal voltage and the frequency and amplitude of change of the battery current. Specifically, in this embodiment, for the battery under test, the frequency of terminal voltage change can be compared with a preset frequency of terminal voltage change, the amplitude of terminal voltage change can be compared with a preset amplitude of terminal voltage change, the frequency of current change can be compared with a preset frequency of current change, and the amplitude of current change can be compared with a preset amplitude of current change. When the frequency of terminal voltage change, the amplitude of terminal voltage change, the frequency of current change, and the amplitude of current change of the battery under test are within their respective preset ranges, the target vehicle is considered to be in dynamic driving conditions. In this embodiment or other embodiments, the target vehicle includes, but is not limited to, new energy vehicles such as pure electric vehicles and hybrid electric vehicles.

[0083] In another exemplary embodiment of the present invention, such as Figures 2 to 4 As shown, this embodiment provides an adaptive estimation method for battery state of charge, including the following steps:

[0084] Step 1: When the current temperature exceeds the preset temperature, determine whether the target vehicle is in dynamic driving condition at the current moment based on the frequency and amplitude of changes in battery terminal voltage and battery current. If the target vehicle is in dynamic driving condition at the current moment, collect battery current and battery terminal voltage. If the target vehicle is not in dynamic driving condition at the current moment, re-compare the dynamic driving condition based on the frequency and amplitude of changes in battery terminal voltage and battery current.

[0085] Step 2: Using the collected battery terminal voltage and battery current, the original OCV estimate of the equivalent circuit model is obtained iteratively through the Multiple Forgetting Factors Recursive Least Squares (MFF-RLS) algorithm. In this embodiment or other embodiments, the equivalent circuit model is generated based on the battery under test, which includes, but is not limited to, lithium iron phosphate batteries. The equivalent circuit model in this embodiment includes, but is not limited to, batteries such as... Figure 2The first-order RC equivalent circuit model is shown. The iterative formula for the MFF-RLS algorithm based on this model is as follows:

[0086]

[0087] In the formula, U t (k) represents the battery terminal voltage at the current moment; I L (k) represents the battery current at the current moment; τ p R represents the time constant of the RC element in the equivalent circuit model; s R represents the ohmic internal resistance in the equivalent circuit model. p U represents the polarization resistance in the equivalent circuit model; oc (k) represents the open-circuit voltage at the current moment; where τ p R s R p U oc (k) represents the system state variables in the equivalent circuit model, denoted as the parameters to be identified.

[0088] Step 3, then... Adjusting to vector form, we have:

[0089] U t (k)=Φ T (k)·θ(k);

[0090]

[0091] θ(k)=[R s +R p R s *τ p τ p U oc (k)] T .

[0092] Step 4: Based on the vectorized adjustment results, the MFF-RLS algorithm is used to calculate the equivalent circuit model, obtaining the real-time identification results of each parameter, which are denoted as the MFF-RLS algorithm calculation results. The calculation process of the MFF-RLS algorithm includes:

[0093] (1) Parameter initialization; including initializing the state of charge of the battery under test and the parameter vector to be identified, and determining the data covariance matrix P(0) = [P1(0) P2(0) P3(0) P4(0)] T ;

[0094] (2) Calculate the update gain L for each single-parameter vector component. i (k), we have:

[0095]

[0096] (3) Covariance P for each single-parameter vector component i (k) is updated, and:

[0097]

[0098] (4) Calculate the update gain L(k) for all parameter vectors, then:

[0099]

[0100] (5) Calculate the updated values ​​of the parameters to be identified:

[0101] θ(k)=θ(k-1)+L(k)*[U t (k)-Φ T (k)·θ(k-1)].

[0102] Step 5: Determine whether the MFF-RLS algorithm calculation result is valid at the current moment. The corresponding judgment criteria include: the identified equivalent circuit model parameter τ. p R s R p U oc (k) Whether it is within its respective preset value range, whether the error between the estimated battery terminal voltage and the measured battery voltage is less than the second preset threshold, whether the data covariance is less than the third preset threshold, and whether the fluctuation value of the identified open-circuit voltage OCV is less than the fourth preset threshold. If parameter τ p Within its corresponding preset value range, parameter R s Within its corresponding preset value range, parameter R p Within its corresponding preset value range, parameter U oc (k) If the error between the estimated battery terminal voltage and the measured battery voltage is less than the second preset threshold, the data covariance is less than the third preset threshold, and the fluctuation of the identified open-circuit voltage (OCV) is less than the fourth preset threshold, then the MFF-RLS algorithm calculation result is determined to be valid; otherwise, the MFF-RLS algorithm calculation result is determined to be invalid. The second, third, and fourth preset thresholds can be set numerically according to the actual application scenario, so this embodiment does not impose specific application scenarios or numerical limitations. τ p R s R p U oc The preset value range of parameters such as (k) can be set according to the actual application scenario. This embodiment does not limit the application scenario or specific values. For example, τ p R sR p U oc The preset value range of parameters such as (k) can be an empirical value range determined based on expert experience.

[0103] Step 6: When the MFF-RLS algorithm calculation result is valid, the last term θ4(k) in the identified equivalent circuit model parameters is used as the estimated value of the open-circuit voltage OCV at the current moment. Meanwhile, to maintain the stability of the parameters to be identified, this embodiment can perform low-pass filtering on the OCV estimated value through a first-order low-pass filter to obtain the filtered OCV estimated value.

[0104] Step 7: Based on the filtered OCV estimate, query the corresponding SOC estimate from the pre-determined or real-time determined OCV-SOC curve. When the queried SOC estimate is less than the first preset threshold, it is considered that the battery under test is in the SOC range below the inflection point of the OCV-SOC curve at the current moment.

[0105] Step 8: When the battery under test is in the SOC range below the inflection point of the OCV-SOC curve, the SOC estimate obtained by the ampere-hour integration method is corrected using the SOC estimate obtained by the MFF-RLS algorithm.

[0106] Repeat steps 1 through 8 above to achieve real-time SOC estimation for the battery under test.

[0107] Therefore, this embodiment uses the collected battery current and battery terminal voltage, and employs the multi-forgetting factor recursive least squares (MFF-RLS) algorithm to online recursively obtain the original OCV estimate of the battery under test at the current moment. This original OCV estimate is then low-pass filtered, and the SOC estimate at the current moment is obtained online by looking up a table based on the offline calibrated OCV-SOC curve. This can, to some extent, solve the problem of SOC accumulation error in the ampere-hour integration method under dynamic driving conditions. Furthermore, as the principle of the MFF-RLS algorithm shows, the SOC estimation accuracy based on the equivalent circuit model is significantly affected by the slope of the OCV-SOC curve. For lithium iron phosphate batteries, their OCV-SOC curve has a voltage "plateau segment" and a "slope segment," such as... Figure 3As shown, the slopes of the curves change significantly on both sides of the inflection point. In the voltage "plateau segment," the slope of the OCV-SOC curve is relatively small. This means that a small OCV estimation error can lead to a large SOC estimation error. Therefore, for lithium iron phosphate batteries, an online recursive OCV SOC estimation method can be used, applicable only to the slope segment of the OCV-SOC curve. Thus, this embodiment uses the MFF-RLS algorithm to iteratively calculate the original OCV estimate online, and then performs a low-pass filter on this original OCV estimate. It then determines in real-time whether the original OCV estimate is below a preset OCV threshold. When the filtered OCV estimate is less than this threshold, the current battery OCV is considered to be below the inflection point of the OCV-SOC curve (i.e., within the slope segment of the OCV-SOC curve), and the SOC estimation method based on the equivalent circuit model is activated to avoid mistakenly activating the equivalent circuit model-based SOC estimation method in the voltage "plateau segment."

[0108] In summary, this invention provides an adaptive estimation method for the state of charge (SOC) of a battery. First, the original estimated value of the open-circuit voltage is calculated using the equivalent circuit model of the battery under test. Then, this original estimated value is filtered. When the filtered estimated value is less than a first preset threshold, the open-circuit voltage of the battery under test is considered to be below the inflection point of the open-circuit voltage-SOC curve and located in the slope segment of the curve. At this point, the calculated original estimated value of the open-circuit voltage is considered valid. Based on this, the corresponding SOC estimate is retrieved from the curve using the correspondence between the open-circuit voltage and SOC curves, and recorded as the first SOC estimate. The first SOC estimate is then used to correct the second SOC estimate, thereby improving the accuracy of SOC estimation under dynamic driving conditions. This prevents the battery from entering a low SOC range, resulting in an excessively high SOC estimate, which could lead to undervoltage and over-discharge risks. If the first state of charge (SOC) estimate is obtained using the MFF-RLS algorithm and the second SOC estimate is obtained using the ampere-hour integration method, this invention can use the SOC estimate obtained by the MFF-RLS algorithm to correct the SOC estimation error of the ampere-hour integration method in real time. Simultaneously, this method iteratively calculates the original OCV estimate using the MFF-RLS algorithm and performs low-pass filtering. When the filtered OCV estimate is less than a preset OCV threshold, the current battery OCV is considered to be below the inflection point of the OCV-SOC curve (i.e., within the slope segment of the OCV-SOC curve). At this point, the OCV calculated by the MFF-RLS algorithm is considered valid. Based on this, the correspondence between OCV and SOC is used to estimate the current SOC of the battery online to correct the SOC estimation error of the ampere-hour integration method in real time, thereby improving the SOC estimation accuracy under dynamic driving conditions. This prevents the battery from entering a low SOC range and resulting in an excessively high SOC estimate, which could lead to undervoltage and over-discharge risks. Furthermore, when the open-circuit voltage of the battery under test is below the inflection point of the open-circuit voltage-state-of-charge (OCC) curve, this invention can simultaneously activate the OCC estimation method based on the equivalent circuit model. This avoids mistakenly activating the OCC estimation method based on the equivalent circuit model during voltage plateau periods, which could lead to incorrect OCC corrections. If the battery under test is a lithium iron phosphate (LFP) battery, this method can effectively identify the inflection point of the OCC-SOC curve under dynamic driving conditions, serving as a prerequisite for activating the model-based SOC estimation method. Simultaneously, it avoids mistakenly activating the model-based SOC estimation method when the LFP battery is in a "plateau" period, which could also lead to incorrect SOC corrections.

[0109] In another exemplary embodiment of the present invention, such as Figure 5 As shown, this embodiment also provides a battery state of charge adaptive estimation device, including:

[0110] The equivalent circuit module 510 acquires a pre-generated or real-time equivalent circuit model based on the battery under test, and determines the battery current and battery terminal voltage corresponding to the battery under test. As an example, in this embodiment or other embodiments, the battery under test includes, but is not limited to, lithium iron phosphate batteries.

[0111] The calculation module 520 is used to calculate the equivalent circuit model based on the battery current and battery terminal voltage to obtain the corresponding parameter identification results. As an example, in this embodiment or other embodiments, when calculating the equivalent circuit model, the equivalent circuit model can be calculated online recursively or iteratively based on the multi-forgetting-factor recursive least squares algorithm to obtain the parameter identification results corresponding to the equivalent circuit model.

[0112] The open-circuit voltage estimation module 530 is used to identify the validity of the parameter identification results, and when the parameter identification results are valid, to filter the original open-circuit voltage estimate in the parameter identification results to obtain a filtered open-circuit voltage estimate. As an example, the filtering method in this embodiment includes, but is not limited to, low-pass filtering.

[0113] The state of charge (SOC) estimation module 540 is used to retrieve the SOC estimate corresponding to the filtered open-circuit voltage estimate from a pre-calibrated or real-time calibrated open-circuit voltage-SOC curve, denoted as the first SOC estimate; and, when the first SOC estimate is less than a first preset threshold, to correct a pre-generated second SOC estimate using the first SOC estimate, so as to perform real-time estimation of the SOC of the battery under test. In this embodiment, the first preset threshold can be set numerically according to the actual application scenario, and this embodiment does not limit the specific application scenario or specific value. When the first SOC estimate is less than the first preset threshold, the battery under test is considered to be in the SOC interval below the inflection point of the open-circuit voltage-SOC curve, which can be used as a prerequisite activation condition for the SOC estimation method based on the equivalent circuit model. If the open-circuit voltage-state-of-charge (OCC) curve is divided into a slope segment and a plateau segment based on the inflection point of the open-circuit voltage-OCC curve, then the battery under test can be considered to be in the slope segment of the open-circuit voltage-OCC curve. This avoids the erroneous activation of the SOC estimation method based on the equivalent circuit model when the battery under test is in the "plateau segment," which could lead to incorrect SOC correction. Furthermore, in this embodiment or other embodiments, the first and second OCC estimates are generated in different ways; or, they are generated in the same way. As an example, in this embodiment, the first OCC estimate is generated using the multi-forgetting-factor recursive least squares method, and the second OCC estimate is generated using the ampere-hour integration method. Therefore, this embodiment uses the OCC estimate obtained by the MFF-RLS algorithm to correct the OCC estimation error of the ampere-hour integration method in real time.

[0114] Therefore, this embodiment first obtains the equivalent circuit model generated in advance or in real time based on the battery under test, and determines the battery current and battery terminal voltage corresponding to the battery under test; then, based on the battery current and battery terminal voltage, the original estimate of the open circuit voltage is calculated through the equivalent circuit model of the battery under test; then, the original estimate of the open circuit voltage is filtered; when the filtered estimate of the open circuit voltage is less than the first preset threshold, it is considered that the open circuit voltage of the battery under test is below the inflection point of the open circuit voltage-state of charge curve and is located in the slope segment of the open circuit voltage-state of charge curve. At this time, the calculated original estimate of the open circuit voltage is considered valid. Then, based on this, the corresponding relationship of the open circuit voltage-state of charge curve is used to find the state of charge estimate corresponding to the original estimate of the open circuit voltage from the curve, which is recorded as the first state of charge estimate. The first state of charge estimate is then used to correct the second state of charge estimate, thereby improving the estimation accuracy of the state of charge under dynamic driving conditions, preventing the battery from entering the low state of charge range and having an excessively high state of charge estimate, which could lead to the risk of undervoltage and over-discharge of the battery. Furthermore, when the open-circuit voltage of the battery under test is below the inflection point of the open-circuit voltage-state-of-charge curve, this embodiment can simultaneously activate the state-of-charge estimation method based on the equivalent circuit model. This avoids mistakenly activating the state-of-charge estimation method based on the equivalent circuit model during the voltage plateau segment, which could lead to incorrect state-of-charge correction. In addition, if the first state-of-charge estimate is obtained using a multi-forgetting-factor recursive least squares algorithm, and the second state-of-charge estimate is obtained using the ampere-hour integration method, this embodiment can use the state-of-charge estimate obtained by the multi-forgetting-factor recursive least squares algorithm to correct the state-of-charge estimation error existing when using the ampere-hour integration method in real time.

[0115] According to the above description, in an exemplary embodiment, the process of calculating the parameters to be identified by using the recursive least squares method with multiple forgetting factors to calculate the equivalent circuit model based on the battery current and battery terminal voltage includes:

[0116]

[0117] In the formula, U t (k) represents the battery terminal voltage at the current moment; I L (k) represents the battery current at the current moment; τ p R represents the time constant of the RC element in the equivalent circuit model; s R represents the ohmic internal resistance in the equivalent circuit model. p U represents the polarization resistance in the equivalent circuit model; oc (k) represents the open-circuit voltage at the current moment; where τ p R s Rp U oc (k) represents the system state variables in the equivalent circuit model, denoted as the parameters to be identified. Therefore, this embodiment uses the collected battery current and battery terminal voltage, employs the MFF-RLS algorithm to recursively derive the current open-circuit voltage (OCV) of the battery online, and obtains an estimated value of the current SOC by looking up a table online based on the offline calibrated OCV-SOC curve. This can, to some extent, solve the problem of accumulated SOC error in the ampere-hour integration method.

[0118] According to the above description, in an exemplary embodiment, when calculating the equivalent circuit model using the recursive least squares method with multiple forgetting factors, this embodiment may further include: vectorizing the calculation formula of the parameters to be identified, that is, adjusting the calculation formula... After vectorization adjustment, we have:

[0119] U t (k)=Φ T (k)·θ(k);

[0120]

[0121] θ(k)=[R s +R p R s *τ p τ p U oc (k)] T .

[0122] According to the above description, in an exemplary embodiment, the process of calculating the equivalent circuit model based on the MFF-RLS algorithm further includes:

[0123] The state of charge (SOC) and parameter vectors of the battery under test are initialized, and the data covariance matrix P(0) = [P1(0) P2(0) P3(0) P4(0)] is determined. T ;

[0124] Based on the vectorization adjustment results and the data covariance matrix, calculate the update gain L for each single-parameter vector component. i (k), we have:

[0125]

[0126] The update gain L is obtained through each single-parameter vector component. i (k) Covariance P for each single-parameter vector component i (k) is updated, and:

[0127]

[0128] Based on the covariance P of each single-parameter vector component i (k) Calculate the update gain L(k) for all parameter vectors, then:

[0129]

[0130] The updated parameter values ​​are calculated based on the update gain L(k) of all parameter vectors, yielding the parameter identification results:

[0131] θ(k)=θ(k-1)+L(k)*[U t (k)-Φ T (k)·θ(k-1)].

[0132] According to the above description, in an exemplary embodiment, the process of validating the parameter identification result includes: determining whether the value of the parameter to be identified in the parameter identification result is within its corresponding preset range, that is, determining whether the time constant of the RC element in the equivalent circuit model is within its corresponding preset value range, whether the ohmic internal resistance in the equivalent circuit model is within its corresponding preset value range, whether the polarization internal resistance in the equivalent circuit model is within its corresponding preset value range, and whether the open-circuit voltage at the current moment is within its corresponding preset value range; calculating the difference between the measured battery voltage and the battery terminal voltage in the parameter identification result, and determining whether the difference is less than a second preset threshold; determining whether the data covariance is less than a third preset threshold; and determining whether the parameter... The parameter identification result is determined to be valid if the fluctuation change of the original estimated value of the open-circuit voltage in the parameter identification result is less than the fourth preset threshold. If the parameter value to be identified is within the corresponding preset range, that is, the time constant of the RC element in the equivalent circuit model is within its corresponding preset value range, the ohmic internal resistance in the equivalent circuit model is within its corresponding preset value range, the polarization internal resistance in the equivalent circuit model is within its corresponding preset value range, the open-circuit voltage at the current moment is within its corresponding preset value range, and the difference between the measured battery voltage and the battery terminal voltage in the parameter identification result is less than the second preset threshold, the data covariance is less than the third preset threshold, and the fluctuation change of the original estimated value of the open-circuit voltage is less than the fourth preset threshold, then the parameter identification result is determined to be valid; otherwise, the parameter identification result is determined to be invalid.

[0133] According to the above description, in an exemplary embodiment, if the battery under test is installed in the target vehicle, before determining the corresponding battery current and battery terminal voltage, this embodiment may further include: when the temperature at the current moment exceeds a preset temperature, determining whether the target vehicle is in dynamic driving condition at the current moment based on the frequency of change of the terminal voltage of the battery under test, the amplitude of change of the terminal voltage, the frequency of change of the current, and the amplitude of change of the current; if the target vehicle is in dynamic driving condition at the current moment, collecting the battery current and battery terminal voltage; if the target vehicle is not in dynamic driving condition at the current moment, re-comparing the dynamic driving condition based on the frequency and amplitude of change of the battery terminal voltage and the frequency and amplitude of change of the battery current. Specifically, in this embodiment, for the battery under test, the frequency of the terminal voltage change can be compared with a preset frequency of terminal voltage change, the amplitude of the terminal voltage change can be compared with a preset amplitude of terminal voltage change, the frequency of the current change can be compared with a preset frequency of current change, and the amplitude of the current change can be compared with a preset amplitude of current change. When the frequency of the terminal voltage change, the amplitude of the terminal voltage change, the frequency of the current change, and the amplitude of the current change of the battery under test are within the corresponding preset ranges, the target vehicle is considered to be in dynamic driving conditions. In this embodiment or other embodiments, the target vehicle includes, but is not limited to, new energy vehicles such as pure electric vehicles and hybrid electric vehicles.

[0134] In another exemplary embodiment of the present invention, this embodiment provides a battery state-of-charge adaptive estimation device, which can be used to perform, for example... Figures 2 to 4 The process is illustrated. In this embodiment, the battery state of charge adaptive estimation device and the battery state of charge adaptive estimation method provided in the above embodiments belong to the same concept. The specific way in which the battery state of charge adaptive estimation method is executed has been described in detail in the above embodiments. Therefore, the technical functions and effects of the battery state of charge adaptive estimation device can be referred to the above method embodiments, and will not be repeated here.

[0135] In summary, this invention provides a battery state-of-charge (POC) adaptive estimation device. First, it calculates the original estimate of the open-circuit voltage using an equivalent circuit model of the battery under test. Then, it filters this original estimate. When the filtered open-circuit voltage estimate is less than a first preset threshold, the open-circuit voltage of the battery under test is considered to be below the inflection point of the open-circuit voltage-POC curve and located in the slope segment of the curve. At this point, the calculated original estimate of the open-circuit voltage is considered valid. Based on this, the corresponding POC estimate is retrieved from the curve using the correspondence between the open-circuit voltage and POC curves, and recorded as the first POC estimate. The first POC estimate is then used to correct the second POC estimate, thereby improving the POC estimation accuracy under dynamic driving conditions. This prevents the battery from entering a low POC range, resulting in an excessively high POC estimate that could lead to undervoltage and over-discharge risks. If the first state of charge (SOC) estimate is obtained using the MFF-RLS algorithm and the second SOC estimate is obtained using the ampere-hour integration method, then this invention can use the SOC estimate obtained by the MFF-RLS algorithm to correct the SOC estimation error of the ampere-hour integration method in real time. Simultaneously, this device iteratively calculates the original OCV estimate using the MFF-RLS algorithm and performs low-pass filtering. When the filtered OCV estimate is less than a preset OCV threshold, the current battery OCV is considered to be below the inflection point of the OCV-SOC curve (i.e., within the slope segment of the OCV-SOC curve). At this point, the OCV calculated by the MFF-RLS algorithm is considered valid. Based on this, the correspondence between OCV and SOC is used to estimate the current SOC of the battery online to correct the SOC estimation error of the ampere-hour integration method in real time, thereby improving the SOC estimation accuracy under dynamic driving conditions and preventing the battery from entering a low SOC range and resulting in an excessively high SOC estimate, which could lead to undervoltage and over-discharge risks. Furthermore, when the open-circuit voltage of the battery under test is below the inflection point of the open-circuit voltage-state-of-charge (OCC) curve, this invention can simultaneously activate the OCC estimation method based on the equivalent circuit model. This avoids mistakenly activating the OCC estimation device based on the equivalent circuit model during voltage plateau conditions, which could lead to incorrect OCC corrections. If the battery under test is a lithium iron phosphate (LFP) battery, this device can effectively identify the inflection point of the OCC-SOC curve under dynamic driving conditions, serving as a prerequisite for activating the model-based SOC estimation method. Simultaneously, it avoids mistakenly activating the model-based SOC estimation method when the LFP battery is in a "plateau" phase, which could lead to incorrect SOC corrections.

[0136] It should be noted that the battery state-of-charge adaptive estimation device provided in the above embodiments and the battery state-of-charge adaptive estimation method provided in the above embodiments belong to the same concept. The specific way in which the battery state-of-charge adaptive estimation method is executed has been described in detail in the above method embodiments. Therefore, in practical applications, the battery state-of-charge adaptive estimation device provided in the above embodiments can allocate the above functions to different functional modules as needed, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described in the above method embodiments. This is not a limitation here.

[0137] In another exemplary embodiment of the present invention, this embodiment also provides a vehicle equipped with the battery state-of-charge adaptive estimation device described in any of the above embodiments. It should be noted that the vehicle provided in this embodiment and the battery state-of-charge adaptive estimation device provided in the above embodiments belong to the same concept, and the specific manner in which the battery state-of-charge adaptive estimation device performs its operation has been described in detail in the above embodiments. Therefore, the technical functions and effects of the vehicle provided in this embodiment in practical applications can be found in the battery state-of-charge adaptive estimation device, and will not be repeated here.

[0138] It should be understood that although the terms first, second, third, etc., may be used in the embodiments of the present invention to describe preset thresholds, these preset thresholds should not be limited to these terms. These terms are only used to distinguish preset thresholds from each other. For example, without departing from the scope of the embodiments of the present invention, a first preset threshold may also be referred to as a second preset threshold, and similarly, a second preset threshold may also be referred to as a first preset threshold.

[0139] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. An adaptive estimation method for battery state of charge, characterized in that, The method includes the following steps: Obtain an equivalent circuit model generated in advance or in real time based on the battery under test, and determine the battery current and battery terminal voltage corresponding to the battery under test; The equivalent circuit model is calculated based on the battery current and the battery terminal voltage to obtain the corresponding parameter identification results; The validity of the parameter identification results is assessed, and when the parameter identification results are valid, the original estimated value of the open-circuit voltage in the parameter identification results is filtered to obtain a filtered estimated value of the open-circuit voltage. The process of assessing the validity of the parameter identification results includes: determining whether the values ​​of the parameters to be identified in the parameter identification results are respectively within the corresponding preset ranges; calculating the difference between the measured battery voltage and the battery terminal voltage in the parameter identification results, and determining whether the difference is less than a second preset threshold; determining whether the data covariance is less than a third preset threshold; and determining whether the fluctuation value of the original estimated value of the open-circuit voltage in the parameter identification results is less than a fourth preset threshold. If the value of the parameter to be identified is within the corresponding preset range, the difference is less than the second preset threshold, the data covariance is less than the third preset threshold, and the fluctuation value of the original estimated value of the open-circuit voltage is less than the fourth preset threshold, then the parameter identification results are determined to be valid; otherwise, the parameter identification results are determined to be invalid. From the pre-calibrated or real-time open-circuit voltage-state-of-charge curves, retrieve the state-of-charge estimate corresponding to the filtered open-circuit voltage estimate, and denote it as the first state-of-charge estimate; and, When the first state of charge estimate is less than the first preset threshold, the first state of charge estimate is used to correct the pre-generated second state of charge estimate in order to estimate the state of charge of the battery under test in real time.

2. The adaptive estimation method for battery state of charge according to claim 1, characterized in that, The first state of charge estimate and the second state of charge estimate are generated in different ways; Alternatively, the first state of charge estimate and the second state of charge estimate can be generated in the same way.

3. The adaptive estimation method for battery state of charge according to claim 2, characterized in that, The method for generating the first state of charge estimate includes: using the original open-circuit voltage estimate obtained by the recursive least squares method with multiple forgetting factors as the original open-circuit voltage estimate in the parameter identification result, filtering the original open-circuit voltage estimate obtained by the recursive least squares method with multiple forgetting factors to obtain the filtered open-circuit voltage estimate, and querying the state of charge estimate corresponding to the filtered open-circuit voltage estimate from the pre-calibrated or real-time open-circuit voltage-state of charge curve as the first state of charge estimate. The method for generating the second state of charge estimate includes the ampere-hour integration method.

4. The adaptive estimation method for battery state of charge according to any one of claims 1 to 3, characterized in that, The process of calculating the parameters to be identified based on the battery current and the battery terminal voltage using the recursive least squares method with multiple forgetting factors includes: In the formula, This indicates the battery terminal voltage at the current moment; This indicates the battery current at the current moment; This represents the time constant of the RC element in the equivalent circuit model; This represents the ohmic internal resistance in the equivalent circuit model; This represents the polarization resistance in the equivalent circuit model; This represents the open-circuit voltage at the current moment; in, , , , Let be the system state variables in the equivalent circuit model, denoted as the parameters to be identified.

5. The adaptive estimation method for battery state of charge according to claim 4, characterized in that, When calculating the equivalent circuit model using the recursive least squares method with multiple forgetting factors, the method further includes: The formulas for calculating the parameters to be identified are vectorized and adjusted, and the results of the vectorization adjustment are used to calculate the parameters to be identified. ； 6. The adaptive estimation method for battery state of charge according to claim 5, characterized in that, The process of calculating the equivalent circuit model also includes: The state of charge (SOC) of the battery under test and the parameter vectors to be identified are initialized, and the data covariance matrix is ​​determined. ; The update gain for each single-parameter vector component is calculated based on the vectorization adjustment results and the data covariance matrix. ,have: Update the gain for each single-parameter vector component. Covariance for each single-parameter vector component Updates are being made, including: Based on the covariance of each single-parameter vector component Calculate the update gain for all parameter vectors ,have: Update gain based on all parameter vectors The updated values ​​of the parameters to be identified are calculated, and the parameter identification results are obtained, including:

7. The adaptive estimation method for battery state of charge according to claim 1, characterized in that, If the battery under test is installed in the target vehicle, the method further includes the following steps before determining the corresponding battery current and battery terminal voltage: When the temperature at the current moment exceeds the preset temperature, the frequency and magnitude of changes in battery terminal voltage and battery current are used to determine whether the target vehicle is in a dynamic driving condition at the current moment. If the target vehicle is in dynamic driving conditions at the current moment, the battery current and battery terminal voltage are collected. If the target vehicle is not in dynamic driving condition at the current moment, a dynamic driving condition comparison will be performed again based on the frequency and magnitude of changes in battery terminal voltage and battery current.

8. A battery state-of-charge adaptive estimation device, characterized in that, The device includes: The equivalent circuit module acquires the equivalent circuit model generated in advance or in real time based on the battery under test, and determines the battery current and battery terminal voltage corresponding to the battery under test. The calculation module is used to calculate the equivalent circuit model based on the battery current and the battery terminal voltage to obtain the corresponding parameter identification results; The open-circuit voltage estimation module is used to identify the validity of the parameter identification result, and when the parameter identification result is valid, to filter the original open-circuit voltage estimate in the parameter identification result to obtain the filtered open-circuit voltage estimate. The process of validating the parameter identification results includes: determining whether the values ​​of the parameters to be identified in the parameter identification results are respectively within the corresponding preset ranges; calculating the difference between the measured battery voltage and the battery terminal voltage in the parameter identification results, and determining whether the difference is less than a second preset threshold; determining whether the data covariance is less than a third preset threshold; and determining whether the fluctuation change value of the original estimated value of the open-circuit voltage in the parameter identification results is less than a fourth preset threshold; if the value of the parameter to be identified is within the corresponding preset range, the difference is less than the second preset threshold, the data covariance is less than the third preset threshold, and the fluctuation change value of the original estimated value of the open-circuit voltage is less than the fourth preset threshold, then the parameter identification results are determined to be valid; otherwise, the parameter identification results are determined to be invalid. The state of charge estimation module is used to query the filtered open-circuit voltage estimate from a pre-calibrated or real-time calibrated open-circuit voltage-state of charge curve, and record it as the first state of charge estimate; and when the first state of charge estimate is less than a first preset threshold, the first state of charge estimate is used to correct the pre-generated second state of charge estimate, so as to estimate the state of charge of the battery under test in real time.

9. A vehicle, characterized in that, The vehicle is equipped with the battery state of charge adaptive estimation device as described in claim 8.

Images