CN-115659649-B - Cloud data-based construction method for real-time complete OCV-SOC curve of lithium battery

CN115659649BCN 115659649 BCN115659649 BCN 115659649BCN-115659649-B

Abstract

The invention discloses a cloud data-based construction method of a real-time complete OCV-SOC curve of a lithium battery, which comprises the following steps of firstly, constructing a battery model, secondly, identifying a discharge segment OCV, identifying the discharge segment OCV based on an analogy method, thirdly, constructing an OCV-SOC model, improving an electrode potential expression by analyzing the relation between the lithium intercalation rate and the SOC of a battery electrode in the electrode potential model, and obtaining the OCV-SOC model, and fourthly, identifying the complete OCV-SOC model based on a charging stage, wherein the complete OCV-SOC model comprises ohmic internal resistance change trend analysis, charging stage characteristic analysis, complete OCV-SOC solution and real-time updating of the OCV-SOC relation. The method has the beneficial effects that the battery OCV is obtained under the condition that the real-vehicle battery pack is not dismounted, the cloud SOC value is corrected, the method is simple and easy to realize, and the problems of low cloud data precision and poor battery state estimation accuracy are solved.

Inventors

WANG LIMEI
CAI YINGFENG
CHEN LONG
SUN HONGLIANG
YAN XUEQING
SUN JINGJING
JIN MENGJIE
GAO KAIXU
LUO FULIN
ZHAO XIULIANG
WANG RUOCHEN
PAN CHAOFENG
SUN XIAODONG

Assignees

江苏大学

Dates

Publication Date: 20260512
Application Date: 20221027

Claims (8)

1. The method for constructing the real-time complete OCV-SOC curve of the lithium battery based on cloud data is characterized by comprising the following steps of: Step one, constructing a battery model; step two, identifying the OCV of the discharge segment, and identifying the OCV of the discharge segment based on an analog method; In the second step, based on analog method, the OCV of the discharge segment is identified, and the functional relation between the terminal voltage U and the time t in the zero input response stage under the working condition of the battery mixed power pulse is adopted, wherein the specific functional relation is as follows: (2) wherein e is a natural constant, =R 1 C 1 is a time constant; the terminal voltage U and the time t of the zero input response stage under the HPPC working condition accord with an exponential function relation, and the exponential function expression is as follows: (3) where y 0 is a generalized representation of U OCV , Is that Is a generalized representation of the number (c), And Representing the coefficient of generalization, Represents a generalized dependent variable; Comparing the formulas (2) and (3), the OCV of the battery is obtained as follows: (4) Taking a data segment of current fluctuation near 0A in cloud data as a zero input response stage, selecting at least 3 voltage and acquisition time data in continuous data segments of current fluctuation near 0A to carry into formula (3) for fitting, further obtaining a parameter y 0 , and obtaining a battery pack open-circuit voltage U OCV according to formula (4); the average open circuit voltage U OCV,dis of the single batteries calculated according to the battery grouping mode is calculated according to the following formula: (5) wherein n is the number of batteries connected in series in the battery pack; Thirdly, constructing an OCV-SOC model, and improving an electrode potential expression by analyzing the relation between the lithium intercalation rate of a battery electrode in the electrode potential model and the SOC to obtain the OCV-SOC model; in the third step, the relation between the lithium intercalation rate and the SOC of the battery electrode in the electrode potential model is analyzed, and the electrode potential expression is improved to obtain an OCV-SOC model, wherein the specific improvement method is as follows: First, the electrode potential model expression is composed of three parts, and the expression is as follows: (6) Wherein U (x) is electrode potential, x is lithium intercalation rate, a 1 ,b 1 ,b 2 ,c 1 ,c 2 ,d i ,e i and f i are related parameters of electrode potential at corresponding temperature, and are rational numbers larger than 0, i is the number of terms of hyperbolic tangent (tanh) function; Analyzing the relation between the lithium intercalation rate of the cathode of the battery and the SOC, changing x in the formula (6) into 1-s, and thus establishing a relation model describing the OCV-SOC of the whole battery, wherein the relation model is specifically expressed as: (7) Wherein s is a battery SOC; And step four, identifying a complete OCV-SOC model based on a charging stage, wherein the complete OCV-SOC model comprises ohmic internal resistance change trend analysis, charging stage characteristic analysis, complete OCV-SOC solution and OCV-SOC relation real-time updating.
2. The method for constructing a real-time complete OCV-SOC curve of a lithium battery based on cloud data according to claim 1, wherein the battery model constructed in the first step is a first-order RC equivalent circuit model, and the equation of the model is as follows: (1) Wherein U OCV is open circuit voltage, I is working current, R 0 is ohmic internal resistance, R 1 is polarized internal resistance, C 1 is polarized capacitance, U 1 is polarized voltage, namely voltage at two ends of R 1 C 1 , and U is terminal voltage.
3. The method for constructing a real-time complete OCV-SOC curve of a lithium battery based on cloud data as claimed in claim 1, wherein the expression (7) of the full-battery OCV-SOC relation model is split into constant terms Index term And tangent function term Wherein, the constant term Wherein a 1 is used for representing the up-down shift of open circuit voltage curve of the battery under different temperatures and aging states, and the index term In (a) And Respectively describing the change trend of two ends of open-circuit voltage curve and hyperbolic tangent function term The voltage plateau used to describe the middle portion of the open circuit voltage curve due to phase change is described by one or more hyperbolic tangent functions, with the term increasing as the corresponding plateau increases.
4. The method for constructing the real-time complete OCV-SOC curve of the lithium battery based on cloud data as claimed in claim 1, wherein in the fourth step, the variation trend of the ohmic internal resistance is analyzed, and the variation rule of R 0 is analyzed based on the relation curve of the ohmic internal resistance R 0 and the SOC at different temperatures measured in a laboratory.
5. The method for constructing the real-time complete OCV-SOC curve of the lithium battery based on cloud data according to claim 1 or 4, wherein the characteristic analysis of the charging stage in the fourth step comprises charging current variation trend analysis, charging voltage and open-circuit voltage relation analysis under a charging condition and charging voltage curve trend analysis in different temperature ranges.
6. The cloud data-based lithium battery real-time complete OCV-SOC curve construction method is characterized in that the charging current change trend analysis is performed on the change condition of the average current of battery unit cells under a typical charging condition based on the collected cloud data; And (3) analyzing the relation between the charging voltage and the open-circuit voltage under the charging working condition, and analyzing the trend of an open-circuit voltage curve and a charging voltage curve based on an open-circuit voltage expression under the low-charging-rate working condition in the formula (8): (8) Wherein U OCV,C is OCV under charging condition, U C is charging voltage, and I C is charging current; the method comprises the steps of analyzing trend of charging voltage curves in different temperature ranges, randomly extracting charging fragment voltages in different temperature ranges from cloud data, calculating average charging voltage of a single battery, analyzing trend of variation of the average charging voltage curves in different temperatures, and comparing and analyzing the average charging voltage curves with OCV-SOC curves measured in a laboratory at the temperature.
7. The method for constructing a real-time complete OCV-SOC curve of a lithium battery based on cloud data according to claim 1, wherein the complete OCV-SOC solution in the fourth step comprises the following steps: firstly, taking an average charging voltage curve of a charging segment as a reference on an OCV-SOC curve measured in a laboratory at the corresponding temperature of the segment, and realizing correction of an SOC value by transversely shifting a certain distance k; Then, taking the curve after correcting the SOC value and the OCV-SOC curve measured by a laboratory at the corresponding temperature of the segment as a reference, and longitudinally translating a certain distance b to enable the curve to coincide with the OCV-SOC curve of the laboratory as much as possible, so as to obtain U OCV,C under the corresponding SOC; Carrying OCV-SOC data measured by a laboratory at a corresponding temperature into a formula (7), fitting by using a least square method to obtain a specific expression of an OCV-SOC relation curve based on an improved electrode potential model, and on the basis of the obtained expression, adding a parameter b in the vertical direction and subtracting a parameter k in the horizontal direction according to the principle of 'left and right addition and right subtraction and upper and lower subtraction' of a function to obtain a charging voltage and SOC relation curve model, wherein the expression is as follows: (9) wherein b is the difference between the charging voltage and OCV in the charging stage, and k is the SOC correction value of the charging segment; The data of the charging segment is brought in, and then the values of b and k can be obtained; Finally, if the translated curve can cover 0-1 SOC and accords with the open circuit voltage characteristic, the following steps are not needed to be continued, otherwise, the following steps are needed to be carried out: And splicing and fusing the translated OCV-SOC of the curve segment which accords with the open circuit voltage characteristic with the OCV-SOC measured by a laboratory at the corresponding temperature, so as to obtain a complete OCV-SOC curve.
8. The method for constructing the real-time complete OCV-SOC curve of the lithium battery based on cloud data according to claim 1, wherein in the fourth step, the OCV-SOC relation is updated in real time, cloud data is divided into a plurality of charge-discharge units according to a charging stage, each charge-discharge unit consists of a complete charging stage and a discharging stage before the charging stage reaches the next charging stage, then the complete OCV-SOC data obtained based on a formula (7) and a charging fragment is fitted by using a least square method, and the OCV-SOC relation parameters of each unit are updated, so that the real-time updating of the OCV-SOC relation curve is realized.

Description

Cloud data-based construction method for real-time complete OCV-SOC curve of lithium battery Technical Field The invention relates to a construction method of an OCV-SOC curve of a power battery, in particular to a construction method of a real-time complete OCV-SOC curve of a lithium battery based on cloud data, and belongs to the technical field of power batteries. Background The accurate estimation of the State of Charge (SOC) of the battery can prevent the battery from being overcharged and overdischarged, improve the battery performance and prolong the service life of the battery. Of the many SOC estimation algorithms, open circuit voltage methods and kalman filter algorithms are the most common methods. The open circuit voltage method is to obtain corresponding open circuit voltages (Open Circuit Voltag e, OCV) under different SOCs through a long-time standing test, so as to establish the relationship between the OCV and the SOCs and realize the SOC estimation based on the open circuit voltages. The Kalman filtering algorithm is to establish a corresponding state equation and an observation equation according to a battery model, and to carry out optimal estimation on the system state in the sense of minimum variance through the observation value and the actual measurement value output by the system, so as to realize SOC estimation. The establishment of the observation equation still takes the OCV-SOC curve as a reference and the terminal voltage as a feedback signal, so that the closed-loop correction of the SOC estimation value is realized. The accuracy of the OCV-SOC relationship has a direct impact on the battery SOC estimation, while the accuracy of the battery State of Health (SOH) estimation also depends on the OCV-SOC relationship. Therefore, an accurate OCV-SOC curve is critical to improve the accuracy of battery state estimation. Since the working condition of long-time standing is not generally existed in the actual vehicle operation engineering, many researchers study the relationship between OCV and SOC through laboratory data in the preliminary stage. However, the experimental environment or simulation is too ideal, and it is often difficult to simulate the complex and variable working conditions in real vehicle operation. In recent years, research on the relation of the OCV and the SOC of a lithium ion battery under actual working conditions is gradually increased, for example, based on the sectional identification of real vehicle data, an identification result is spliced into a long OCV-capacity (Ah) curve, two ends of the identified OCV-Ah curve are supplemented through table lookup of an OCV-Ah database constructed in a laboratory, so that a complete OCV-Ah curve is obtained, based on the sectional identification of the OCV-SOC curve of the real vehicle data, a real vehicle data fragment set is constructed according to data fragments, a reference data set is established according to the actually measured SOC and capacity, and the identification result of the real vehicle data fragment set is spliced into the complete OCV-SOC curve according to the reference data set. However, in the previous researches, a large number of reference databases are required to be established, and the OCV is calculated through the parameter of the battery capacity, so that the time cost is greatly increased, and the initial value of the data segment SOC has a great influence on the identification result. In fact, data uploaded to a large data center, i.e., the cloud, generally reduces accuracy, and the above study does not consider correcting the SOC when identifying the OCV-SOC. Meanwhile, the above-mentioned researches fail to consider the real-time change of the voltage platform trend in the middle of the OCV-SOC curve caused by different phase changes of each charge and discharge process of the battery. Therefore, a model which can adapt to different temperatures and accurately describe the OCV-SOC cannot be constructed, an OCV-SOC curve based on the complete charging stage cannot be constructed, and real-time updating of the OCV-SOC curve is realized. Disclosure of Invention Aiming at the problems in the prior art, the invention provides a method for constructing a real-time complete OCV-SOC curve of a lithium battery based on cloud data. The cloud data discharge segment OCV-SOC relationship is identified by adopting an analogy method, the cloud data discharge segment OCV-SOC relationship is compared with an OCV-SOC curve measured in a laboratory, battery characteristics under actual vehicle working conditions and experimental tests are analyzed, models which can adapt to different temperatures and accurately describe the OCV-SOC are constructed by analyzing the battery OCV-SOC curve characteristics, an OCV-SOC curve with complete charging stage is reconstructed by analyzing charging stage characteristics, and real-time updating of the OCV-SOC curve is realized. The technical scheme is that the method