Method and device for estimating health state of battery pack based on cloud data

Abstract

The invention discloses a method and device for estimating the health state of a battery pack based on cloud data. The method comprises the steps that S1, a virtual battery model is constructed based on the cloud data; s2, obtaining an effective charging segment, and constructing a reference curve OCV-SOC; s3, calculating the standardized battery capacity of the effective charging segment, and obtaining the health state of the standardized battery; s4, acquiring an estimated battery capacity by adopting a recursive capacity estimation method considering measurement and estimation errors, and acquiring an estimated battery health state; and S5, comparing the SOH estimated in the step S4 with the standardized SOH in the step S3, and verifying the effectiveness of the estimation method in the step S4. The SOH of the battery is estimated on line through information in cloud data, the estimation accuracy is improved, and the actual SOH of the battery pack is reflected. For a lithium iron phosphate battery, an effective charging fragment selection method is provided, the estimation precision of the health state of the battery is improved, and the increasingly accurate evaluation requirement of a new energy automobile on the health state of the battery pack is met.

Description

Technical Field

[0001] The present invention relates to a method and device for estimating the health status of a battery pack based on cloud data, and belongs to the technical field of batteries. Background Art

[0002] Accurate estimation of the battery's state of health (SOH) not only reveals the degree of battery aging and performance degradation, but is also crucial to preventing safety issues such as overcharging and over-discharging of the battery. In addition, accurate SOH estimation can help formulate more reasonable charging and discharging strategies, optimize energy management, and ensure that the battery can maximize its energy storage and release capabilities under the premise of safety. Therefore, accurate estimation of battery SOH is of great significance to improving the overall performance and service life of the power battery system.

[0003] At present, the common methods used for battery SOH estimation are mainly divided into direct estimation method and indirect estimation method. Direct estimation methods generally include coulomb counting method, internal resistance measurement method, impedance spectrum measurement method, etc., but they are rarely used due to the problem of measurement and estimation accuracy. Indirect estimation methods are mainly divided into three categories: model-driven methods, data-driven methods and hybrid methods, among which hybrid methods refer to the integration of two or more methods to effectively integrate their respective advantages, so as to achieve high-precision estimation of battery SOH. Current battery SOH estimation methods often involve complex algorithms and are accompanied by huge data storage and processing work. However, in the actual operating environment of the vehicle, due to the limited computing power and insufficient storage space of the on-board battery management system BMS, the cost of such huge data storage and processing is far higher than the benefits it brings. In addition, the complexity of the battery pack itself, such as the inconsistency of single cell performance, uneven temperature distribution and differences in aging rates, further increase the difficulty of estimating the battery pack SOH and put forward higher requirements for computing and storage resources. The cloud platform can receive real-time battery characteristic data transmitted from the vehicle via network communication, and has the ability to store and calculate massive amounts of data, which lays the foundation for the application of complex and high-precision SOH algorithms. However, low-quality data from the cloud platform will lead to large errors in SOH estimation. In addition, due to the unique charging voltage curve of lithium iron phosphate batteries, it is impossible to obtain a charging voltage range that can be effectively calculated, making SOH estimation of lithium iron phosphate batteries more difficult than that of ternary lithium-ion batteries. Summary of the invention

[0004] Purpose of the invention: In view of the deficiencies in the prior art, the present invention provides a method and device for estimating the health status of a battery pack based on cloud data. The present invention analyzes the charging voltage curve characteristics unique to lithium iron phosphate batteries and adopts a recursive capacity estimation method that takes into account measurement and estimation errors to obtain an estimated battery capacity SOH, thereby realizing the estimation of the battery health status.

[0005] Technical solution: A method for estimating the health status of a battery pack based on cloud data, comprising the following steps:

[0006] S1: Build a virtual battery model based on cloud data, select the highest voltage curve during charging and the lowest voltage curve during discharging, build a virtual battery voltage curve during charging and discharging, and calculate the open circuit voltage U OCV ;

[0007] S2: intercept the effective charging segment according to the virtual battery characteristics and construct the benchmark curve OCV-SOC;

[0008] S3: Calculate the standardized battery capacity of the effective charging segment in S2 based on cloud data, and obtain the standardized battery health status SOH 标 , make temperature and rate corrections to the standardized battery capacity;

[0009] S4: The estimated battery capacity is obtained by using a recursive capacity estimation method that takes into account measurement and estimation errors, and the estimated battery health status SOH is obtained based on the reference curve OCV-SOC in S2 估 ;

[0010] S5: The SOH estimated by S4 is compared with the standardized SOH of S3 to verify the effectiveness of the recursive capacity estimation method in S4 that takes into account measurement and estimation errors, based on the trend of time versus mileage and the accuracy of the estimated SOH.

[0011] Preferably, the S1 is specifically:

[0012] According to the highest voltage curve, the lowest voltage curve, and the average voltage curve of the battery pack during charging in the cloud data, the highest voltage curve during charging and the lowest voltage curve during discharging are selected as the virtual battery charging and discharging voltage curves;

[0013] During the charging process, the terminal voltage is determined by the open circuit voltage, ohmic internal resistance and polarization internal resistance. Therefore, according to Kirchhoff's voltage law, the terminal voltage of the battery during the charging process can be expressed as:

[0014] U=U OCV +IR0+U R1 (1)

[0015] In the formula, R0 represents the ohmic internal resistance, U OCV and U represent the open circuit voltage and terminal voltage respectively, U R1 Polarization voltage, I represents the current flowing;

[0016] During the battery aging process, the open circuit voltage remains unchanged, while the ohmic internal resistance and polarization internal resistance gradually increase. Therefore, the terminal voltage of the virtual battery during charging mainly refers to the highest voltage of the battery pack during charging.

[0017] Preferably, S2 is specifically:

[0018] S201: preprocessing the virtual battery charging voltage curve in S1 to obtain a smooth charging voltage curve;

[0019] S202: Based on the smooth charging voltage curve of S201, analyze the charging voltage segment characteristics of the virtual battery:

[0020] Based on the SOC distribution at the beginning and end of charging according to the charging segment statistics, the SOC range is obtained, that is, the SOC range frequently used by users. The charging voltage curve characteristics of the virtual battery are analyzed, and it is found that the curve between the two voltage platforms has an upward trend and is centrally symmetrical.

[0021] S203: Determine the symmetrical center point of the curve in S202 using the capacity increment IC curve. By analyzing the peak-valley position between the two main peaks on the capacity increment IC curve, the symmetrical center point voltage can be determined. With the symmetrical center point as the center, the average value of the voltage difference between the two voltage platforms is selected as the charging voltage range, thereby intercepting the charging segment with the largest voltage change difference, that is, the effective charging segment. Based on the effective charging segment range, the maximum likelihood method is used to perform curve fitting and curve translation based on the intermediate stage charging voltage curve to construct the reference curve OCV-SOC. Among them, the intermediate stage is the charging segment between the two voltage platforms, that is, the curve with a large voltage value change between the two curves where the voltage value changes tend to be stable.

[0022] Preferably, the curve fitting in S203 is specifically:

[0023] The electrode potential model mainly consists of three parts: constant term, exponential term and tangent term. Its specific expression is shown in formula (2), where the number of tangent terms corresponds to the number of voltage platforms in the battery open circuit voltage curve.

[0024]

[0025] Where a1, b1, b2, c1, c2, d1, d2, e1, e2, f1, f2 are all related parameters of electrode potential, and are all greater than 0, and s represents the SOC of the battery;

[0026] The OCV-SOC value is brought in for fitting, and the relevant parameters of the lithium iron phosphate battery based on the electrode potential model are obtained, and the specific fitting expression is obtained.

[0027] Preferably, the curve translation in S203 is specifically:

[0028] Curve translation refers to moving the curve a certain distance in a specific direction on a plane; after translation, the shape and size of the curve remain unchanged, only the position changes; the OCV-SOC curve has a longitudinal offset caused by the internal resistance of the battery, and the accumulated error of SOC will cause the OCV-SOC curve to have a lateral offset; a represents the unit distance of the horizontal translation of the charging voltage curve and the open circuit voltage curve, and b represents the unit distance of the vertical translation of the charging voltage curve and the open circuit voltage curve. The specific process of curve translation is as follows:

[0029]

[0030] Preferably, S3 obtains the standardized battery capacity Q 标 And get the standardized battery health status SOH as follows:

[0031] The ampere-hour integration method is used to calculate the standardized capacity of the virtual battery and used as a verification benchmark. The calculation formula of the ampere-hour integration method is as follows:

[0032]

[0033] In the formula, SOC S Indicates the SOC at the start of the current charging segment, SOC E Indicates the SOC at the end of the current charging segment, It indicates the amount of electricity charged into the battery during the entire current charging segment, and Q indicates the battery capacity corresponding to the current segment;

[0034] By transforming formula (4), we can get:

[0035]

[0036] In the formula, Q 标 represents the standardized battery capacity, ΔQ represents the amount of electricity charged into the battery within time t, and ΔSOC represents the change in battery SOC within time t;

[0037] Standardized battery state of health SOH 标 for:

[0038] Q 标 / Q 额 =SOH 标 (6)

[0039] In the formula, Q 标 represents the standardized battery capacity, Q 额 Indicates the rated capacity of the battery.

[0040] Preferably, the step S3 performs temperature and rate correction on the standardized battery capacity as follows:

[0041] The specific temperature capacity correction process is as follows: The battery capacity when the average battery temperature is close to 25°C is taken as the reference capacity Q 25 ; Calculate the battery capacity Q at any temperature T With the reference capacity Q 25 The ratio of temperature to capacity K T ; Utilize temperature capacity ratio K T To correct the battery capacity;

[0042] The average temperature T under the charging segment is calculated as shown below:

[0043]

[0044] Where, T max and T min Respectively represent the average values ​​of the highest and lowest battery temperatures in this charging segment, and T represents the average temperature of the entire charging segment;

[0045] Calculate the temperature capacity ratio of the battery at different temperatures. The temperature capacity ratio K of the battery T The calculation is as follows:

[0046]

[0047] In the formula, Q T Indicates the battery capacity at different temperatures, Q 25 Indicates the base capacity of the battery at 25°C;

[0048] The specific rate capacity correction process is as follows: the battery capacity when the average charging rate is close to 1 / 3C is used as the reference capacity Q 1 / 3C ; Calculate the battery capacity Q at any rate C With the reference capacity Q 1 / 3C The ratio of the rate capacity ratio K C ; Finally, using the rate capacity ratio K C To correct the battery capacity;

[0049] Calculate the average rate C under the charging segment, the calculation formula is as follows:

[0050]

[0051] Where I represents the average current of the current charging segment, Q 额 Indicates the rated capacity of the battery, and C represents the average charging rate of the current charging segment;

[0052] Calculate the rate capacity ratio of the battery at different rates and the battery rate correction factor KC The calculation is as follows:

[0053]

[0054] In the formula, Q C Indicates the battery capacity at different rates, Q 1 / 3C Indicates the base capacity of the battery at a 1 / 3C rate.

[0055] Preferably, the S4 is specifically:

[0056] According to the calculated open circuit voltage value U OCV The reference curve OCV-SOC is looked up in the table to obtain the corresponding SOC for the following calculations. Combining the input variable noise, that is, the noise of the battery SOC difference, and the variable noise, that is, the noise of the current sensor, the battery capacity is estimated by the recursive least squares method:

[0057]

[0058] In the formula, and They represent the SOC of the battery at time t1 and time t2, Q 待 Indicates the battery capacity to be estimated, I t It represents the current size of the battery when it is charging at time t. Represents the noise of SOC difference, W t represents the measurement noise of the current sensor at time t, where the noise is all Gaussian white noise;

[0059] In formula (11), and V k The capacity estimation model is transformed into a mathematical model with Gaussian white noise representing zero mean, where η is the coulomb efficiency coefficient when the battery is charged, specifically:

[0060] y k =Q 待 (x k +V k )+W k (12)

[0061] Based on formula (12), the recursive method is used to implement Q 待 Calculate the estimated battery health state SOH 估 As follows:

[0062] Q 待 / Q 额 =SOH 估 (13)

[0063] In the formula, Q 待Indicates the battery capacity to be estimated, Q 额 Indicates the rated capacity of the battery.

[0064] Preferably, in S4, a recursive method is used to implement Q by recursive iteration. 待 The calculation is as follows:

[0065] For the convenience of analysis, the following three matrices α1, α2 and α3 are defined, and their changes with time parameter k are expressed as α 1,k , α 2,k and α 3,k . Parameter initialization α 1，0 , α 2,0 , α 3,0 , Q0, the calculation formula is as follows:

[0066] α1=E{x·x T}(14)

[0067]

[0068] Autocorrelation matrix update α 1,k , α 2,k , α 3,k , the calculation formula is as follows:

[0069]

[0070] Parameter update β 1,k , β 2,k , β 3,k , the calculation formula is as follows:

[0071]

[0072] In the formula, represents the ratio between the current sensor noise and the SOC difference noise variance, where represents the current sensor noise variance, represents the SOC difference noise variance;

[0073] Gain coefficient update K k , the calculation formula is as follows:

[0074]

[0075] Capacity Update Q k , the calculation formula is as follows:

[0076]

[0077] The recursive complete least squares method first requires the parameter α 1，0 , α 2,0 , α 3,0, Q0 are initialized to 1, Q Rated , Q Rated , Q Rated is the rated capacity of the battery, and then the iteration is repeated until the iteration reaches the length of the recursive unit.

[0078] A device for realizing a method for estimating a battery pack health status based on cloud data comprises a cloud data acquisition module, a cloud data storage module, a cloud data processing module, and a capacity calculation module which are sequentially connected by signals.

[0079] The cloud data acquisition module is used to collect the charging and discharging voltage of the virtual battery, the virtual battery temperature, and the virtual battery SOC;

[0080] The cloud data storage module is used to store the data collected by the cloud data collection module;

[0081] The cloud data processing module is used to screen, delete, and make up for missing and abnormal data, and to construct a virtual battery based on the stored data in the cloud data storage module, to smooth the charging voltage curve of the virtual battery, and to intercept effective charging segments;

[0082] The capacity calculation module is used to construct a reference curve OCV-SOC, and estimate the battery capacity by recursive least squares method in combination with battery SOC difference noise and current sensor noise.

[0083] This device takes into full account the characteristics of battery types and the complexity of battery pack capacity estimation, aiming to provide a simple, universal and efficient battery pack capacity estimation method. By utilizing the efficient data management and processing capabilities of the cloud platform, the battery pack health status estimation process is greatly optimized to ensure the accuracy and efficiency of capacity calculation.

[0084] Beneficial effects: The present invention estimates the battery's state of health SOH online by utilizing a large amount of information in cloud data, thereby improving the accuracy of the estimation and truly reflecting the actual health status of the battery pack. At the same time, for lithium iron phosphate batteries, an effective charging segment selection method is proposed, which greatly improves the accuracy of health status estimation for such batteries, meets the increasingly accurate assessment needs of new energy vehicles for the health status of battery packs, and provides important support for battery management and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0086] Figure 1 is a flow chart of the method of the present invention;

[0087] Figure 2 It is the relationship diagram between the single cell voltage and the highest / lowest voltage;

[0088] Figure 3 This is the Thevenin model diagram;

[0089] Figure 4 This is the charging voltage curve after smoothing by the Savitzky-Golay method;

[0090] Figure 5 This is a charging voltage curve of a lithium iron phosphate battery;

[0091] Figure 6 It is a comparison chart of IC curves of the complete charging fragment and the intercepted fragment;

[0092] Figure 7 This is a comparison chart before and after the curve is translated;

[0093] Figure 8 It is a scatter plot of standardized capacity and mileage;

[0094] Fig. 9 Plot of estimated SOH versus normalized SOH for a battery pack. DETAILED DESCRIPTION

[0095] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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.

[0096] In the description of the present invention, it is necessary to understand that the terms "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship are based on the orientation or position 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 a limitation on the present invention.

[0097] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may include that the first and second features are in direct contact, or may include that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, a first feature being "above", "above" and "above" a second feature includes that the first feature is directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below", "below" and "below" a second feature includes that the first feature is directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0098] like Figure 1 As shown, a method for estimating the health status of a battery pack based on cloud data includes the following steps:

[0099] S1: Build a virtual battery model based on cloud data, select the highest voltage curve during charging and the lowest voltage curve during discharging, build a virtual battery voltage curve during charging and discharging, and calculate the open circuit voltage U OCV ;

[0100] In practical applications, multiple cells are usually combined into a battery pack in series-parallel mode. The overall performance of the battery pack is limited by the cell with the worst performance. In the actual charging and discharging process, the cell that is overcharged or over-discharged first is usually regarded as the cell with the worst performance. By establishing a relationship between the worst cell parameters and the battery pack parameters, the worst cell can be used to describe the status of the entire battery pack.

[0101] The present invention mainly studies the charging stage. Multiple key parameters in the charging process, including open circuit voltage, ohmic internal resistance and polarization internal resistance, jointly determine the size of the charging voltage. During the aging of the battery, the open circuit voltage usually remains unchanged, while the ohmic internal resistance and polarization internal resistance will gradually increase. For this reason, the virtual battery voltage mainly refers to the highest voltage of the battery pack during the charging process. Subsequently, the highest voltage of the single cell in the cloud data will be used to construct a virtual battery charging voltage to realize the estimation of the battery pack SOH.

[0102] According to the voltage values ​​of the battery pack during charging and discharging in the cloud data, the highest voltage curve, the lowest voltage curve, and the average voltage curve are constructed, and the highest voltage curve during charging and the lowest voltage curve during discharging are selected as the virtual battery charging and discharging voltage curves;

[0103] The basic parameters of the lithium iron phosphate battery of this embodiment are shown in Table 1:

[0104] Table 1 Basic parameters of lithium iron phosphate battery

[0105] parameter Specification Battery Type Lithium iron phosphate battery Nominal voltage 3.2V Operating voltage 2.5V-3.65V Nominal capacity 20Ah

[0106] The highest voltage curve, the lowest voltage curve, and the average voltage curve of lithium iron phosphate battery are as follows: Figure 2 As shown, from Figure 2 It can be seen from the comparison that during the charging process, the single cell voltage is contained within the envelope of the highest and lowest voltages, which means that the virtual battery state constructed using the highest or lowest voltage can be used to summarize the voltage states of all single cells.

[0107] This paper chooses to use the first-order RC model, namely the Thevenin model, to simulate the external characteristic behavior of the single battery. Figure 3 As shown. During the charging process, the terminal voltage is determined by the open circuit voltage, ohmic internal resistance and polarization internal resistance. Therefore, according to Kirchhoff's voltage law, the terminal voltage of the battery during the charging process can be expressed as:

[0108] U=U OCV +IR0+U R1 (1)

[0109] In the formula, R0 represents the ohmic internal resistance, U OCV and U represent the open circuit voltage and terminal voltage respectively, U R1 Polarization voltage, I represents the current flowing;

[0110] During the battery aging process, the open circuit voltage remains unchanged, while the ohmic internal resistance and polarization internal resistance gradually increase. Therefore, the terminal voltage of the virtual battery during the charging process mainly refers to the highest voltage of the battery pack during the charging process. Subsequently, the highest voltage of the single cell in the cloud data will be used to construct the virtual battery voltage to realize the estimation of the battery pack SOH.

[0111] S2: intercepting effective charging segments according to the virtual battery characteristics;

[0112] S201: preprocessing the virtual battery charging voltage curve in S1 to obtain a smooth charging voltage curve;

[0113] like Figure 4 As shown in the figure, due to the problems of low collection accuracy and frequency of cloud data, there are relatively more glitches on the charging voltage curve, which has a certain impact on the subsequent results. The Savitzky-Golay filter fitting method is used to process the charging voltage curve, which can effectively retain the overall trend and shape of the curve while significantly suppressing high-frequency noise, thereby obtaining a smooth charging voltage curve.

[0114] S202: Based on the smooth charging voltage curve of S201, analyze the charging voltage segment characteristics of the virtual battery:

[0115] Based on the SOC distribution at the beginning and end of charging according to the charging segment statistics, the SOC range is obtained, that is, the SOC range frequently used by users. The charging voltage curve characteristics of the virtual battery are analyzed, and it is found that the curve between the two voltage platforms has an upward trend and is centrally symmetrical.

[0116] Since there are obvious differences between different charging voltage curves in the cloud data charging voltage curves at the beginning and end of charging. Therefore, it is necessary to select a charging segment that covers the usage habits of most users. Here, the distribution of SOC at the beginning and end of charging is based on the statistics of charging segments. By statistics and analysis of the distribution of charging start / end SOC in cloud data, it can be seen that the starting SOC of users when charging is generally concentrated between 0.2 and 0.8, indicating that users usually do not charge when the battery is too low or completely exhausted. On the contrary, when users finish charging, the SOC is mostly concentrated in the range of 0.9 to 1, indicating that users tend to charge the battery to a nearly full state. Statistical analysis found that the commonly used SOC range is 0.2 to 1.

[0117] In the above SOC range, there are two obvious voltage platforms in the charging segment of lithium iron phosphate battery, such as Figure 5 The voltage change between the two platforms is very small. This small voltage change increases the uncertainty of estimating the battery SOC by using the method of reverse lookup of the open circuit voltage and state of charge curve OCV-SOC after the charging voltage is shifted, which can easily lead to large errors.

[0118] By analyzing the charging voltage curve characteristics of lithium iron phosphate batteries, it is found that there is a stage between the two voltage platforms that shows an obvious upward trend and is centrally symmetrical. Once this symmetric point is found, the voltage range can be controlled using this point, thereby effectively intercepting the required charging segment.

[0119] S203: Determine the symmetrical center point of the curve in S202 using the capacity increment IC curve. By analyzing the peak-valley position between the two main peaks on the capacity increment IC curve, the symmetrical center point voltage can be determined. With the symmetrical center point as the center, the average value of the voltage difference between the two voltage platforms is selected as the charging voltage range, thereby intercepting the charging segment with the largest voltage change difference, that is, the effective charging segment. Based on the effective charging segment range, the maximum likelihood method is used to perform curve fitting and curve translation based on the intermediate stage charging voltage curve to construct the reference curve OCV-SOC. Among them, the intermediate stage is the charging segment between the two voltage platforms, that is, the curve with a large voltage value change between the two curves where the voltage value changes tend to be stable.

[0120] like Figure 6As shown, it is a comparison of IC curves of complete segments and intercepted segments of lithium iron phosphate batteries under laboratory conditions. Following the principle of selecting segments with larger voltage changes, it is found that the voltage difference between the two voltage platforms of the lithium iron phosphate battery used in the research is about 30mV. Based on the above analysis, for the interception of the charging segment of the lithium iron phosphate battery, this embodiment selects a voltage range of 15mV above and below the symmetrical center point to intercept the required charging segment.

[0121] The curve fitting is as follows:

[0122] The electrode potential model mainly consists of three parts: constant term, exponential term and tangent term. Its specific expression is shown in formula (2), where the number of tangent terms corresponds to the number of voltage platforms in the battery open circuit voltage curve.

[0123]

[0124] Where a1, b1, b2, c1, c2, d1, d2, e1, e2, f1, f2 are all related parameters of electrode potential, and are all greater than 0, and s represents the SOC of the battery;

[0125] The OCV-SOC value is brought in for fitting, and the relevant parameters of the lithium iron phosphate battery based on the electrode potential model are obtained, and the specific fitting expression is obtained.

[0126] The specific curve translation is:

[0127] Curve translation refers to moving the curve a certain distance in a specific direction on a plane; after translation, the shape and size of the curve remain unchanged, only the position changes; the OCV-SOC curve has a longitudinal offset caused by the internal resistance of the battery, and the accumulated error of SOC will cause the OCV-SOC curve to have a lateral offset; a represents the unit distance of the horizontal translation of the charging voltage curve and the open circuit voltage curve, and b represents the unit distance of the vertical translation of the charging voltage curve and the open circuit voltage curve. The specific process of curve translation is as follows:

[0128]

[0129] Figure 7 It shows the result of shifting the curve of lithium iron phosphate battery using the maximum likelihood method. As can be seen from the figure, after the maximum likelihood method shift, the curve has a good overlap with the OCV-SOC curve.

[0130] S3: Calculate the standardized battery capacity of the effective charging segment in S2 based on cloud data, and obtain the standardized battery health status SOH 标 , make temperature and rate corrections to the standardized battery capacity;

[0131] The ampere-hour integration method is used to calculate the standardized capacity of the virtual battery and used as a verification benchmark. The calculation formula of the ampere-hour integration method is as follows:

[0132]

[0133] In the formula, SOC S Indicates the SOC at the start of the current charging segment, SOC E Indicates the SOC at the end of the current charging segment, It indicates the amount of electricity charged into the battery during the entire current charging segment, and Q indicates the battery capacity corresponding to the current segment;

[0134] By transforming formula (4), we can get:

[0135]

[0136] In the formula, Q 标 represents the standardized battery capacity, ΔQ represents the amount of electricity charged into the battery within time t, and ΔSOC represents the change in battery SOC within time t;

[0137] Standardized battery state of health SOH 标 for:

[0138] Q 标 / Q 额 =SOH 标 (6)

[0139] In the formula, Q 标 represents the standardized battery capacity, Q 额 Indicates the rated capacity of the battery.

[0140] The specific temperature capacity correction process is as follows: The battery capacity when the average battery temperature is close to 25°C is taken as the reference capacity Q 25 ; Calculate the battery capacity Q at any temperature T With the reference capacity Q 25 The ratio of temperature to capacity K T ; Utilize temperature capacity ratio K T To correct the battery capacity;

[0141] The average temperature T under the charging segment is calculated as shown below:

[0142]

[0143] Where, T max and T min Respectively represent the average values ​​of the highest and lowest battery temperatures in this charging segment, and T represents the average temperature of the entire charging segment;

[0144] Calculate the temperature capacity ratio of the battery at different temperatures. The temperature capacity ratio K of the battery T The calculation is as follows:

[0145]

[0146] In the formula, Q T Indicates the battery capacity at different temperatures, Q 25 Indicates the base capacity of the battery at 25°C;

[0147] The specific rate capacity correction process is as follows: the battery capacity when the average charging rate is close to 1 / 3C is used as the reference capacity Q 1 / 3C ; Calculate the battery capacity Q at any rate C With the reference capacity Q 1 / 3C The ratio of the rate capacity ratio K C ; Finally, using the rate capacity ratio K C To correct the battery capacity;

[0148] Calculate the average rate C under the charging segment, the calculation formula is as follows:

[0149]

[0150] Where I represents the average current of the current charging segment, Q 额 Indicates the rated capacity of the battery, and C represents the average charging rate of the current charging segment;

[0151] Calculate the rate capacity ratio of the battery at different rates and the battery rate correction factor K C The calculation is as follows:

[0152]

[0153] In the formula, Q C Indicates the battery capacity at different rates, Q 1 / 3C Indicates the base capacity of the battery at a 1 / 3C rate.

[0154] Draw a scatter plot reflecting the relationship between standardized capacity and mileage, such as Figure 8 As shown, the influence of temperature and charging rate on battery capacity can be restored through correction and completely standardized.

[0155] S4: Use the recursive capacity estimation method that takes into account measurement and estimation errors to obtain the estimated battery capacity and the estimated battery health status SOH 估 ;

[0156] The open circuit voltage value U calculated based on the cloud data OCV, and then look up the OCV-SOC table to get the corresponding SOC value for the following calculations. When using low-precision cloud data to estimate battery capacity through recursive least squares, additional current noise will be introduced, which will affect the accuracy of the estimation. Traditional recursive least squares usually mainly considers input variable noise, that is, the noise of the battery SOC difference, and often ignores output variable noise, that is, the noise of the current sensor.

[0157] Combining the input variable noise, that is, the noise of the battery SOC difference, and the variable noise, that is, the noise of the current sensor, the battery capacity is estimated by recursive least squares method:

[0158]

[0159] In the formula, and They represent the SOC of the battery at time t1 and time t2, Q 待 Indicates the battery capacity to be estimated, I t It represents the current size of the battery when it is charging at time t. Represents the noise of SOC difference, W t represents the measurement noise of the current sensor at time t, where the noise is all Gaussian white noise;

[0160] In formula (11), and V k The capacity estimation model is transformed into a mathematical model with Gaussian white noise representing zero mean, where η is the coulomb efficiency coefficient when the battery is charged, specifically:

[0161] y k =Q 待 (x k +V k )+W k (12)

[0162] Based on formula (12), the recursive method is used to implement Q 待 Calculate the estimated battery health state SOH 估 As follows:

[0163] Q 待 / Q 额 =SOH 估 (13)

[0164] In the formula, Q 标 represents the standardized battery capacity, Q 额 Indicates the rated capacity of the battery.

[0165] For the convenience of analysis, the following three matrices α1, α2 and α3 are defined, and their changes with time parameter k are expressed as α 1,k , α 2,k and α 3,k . Parameter initialization α 1，0 , α 2,0 , α 3,0 , Q0, the calculation formula is as follows:

[0166] α1=E{x·x T} (14)

[0167]

[0168] Autocorrelation matrix update α 1,k , α 2,k , α 3,k , the calculation formula is as follows:

[0169]

[0170] Parameter update β 1,k , β 2,k , β 3,k , the calculation formula is as follows:

[0171]

[0172] In the formula, represents the ratio between the current sensor noise and the SOC difference noise variance, where represents the current sensor noise variance, represents the SOC difference noise variance;

[0173] Gain coefficient update K k , the calculation formula is as follows:

[0174]

[0175] Capacity Update Q k , the calculation formula is as follows:

[0176]

[0177] The recursive complete least squares method first requires the parameter α 1，0 , α 2,0 , α 3,0 , Q0 are initialized to 1, Q Rated , Q Rated , Q Rated is the rated capacity of the battery, and then the iteration is repeated until the iteration reaches the length of the recursive unit.

[0178] like Fig. 9 As shown in Figure 5, S5: The estimated SOH of S4 is compared with the standardized SOH of S3, and the effectiveness of the recursive capacity estimation method in S4 that considers measurement and estimation errors is verified from the trend of time changing with mileage and the accuracy of the estimated SOH. As can be seen from the figure, although the estimated battery pack SOH has certain fluctuations, the estimated battery pack SOH curve and the standardized SOH curve show a similar attenuation trend as a whole, and the estimation error of both batteries does not exceed 3%.

[0179] A device for realizing a method for estimating a battery pack health status based on cloud data comprises a cloud data acquisition module, a cloud data storage module, a cloud data processing module, and a capacity calculation module which are sequentially connected by signals.

[0180] The cloud data acquisition module is used to collect the charging and discharging voltage of the virtual battery, the virtual battery temperature, and the virtual battery SOC;

[0181] The cloud data storage module is used to store the data collected by the cloud data collection module;

[0182] The cloud data processing module is used to screen, delete, and make up for missing and abnormal data, and to construct a virtual battery based on the stored data in the cloud data storage module, to smooth the charging voltage curve of the virtual battery, and to intercept effective charging segments;

[0183] The capacity calculation module is used to construct a reference curve OCV-SOC, and estimate the battery capacity by recursive least squares method in combination with battery SOC difference noise and current sensor noise.

[0184] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0185] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for estimating the health status of a battery pack based on cloud data, characterized in that: The following steps are involved: S1: Build a virtual battery model based on cloud data, select the highest voltage curve during charging and the lowest voltage curve during discharging, build a virtual battery voltage curve during charging and discharging, and calculate the open circuit voltage U OCV ; S2: intercept the effective charging segment according to the virtual battery characteristics and construct the benchmark curve OCV-SOC; S3: Calculate the standardized battery capacity of the effective charging segment in S2 based on cloud data, and obtain the standardized battery health status SOH 标 , make temperature and rate corrections to the standardized battery capacity; S4: The estimated battery capacity is obtained by using a recursive capacity estimation method that takes into account measurement and estimation errors, and the estimated battery health status SOH is obtained based on the reference curve OCV-SOC in S2 估 ; S5: The SOH estimated by S4 is compared with the standardized SOH of S3 to verify the effectiveness of the recursive capacity estimation method in S4 that takes into account measurement and estimation errors, based on the trend of time versus mileage and the accuracy of the estimated SOH.

2. The method for estimating the health status of a battery pack based on cloud data according to claim 1, characterized in that: The S1 is specifically: According to the highest voltage curve, the lowest voltage curve, and the average voltage curve of the battery pack during charging in the cloud data, the highest voltage curve during charging and the lowest voltage curve during discharging are selected as the virtual battery charging and discharging voltage curves; During the charging process, the terminal voltage is determined by the open circuit voltage, ohmic internal resistance and polarization internal resistance. Therefore, according to Kirchhoff's voltage law, the terminal voltage of the battery during the charging process can be expressed as: U=U OCV +IR0+U R1 (1) In the formula, R0 represents the ohmic internal resistance, U OCV and U represent the open circuit voltage and terminal voltage respectively, U R1 Polarization voltage, I represents the current flowing; During the battery aging process, the open circuit voltage remains unchanged, while the ohmic internal resistance and polarization internal resistance gradually increase. Therefore, the terminal voltage of the virtual battery during the charging process mainly refers to the highest voltage of the battery pack during the charging process. From this, the open circuit voltage U can be calculated from the above content. OCV .

3. The method for estimating the health status of a battery pack based on cloud data according to claim 1, characterized in that: The S2 is specifically: S201: preprocessing the virtual battery charging voltage curve in S1 to obtain a smooth charging voltage curve; S202: Based on the smooth charging voltage curve of S201, analyze the charging voltage segment characteristics of the virtual battery: Based on the SOC distribution at the beginning and end of charging according to the charging segment statistics, the SOC range is obtained, that is, the SOC range frequently used by users. The charging voltage curve characteristics of the virtual battery are analyzed, and it is found that the curve between the two voltage platforms has an upward trend and is centrally symmetrical. S203: Determine the symmetrical center point of the curve in S202 using the capacity increment IC curve. By analyzing the peak-to-valley position between the two main peaks on the capacity increment IC curve, the symmetrical center point voltage can be determined. With the symmetrical center point as the center, the average value of the voltage difference between the two voltage platforms is selected as the charging voltage range, thereby intercepting the charging segment with the largest voltage change difference, that is, the effective charging segment. Based on the effective charging segment range, the maximum likelihood method is used to perform curve fitting and curve shifting based on the intermediate stage charging voltage curve to construct a reference curve OCV-SOC, wherein the intermediate stage is the charging segment between the two voltage platforms, that is, the curve with a large voltage value change between the two curves where the voltage value changes tend to be stable.

4. The method for estimating the health status of a battery pack based on cloud data according to claim 3, characterized in that: The curve fitting in S203 is specifically as follows: The electrode potential model mainly consists of three parts: constant term, exponential term and tangent term. Its specific expression is shown in formula (2), where the number of tangent terms corresponds to the number of voltage platforms in the battery open circuit voltage curve. Where a1, b1, b2, c1, c2, d1, d2, e1, e2, f1, f2 are all related parameters of electrode potential, and are all greater than 0, and s represents the SOC of the battery; The OCV-SOC value is brought in for fitting, and the relevant parameters of the lithium iron phosphate battery based on the electrode potential model are obtained, and the specific fitting expression is obtained.

5. The method for estimating the health status of a battery pack based on cloud data according to claim 3, characterized in that: The curve translation in S203 is specifically: Curve translation refers to moving the curve a certain distance in a specific direction on a plane; after translation, the shape and size of the curve remain unchanged, only the position changes; the OCV-SOC curve has a longitudinal offset caused by the internal resistance of the battery, and the accumulated error of SOC will cause the OCV-SOC curve to have a lateral offset; a represents the unit distance of the horizontal translation of the charging voltage curve and the open circuit voltage curve, and b represents the unit distance of the vertical translation of the charging voltage curve and the open circuit voltage curve. The specific process of curve translation is as follows:

6. The method for estimating the health status of a battery pack based on cloud data according to claim 1, characterized in that: S3 obtains the standardized battery capacity Q 标 And get the standardized battery health status SOH as follows: The ampere-hour integration method is used to calculate the standardized capacity of the virtual battery and used as a verification benchmark. The calculation formula of the ampere-hour integration method is as follows: In the formula, SOC S Indicates the SOC at the start of the current charging segment, SOC E Indicates the SOC at the end of the current charging segment, It indicates the amount of electricity charged into the battery during the entire current charging segment, and Q indicates the battery capacity corresponding to the current segment; By transforming formula (4), we can get: In the formula, Q 标 represents the standardized battery capacity, ΔQ represents the amount of electricity charged into the battery within time t, and ΔSOC represents the change in battery SOC within time t; Standardized battery state of health SOH 标 for: Q 标 / Q 额 =SOH 标 (6) In the formula, Q 标 represents the standardized battery capacity, Q 额 Indicates the rated capacity of the battery.

7. The method for estimating the health status of a battery pack based on cloud data according to claim 1, characterized in that: The S3 performs temperature and rate correction on the standardized battery capacity as follows: The specific temperature capacity correction process is as follows: The battery capacity when the average battery temperature is close to 25°C is taken as the reference capacity Q 25 ; Calculate the battery capacity Q at any temperature T With the reference capacity Q 25 The ratio of temperature to capacity K T ; Utilize temperature capacity ratio K T To correct the battery capacity; The average temperature T under the charging segment is calculated as shown below: Where, T max and T min Respectively represent the average values ​​of the highest and lowest battery temperatures in this charging segment, and T represents the average temperature of the entire charging segment; Calculate the temperature capacity ratio of the battery at different temperatures. The temperature capacity ratio K of the battery T The calculation is as follows: In the formula, Q T Indicates the battery capacity at different temperatures, Q 25 Indicates the base capacity of the battery at 25°C; The specific rate capacity correction process is as follows: the battery capacity when the average charging rate is close to 1 / 3C is used as the reference capacity Q 1 / 3C ; Calculate the capacity Q of the battery at any rate C With the reference capacity Q 1 / 3C The ratio of the rate capacity ratio K C ; Finally, using the rate capacity ratio K C To correct the battery capacity; Calculate the average rate C under the charging segment, the calculation formula is as follows: Where I represents the average current of the current charging segment, Q 额 Indicates the rated capacity of the battery, and C represents the average charging rate of the current charging segment; Calculate the rate capacity ratio of the battery at different rates and the battery rate correction factor K C The calculation is as follows: In the formula, Q C Indicates the battery capacity at different rates, Q 1 / 3C Indicates the base capacity of the battery at a 1 / 3C rate.

8. The method for estimating the health status of a battery pack based on cloud data according to claim 1, characterized in that: The S4 is specifically: According to the calculated open circuit voltage value U OCV The reference curve OCV-SOC is looked up in the table to obtain the corresponding SOC for the following calculations. The battery capacity is estimated by the recursive least squares method, combining the input variable noise, that is, the noise of the battery SOC difference, and the variable noise, that is, the noise of the current sensor: In the formula, SOC t1 and SOC t2 They represent the SOC of the battery at time t1 and time t2, Q 待 Indicates the battery capacity to be estimated, I t It represents the current size of the battery when it is charging at time t, V t1～t2 Represents the noise of SOC difference, W t represents the measurement noise of the current sensor at time t, where the noise is all Gaussian white noise; In formula (11), let x k =SOC t2 -SOC t1 , and V k The capacity estimation model is transformed into a mathematical model, where η is the coulomb efficiency coefficient when the battery is charged, Q 待 Indicates that the battery capacity to be estimated is: y k =Q 待 (x k +V k )+W k (12) Based on formula (12), the recursive method is used to implement Q 待 Calculate the estimated battery health state SOH 估 As follows: Q 待 / Q 额 =SOH 估 (13) In the formula, Q 待 Indicates the battery capacity to be estimated, Q 额 Indicates the rated capacity of the battery.

9. The method for estimating the health status of a battery pack based on cloud data according to claim 8, characterized in that: In S4, a recursive method is used to implement Q through recursive iteration. 待 The calculation is as follows: For the convenience of analysis, the following three matrices α1, α2 and α3 are defined, and their changes with time parameter k are expressed as α 1,k , α 2,k and α 3,k , parameter initialization α 1，0 , α 2,0 , α 3,0 , Q0, the calculation formula is as follows: α1=E{x·x T } (14) Autocorrelation matrix update α 1,k , α 2,k , α 3,k , the calculation formula is as follows: Parameter update β 1,k , β 2,k , β 3,k , the calculation formula is as follows: In the formula, represents the ratio between the current sensor noise and the SOC difference noise variance, where represents the current sensor noise variance, represents the SOC difference noise variance; Gain coefficient update K k , the calculation formula is as follows: Capacity Update Q k , the calculation formula is as follows: The recursive complete least squares method first requires the parameter α 1，0 , α 2,0 , α 3,0 , Q0 are initialized to 1, Q Rated , Q Rated , Q Rated is the rated capacity of the battery, and then the iteration is repeated until the iteration reaches the length of the recursive unit.

10. A device for implementing the method for estimating the health status of a battery pack based on cloud data as described in any one of claims 1 to 9, characterized in that: It includes a cloud data acquisition module, a cloud data storage module, a cloud data processing module, and a capacity calculation module, which are connected in sequence. The cloud data acquisition module is used to collect the charging and discharging voltage of the virtual battery, the virtual battery temperature, and the virtual battery SOC; The cloud data storage module is used to store the data collected by the cloud data collection module; The cloud data processing module is used to screen, delete, and make up for missing and abnormal data, and to construct a virtual battery based on the stored data in the cloud data storage module, to smooth the charging voltage curve of the virtual battery, and to intercept effective charging segments; The capacity calculation module is used to construct a reference curve OCV-SOC, and estimate the battery capacity by recursive least squares method in combination with battery SOC difference noise and current sensor noise.

Images