CN112379295A - Method, system and storage medium for predicting health state of power battery - Google Patents

Method, system and storage medium for predicting health state of power battery Download PDF

Info

Publication number
CN112379295A
CN112379295A CN202011050925.9A CN202011050925A CN112379295A CN 112379295 A CN112379295 A CN 112379295A CN 202011050925 A CN202011050925 A CN 202011050925A CN 112379295 A CN112379295 A CN 112379295A
Authority
CN
China
Prior art keywords
soh
power battery
value
round
internal resistance
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011050925.9A
Other languages
Chinese (zh)
Other versions
CN112379295B (en
Inventor
刘兴涛
刘晓剑
武骥
何耀
刘新天
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hefei University of Technology
Original Assignee
Hefei University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hefei University of Technology filed Critical Hefei University of Technology
Priority to CN202011050925.9A priority Critical patent/CN112379295B/en
Publication of CN112379295A publication Critical patent/CN112379295A/en
Application granted granted Critical
Publication of CN112379295B publication Critical patent/CN112379295B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/392—Determining battery ageing or deterioration, e.g. state of health
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/60—Other road transportation technologies with climate change mitigation effect
    • Y02T10/70—Energy storage systems for electromobility, e.g. batteries

Abstract

The embodiment of the invention provides a method, a system and a storage medium for predicting the health state of a power battery, belonging to the technical field of maintenance of the power battery. The method comprises the following steps: building a first-order Thevenin model based on the working condition of the power battery; carrying out parameter identification on the internal resistance of the first-order Thevenin model; estimating the internal resistance of the power battery according to the result of parameter identification by adopting a Kalman filtering algorithm; determining a first SOH of the power battery according to the internal resistance; acquiring charging voltage data of the power battery; inputting the charging voltage data into a trained random forest algorithm model to obtain a second SOH of the power battery; and weighting the first SOH and the second SOH to obtain the SOH of the power battery. The method, the system and the storage medium can accurately estimate the state of health of the power battery.

Description

Method, system and storage medium for predicting health state of power battery
Technical Field
The invention relates to the technical field of maintenance of power batteries, in particular to a method, a system and a storage medium for predicting the health state of a power battery.
Background
The continuous and persistent energy crisis and environmental protection issues have prompted the newer iterative lithium ion batteries in electric vehicle technology to provide power for electric vehicles as power batteries with their advantages of high energy density, low self-discharge rate, long cycle life, and the like. The aging degree of the lithium battery is represented by the indexes of the State of Health (SOH), and the charging and discharging performance and safety of the whole battery pack can be guaranteed by estimating and predicting the SOH in real time.
There are two main categories of current state of health (SOH) estimation methods for batteries. One is a direct acquisition method, which is a method of processing the SOH obtained by directly measuring the battery parameters, specifically, a resistance method, an open circuit voltage method, and the like; the second method is a data-driven method, which is independent of a battery model and constructs a model by using measurement data of a power battery, specifically, a support vector machine, a BP neural network and the like.
Disclosure of Invention
An object of embodiments of the present invention is to provide a method, system and storage medium for predicting the state of health of a power battery, which are capable of accurately estimating the state of health of the power battery.
In order to achieve the above object, an embodiment of the present invention provides a method for predicting a state of health of a power battery, the method including:
building a first-order Thevenin model based on the working condition of the power battery;
carrying out parameter identification on the internal resistance of the first-order Thevenin model;
estimating the internal resistance of the power battery according to the result of parameter identification by adopting a Kalman filtering algorithm;
determining a first SOH of the power battery according to the internal resistance;
acquiring charging voltage data of the power battery;
inputting the charging voltage data into a trained random forest algorithm model to obtain a second SOH of the power battery;
and weighting the first SOH and the second SOH to obtain the SOH of the power battery.
Optionally, the first order Thevenin model comprises:
the negative pole of the voltage source is used for outputting the negative pole of the output voltage of the power battery;
one end of the first resistor is connected with the positive electrode of the voltage source;
one end of the capacitor is connected with one end of the first resistor, and the other end of the capacitor is connected with the other end of the first resistor; and
one end of the second resistor is connected with the other end of the first resistor, and the other end of the second resistor is used for outputting the positive electrode of the output voltage of the power battery;
the first-order Thevenin model is built based on the working condition of the power battery and specifically comprises the following steps:
establishing equation (1) according to the first-order Thevenin model,
Ut＝UOC-IPRP-I0R0， (1)
wherein ,UtFor said output voltage, UOCIs the voltage source, IPTo pass through a first resistor RPCurrent of (I)0To pass through a second resistor R0The current of (2).
Optionally, the performing parameter identification on the internal resistance of the first-order Thevenin model comprises:
and performing parameter identification by using a voltage characteristic curve obtained by pulse discharge at two ends of the power battery by adopting a direct current discharge method.
Optionally, the estimating, by using a kalman filter algorithm, the internal resistance of the power battery according to the result of the parameter identification specifically includes:
initializing the kalman filtering algorithm according to equation (2),
Figure BDA0002709535920000021
wherein ,
Figure BDA0002709535920000031
is an initial value, x, of the state variable estimate0Is a state variable, D0The initial value of the state error covariance is E, and the expected calculation operator is E;
the time update operation is performed according to the formula (3) and the formula (4),
Figure BDA0002709535920000032
wherein ,
Figure BDA0002709535920000033
is a predicted value of the state of the kth round, AkIs the state transition matrix for the k-th round,
Figure BDA0002709535920000034
is an estimated value of the state variable of the k-1 th round, BkIs the input matrix of the k-th round, uk-1Is the input variable of the k-1 th round;
Figure BDA0002709535920000035
wherein ,Pk|k-1Is a prediction value of the error covariance of the kth round, Ak-1Is the state transition matrix of the k-1 th round, Pk-1Is the state error covariance for round k-1,
Figure BDA0002709535920000036
is a state transition matrix Ak-1Q is the covariance of the process noise;
performing measurement update according to the formula (5) to the formula (7),
Figure BDA0002709535920000037
Figure BDA0002709535920000038
Pk＝(I-KkCk)Pk|k-1， (7)
wherein ,KkGain for the k-th wheel, CkIs the observation matrix of the k-th round,
Figure BDA0002709535920000039
for observing matrix CkR is the covariance of the measurement noise,
Figure BDA00027095359200000310
is the state variable estimate of the k-th round, ykIs the observed variable of the k-th round, DkFor a feed-forward matrix, ukIs an input variable of the k-th round, PkThe state error covariance of the kth round is shown, and I is an identity matrix;
determining an equation of state for the Kalman filtering algorithm according to equation (8),
Figure BDA00027095359200000311
wherein ,R0(k)、R0(k-1) second resistances R of the k-th round and the k-1 round respectively0Value of (A), Rp(k)、Rp(k-1) first resistances R of the k-th wheel and the k-1-th wheel respectivelypValue of (d), τp(k) Is tau of the k-th and k-1-th roundspValue of (A), Cp＝τp/Rp，CpIs the capacitance value of the capacitor, ω (k) is the process noise;
and (5) inputting the equation (8) into a Kalman filtering algorithm to obtain the internal resistance of the power battery.
Optionally, the determining the first SOH of the power battery according to the internal resistance includes:
estimating the first SOH according to equation (9),
Figure BDA0002709535920000041
wherein ,SOH1Is said first SOH, REOLOhmic internal resistance value, R, for the end of the life of the power cellnowIs the current internal resistance of the power battery, RnewAnd the ohmic internal resistance value of the power battery when the power battery leaves the factory is obtained.
Optionally, the method comprises:
acquiring cyclic charging voltage data of the power battery;
dividing each cyclic voltage interval in the cyclic charging voltage data into a plurality of voltage values;
respectively acquiring the capacity value of the corresponding power battery according to each voltage value;
for each cycle, combining the plurality of capacity values and the corresponding maximum capacity to form a single data sample;
combining a plurality of the single data samples to form a data set for training the random forest algorithm model.
Optionally, the method comprises:
dividing the data set by a preset probability value to form a training set and a test set;
training the random forest algorithm model by adopting the training set;
testing the random forest algorithm model by adopting the test set;
and calculating the accuracy of the random forest algorithm model by adopting a root mean square error method.
Optionally, the weighting the first SOH and the second SOH to obtain the SOH of the power battery specifically includes:
calculating the SOH of the power battery according to the formula (10),
SOHfruit of Chinese wolfberry＝x1SOH1+x2SOH2， (10)
wherein ,SOHFruit of Chinese wolfberryIs the SOH (true value), SOH, of the power cell1Is the first SOH, SOH2Is a second SOH, x1、x2Are respectively corresponding weighting coefficients and x1+x2＝1。
In another aspect, the present invention also provides a system for predicting the state of health of a power battery, the system comprising a processor configured to perform the method as described in any one of the above.
In yet another aspect, the present invention also provides a storage medium storing instructions for reading by a machine to cause the machine to perform a method as claimed in any one of the above.
According to the technical scheme, on one hand, the method, the system and the storage medium for predicting the health state of the power battery estimate the first SOH of the power battery by establishing a first-order Thevenin model and adopting an improved Kalman filtering method; on the other hand, a random forest algorithm model is adopted to estimate a second SOH of the power battery based on the charging voltage change condition of the power battery; and finally, estimating the true SOH value of the power battery by weighting the first SOH and the second SOH. The method overcomes the defect that the two methods in the prior art are easily influenced by the test conditions when being independently estimated, and improves the estimation precision of the SOH.
Additional features and advantages of embodiments of the invention will be set forth in the detailed description which follows.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the embodiments of the invention without limiting the embodiments of the invention. In the drawings:
FIG. 1 is a flow diagram of a method of predicting a power cell according to one embodiment of the present disclosure;
fig. 2 is a circuit diagram of a first order Thevenin model according to an embodiment of the present invention;
FIG. 3 is a graph of a first order Thevenin model based two terminal voltage according to an embodiment of the present invention;
FIG. 4 is a flow diagram of a method of calculating an internal resistance based on a Kalman filtering algorithm in accordance with an embodiment of the present invention;
FIG. 5 is a flow diagram of generating a data set according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of a workflow of a random forest algorithm model according to one embodiment of the invention; and
FIG. 7 is a block diagram of determining x according to an embodiment of the present invention1、x2A flow chart of the method of (1).
Detailed Description
The following detailed description of embodiments of the invention refers to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating embodiments of the invention, are given by way of illustration and explanation only, not limitation.
In the embodiments of the present invention, unless otherwise specified, the use of directional terms such as "upper, lower, top, and bottom" is generally used with respect to the orientation shown in the drawings or the positional relationship of the components with respect to each other in the vertical, or gravitational direction.
In addition, if there is a description of "first", "second", etc. in the embodiments of the present invention, the description of "first", "second", etc. is for descriptive purposes only and is not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In addition, technical solutions between the various embodiments can be combined with each other, but must be based on the realization of the technical solutions by a person skilled in the art, and when the technical solutions are contradictory or cannot be realized, the combination of the technical solutions should be considered to be absent and not to be within the protection scope of the present invention.
Fig. 1 is a flow chart illustrating a method for predicting the state of health of a power battery according to an embodiment of the invention. In fig. 1, the method may include:
in step S10, a first-order Thevenin model is constructed based on the operating conditions of the power battery. When designing the first-order Thevenin model, the inventor finds that the internal resistance of the power battery is gradually increased along with the increase of the working time, a voltage source which is originally approximate to 0 internal resistance is gradually changed into a circuit structure that the voltage source is connected with a resistor in series, and due to the polarization characteristic inside the power battery, an RC parallel circuit is equivalently connected between the resistor in series and the voltage source. Accordingly, the inventors have devised a circuit configuration as shown in fig. 2. In fig. 2, the first order Thevenin model may include a voltage source UOCA first resistor RpA second resistor R0And a capacitor Cp. Wherein the voltage source UOCMay be used as the negative electrode for outputting the output voltage of the power cell. A first resistor RpCan be connected to a voltage source UOCIs connected to the positive electrode. Capacitor CpMay be connected to the first resistor RpIs connected to a capacitor CpMay be connected to the first resistor RpThe other end of the connecting rod is connected. A second resistor R0May be connected to the first resistor RpIs connected to the other end of the first resistor R0The other end of the positive electrode may be used for a positive electrode that outputs the output voltage of the power cell. Then, based on the circuit in fig. 2, when building the first-order Thevenin model, equation (1) can be built based on the model,
Ut＝Uoc-IpRp-I0R0， (1)
wherein ,UtTo output a voltage, UOCIs a voltage source, IpTo pass through a first resistor RpCurrent of (I)0To pass through a second resistor R0The current of (2).
In step S11, parameter identification is performed on the internal resistance of the first-order Thevenin model. In this embodiment, the manner in which the parameter is identified may be in a variety of forms known to those skilled in the art. In a preferred example of the present invention, in consideration of the simplicity of operation, a dc discharge method may be used to identify parameters using a voltage characteristic curve obtained by pulse discharge across the power battery. Specifically, the power battery may be discharged first, and a voltage characteristic curve across the power battery is recorded in real time during the discharging process, where the voltage characteristic curve is shown in fig. 3.
Further, based on the voltage characteristic curve as shown in fig. 3, ab segment can be selected to establish equation (2),
Figure BDA0002709535920000071
wherein, the delta U is the voltage difference of the ab segment; then selecting bc section to obtain tau by least square methodp and RpI.e., equation (3),
Figure BDA0002709535920000081
wherein ,Cp＝τp/Rp。CpIs a polarization capacitance, i.e. the capacitance in fig. 2. RpIs the polarization resistance, i.e. the first resistance in fig. 2.
In step S12, a kalman filter algorithm is used to estimate the internal resistance of the power battery according to the result of the parameter identification. Specifically, the process of estimating the internal resistance of the power cell may include steps as shown in fig. 4. In fig. 4, the process may include:
in step S20, the kalman filter algorithm is initialized according to equation (2),
Figure BDA0002709535920000082
wherein ,
Figure BDA0002709535920000083
is an initial value, x, of the state variable estimate0Is a state variable, D0The initial value of the state error covariance, and E is the desired calculation operator.
In step S21, a time update operation is performed according to equation (3) and equation (4),
Figure BDA0002709535920000084
wherein ,
Figure BDA0002709535920000085
is a predicted value of the state of the kth round, AkIs the state transition matrix for the k-th round,
Figure BDA0002709535920000086
is an estimated value of the state variable of the k-1 th round, BkIs the input matrix of the k-th round, uk-1Is the input variable of the k-1 th round;
Figure BDA0002709535920000087
wherein ,Pk|k-1Is a prediction value of the error covariance of the kth round, Ak-1Is the state transition matrix of the k-1 th round, Pk-1Is the state error covariance for round k-1,
Figure BDA0002709535920000088
is a state transition matrix Ak-1Q is the covariance of the process noise;
in step S22, the measurement is updated according to the formula (5) to the formula (7),
Figure BDA0002709535920000089
Figure BDA00027095359200000810
Pk＝(I-KkCk)Pk|k-1， (7)
wherein ,KkGain for the k-th wheel, CkIs the observation matrix of the k-th round,
Figure BDA00027095359200000811
for observing matrix CkR is the covariance of the measurement noise,
Figure BDA0002709535920000091
is the state variable estimate of the k-th round, ykIs the observed variable of the k-th round, DkFor a feed-forward matrix, ukIs an input variable of the k-th round, PkAnd I is the state error covariance of the kth round and is an identity matrix.
In step S23, the state equation of the kalman filter algorithm is determined according to equation (8),
Figure BDA0002709535920000092
wherein ,R0(k)、R0(k-1) second resistances R of the k-th round and the k-1 round respectively0Value of (A), Rp(k)、Rp(k-1) first resistances R of the k-th wheel and the k-1-th wheel respectivelypValue of (d), τp(k) Is tau of the k-th and k-1-th roundspValue of (A), Cp＝τp/Rp，CpIs the capacitance value of the capacitor and ω (k) is the process noise.
In step S23, the current flows through the first resistor RpCurrent of (I)pCan be expressed by the formula (9),
Figure BDA0002709535920000093
wherein ,UpIs a first resistor RpThe voltage across;
by substituting this equation (9) into equation (1), equation (10) shown below can be obtained,
Figure BDA0002709535920000094
wherein, SOC is the SOC value of the power battery, and t represents the current time;
discretizing the equation (9) to obtain equation (11),
Figure BDA0002709535920000095
wherein ,Ut(k)、Ut(k-1) output voltages U of the k-th and k-1-th roundstValue of (A), Uoc(k)、Uoc(k-1) voltage source U for the k-th wheel and the k-1-th wheel respectivelyocValue of (A), I0(k)、I0(k-1) current I of the k-th and k-1-th rounds0T is the sampling time (period), v (k) is the measurement noise;
in addition, the inventors have considered that the model parameters (SOC, maximum capacity, etc.) of the power battery can be regarded as constant in a short time, and thus can obtain equation (8).
In step S24, equation (8) is input into the kalman filtering algorithm to obtain the internal resistance of the power cell.
In step S13, a first SOH of the power battery is determined according to the internal resistance. Specifically, the first SOH may be estimated according to equation (12),
Figure BDA0002709535920000101
wherein ,SOH1Is a first SOH, REOLOhmic internal resistance value, R, for the end of the life of a power cellnowIs the internal resistance of the current power battery, RnewThe ohmic internal resistance value of the power battery when the power battery leaves the factory is obtained.
In step S14, charging voltage data of the power battery is acquired.
In step S15, the charging voltage data is input into the trained random forest algorithm model to obtain a second SOH of the power battery. In this embodiment, the random forest algorithm model may be trained from a preset data set, and the training process may specifically include the steps as shown in fig. 5. In fig. 5, the process may include:
in step S30, the cyclic charge voltage data of the power battery is acquired.
In step S31, each cyclic voltage interval in the cyclic charging voltage data is divided into a plurality of voltage values.
In step S32, the capacity value of the corresponding power battery is acquired for each voltage value.
In step S33, for each cycle, the plurality of capacity values and the corresponding maximum capacity are combined to constitute a single data sample.
In step S34, a plurality of individual data samples are combined to form a data set for training a random forest algorithm model.
In steps S30 to S34, the cyclic charging voltage data may be the variation curve of the voltages at two ends of the power batteries with different maximum capacities during the charging process. And the cycle voltage interval can be an interval capable of reflecting the voltage change characteristics of the two ends of the power battery in the charging process. In this embodiment, the cyclic voltage interval may be first divided into a plurality of voltage values Δ V. The voltage value Δ V may be a point value or a subinterval determined by two point values. Then, for each voltage value Δ V, the current capacity of the corresponding power battery is determined, which can be denoted as Q1、Q2……Qm. The current capacity is then combined with the maximum capacity of the power battery to form the data set, denoted Sn＝{(X1， Y1)，(X2,Y2)，……，(Xn,Yn)}. Wherein, X1、X2…XnRepresents Q of one power battery1、Q2……Qm，Y1、Y2…YnIndicating the corresponding maximum capacity.
After the data set is determined, the data set may be divided by a preset probability value (e.g., 7:3) to form a training set and a testing set, and then the training set is used to train the random forest algorithm model, and then the testing set is used to test the random forest algorithm model. And finally, calculating the accuracy of the random forest algorithm model by adopting a root mean square error method. And determining that the training of the random forest algorithm model is finished under the condition that the accuracy of the random forest algorithm model reaches a preset condition.
In addition, in the process of the random forest algorithm, S is firstly selected fromnRandomly selecting n observed values, replacing to obtain Bootstrap samples, wherein each observed value has the probability of 1/n being selected, and constructing q decision trees, wherein the Bootstrap samples are
Figure BDA0002709535920000111
In the training process, the algorithm divides the input data on each node, and simultaneously, all variables can be optimally divided, the division process starts from a root node, and each node applies a corresponding division function to a new input X. This process is repeated recursively until the last node cannot be resegmented. When the maximum node depth is reached; or when the nodes contain fewer than a predetermined number of observations, the tree stops growing. At the end of this training process, at SnConstructing a prediction function
Figure BDA0002709535920000112
The output value corresponding to each tree is respectively
Figure BDA0002709535920000113
Figure BDA0002709535920000114
Finally by averaging the outputs of all trees, i.e.:
Figure BDA0002709535920000115
wherein
Figure BDA0002709535920000116
Is the output value of the corresponding ith tree, i ═ 1,2, … …, q. The corresponding flow chart is shown in fig. 6.
In step S16, the first SOH and the second SOH are weighted to obtain the SOH of the power battery. Specifically, the weighting operation may be performed according to equation (13),
SOHfruit of Chinese wolfberry＝x1SOH1+x2SOH2， (13)
wherein ,SOHFruit of Chinese wolfberryIs the SOH (true value), SOH, of the power cell1Is the first SOH, SOH2Is a second SOH, x1、x2Are respectively corresponding weighting coefficients and x1+x2＝1。
In addition, for the x1、x2The determination of a particular value can be by a variety of methods known to those skilled in the art. However, consider if x is paired1、x2An improper determination of the value may result in an inaccurate determination of the first SOH and the second SOH, which in turn may result in an SOHFruit of Chinese wolfberryThe accuracy of (2) is reduced. Thus, in one embodiment of the present invention, the inventors devised a method as shown in FIG. 7. In FIG. 7, x is determined1、x2The specific process of numerical value may include:
in step S40, x is defined1、x2An initial value of (d);
in step S41, x is added1、x2Substituted into equation (13) to obtain SOHFruit of Chinese wolfberryA value of (d);
in step S42, the SOH is calculated based on the power battery capacity aging curve and the internal resistance aging curveFruit of Chinese wolfberryAverage errors respectively with a capacity aging curve and an internal resistance aging curve of the power battery;
in step S43, the sum of the two average errors is calculated;
in step S44, it is determined whether the sum is less than or equal to a preset threshold;
in step S45, in the case where it is judged that the sum is larger than the threshold value, x is updated according to formula (14) and formula (15)1 and x2And returns to the execution of step S41,
x1 1＝x1+Δt， (14)
x2 1＝x2-Δt， (15)
wherein ,x1 1、x2 1Is updated x1、x2Δ t is a preset recurrence factor;
in step S46, in the case where it is judged that the sum is less than or equal to the threshold value, the current x is determined1 and x2Is the optimal solution.
By the steps as shown in FIG. 7, for x1、x2Performing increment and decrement respectively, and calculating the sum of average errors of the capacity aging curve and the internal resistance aging curve of the power battery in real time during the increment and decrement processes so as to be based on x1、x2Resulting SOHFruit of Chinese wolfberryAnd is more accurate.
Furthermore, it is noted that the sum of the average errors is expressed as SOH for the actual operating conditionsFruit of Chinese wolfberryAnd any other curve of the power battery. However, in the embodiment of the present invention, the inventor considers that the SOH value of the power battery is mainly affected by the maximum capacity and the internal resistance of the power battery, and therefore, the reasonable x can be more accurately determined by using the power battery capacity aging curve and the internal resistance aging curve of the power battery as evaluation criteria1、x2To ultimately yield a more accurate true value of SOH.
In another aspect, the present invention also provides a system for predicting the state of health of a power battery, which may include a processor configured to perform any of the methods described above.
In yet another aspect, the present invention also provides a storage medium which may store instructions which are readable by a machine to cause the machine to perform a method as described in any one of the above.
According to the technical scheme, on one hand, the method, the system and the storage medium for predicting the health state of the power battery estimate the first SOH of the power battery by establishing a first-order Thevenin model and adopting an improved Kalman filtering method; on the other hand, a random forest algorithm model is adopted to estimate a second SOH of the power battery based on the charging voltage change condition of the power battery; and finally, estimating the true SOH value of the power battery by weighting the first SOH and the second SOH. The method overcomes the defect that the two methods in the prior art are easily influenced by the test conditions when being independently estimated, and improves the estimation precision of the SOH.
Although the embodiments of the present invention have been described in detail with reference to the accompanying drawings, the embodiments of the present invention are not limited to the details of the above embodiments, and various simple modifications can be made to the technical solutions of the embodiments of the present invention within the technical concept of the embodiments of the present invention, and the simple modifications all fall within the scope of the embodiments of the present invention.
It should be noted that the various features described in the above embodiments may be combined in any suitable manner without departing from the scope of the invention. In order to avoid unnecessary repetition, the embodiments of the present invention will not be described separately for the various possible combinations.
Those skilled in the art can understand that all or part of the steps in the method for implementing the above embodiments may be implemented by a program to instruct related hardware, where the program is stored in a storage medium and includes several instructions to enable a (may be a single chip, a chip, etc.) or a processor (processor) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
In addition, various different embodiments of the present invention may be arbitrarily combined with each other, and the embodiments of the present invention should be considered as disclosed in the disclosure of the embodiments of the present invention as long as the idea of the embodiments of the present invention is not violated.

Claims (10)

1. A method of predicting a state of health of a power cell, the method comprising:
building a first-order Thevenin model based on the working condition of the power battery;
carrying out parameter identification on the internal resistance of the first-order Thevenin model;
estimating the internal resistance of the power battery according to the result of parameter identification by adopting a Kalman filtering algorithm;
determining a first SOH of the power battery according to the internal resistance;
acquiring charging voltage data of the power battery;
inputting the charging voltage data into a trained random forest algorithm model to obtain a second SOH of the power battery;
and weighting the first SOH and the second SOH to obtain the SOH of the power battery.
2. The method of claim 1, wherein the first order Thevenin model comprises:
the negative pole of the voltage source is used for outputting the negative pole of the output voltage of the power battery;
one end of the first resistor is connected with the positive electrode of the voltage source;
one end of the capacitor is connected with one end of the first resistor, and the other end of the capacitor is connected with the other end of the first resistor; and
one end of the second resistor is connected with the other end of the first resistor, and the other end of the second resistor is used for outputting the positive electrode of the output voltage of the power battery;
the first-order Thevenin model is built based on the working condition of the power battery and specifically comprises the following steps:
establishing equation (1) according to the first-order Thevenin model,
Ut＝UOC-IPRP-I0R0，(1)
wherein ,UtFor said output voltage, UOCIs the voltage source, IPTo pass through a first resistor RPCurrent of (I)0To pass through a second resistor R0The current of (2).
3. The method of claim 1, wherein the parametrically identifying the internal resistance of the first-order Thevenin model comprises:
and performing parameter identification by using a voltage characteristic curve obtained by pulse discharge at two ends of the power battery by adopting a direct current discharge method.
4. The method according to claim 2, wherein the estimating the internal resistance of the power battery according to the result of the parameter identification by using the kalman filter algorithm specifically comprises:
initializing the kalman filtering algorithm according to equation (2),
Figure FDA0002709535910000021
wherein ,
Figure FDA0002709535910000022
is an initial value, x, of the state variable estimate0Is a state variable, D0An initial value of the covariance of the state errors, and E is an expected calculation operator;
the time update operation is performed according to the formula (3) and the formula (4),
Figure FDA0002709535910000023
wherein ,
Figure FDA0002709535910000024
is a predicted value of the state of the kth round, AkIs the state transition matrix for the k-th round,
Figure FDA0002709535910000025
is an estimated value of the state variable of the k-1 th round, BkIs as followsInput matrix of k rounds, uk-1Is the input variable of the k-1 th round;
Figure FDA0002709535910000026
wherein ,Pk|k-1Is a prediction value of the error covariance of the kth round, Ak-1Is the state transition matrix of the k-1 th round, Pk-1Is the state error covariance for round k-1,
Figure FDA0002709535910000027
is a state transition matrix Ak-1Q is the covariance of the process noise;
performing measurement update according to the formula (5) to the formula (7),
Figure FDA0002709535910000028
Figure FDA0002709535910000029
Pk＝(I-KkCk)Pk|k-1，(7)
wherein ,KkGain for the k-th wheel, CkIs the observation matrix of the k-th round,
Figure FDA00027095359100000210
for observing matrix CkR is the covariance of the measurement noise,
Figure FDA0002709535910000031
is the state variable estimate of the k-th round, ykAs observed variable of the k-th round, DkFor a feed-forward matrix, ukIs an input variable of the k-th round, PkThe state error covariance of the kth round is shown, and I is an identity matrix;
determining an equation of state for the Kalman filtering algorithm according to equation (8),
Figure FDA0002709535910000032
wherein ,R0(k)、R0(k-1) second resistances R of the k-th round and the k-1 round respectively0Value of (A), Rp(k)、Rp(k-1) first resistances R of the k-th wheel and the k-1-th wheel respectivelypValue of (d), τp(k) Is tau of the k-th and k-1-th roundspValue of (A), Cp＝τp/Rp，CpIs the capacitance value of the capacitor, ω (k) is the process noise;
and (5) inputting the equation (8) into a Kalman filtering algorithm to obtain the internal resistance of the power battery.
5. The method of claim 1, wherein the determining the first SOH of the power cell based on the internal resistance comprises:
estimating the first SOH according to equation (9),
Figure FDA0002709535910000033
wherein ,SOH1Is said first SOH, REOLOhmic internal resistance value, R, for the end of the life of the power cellnowIs the current internal resistance of the power battery, RnewAnd the ohmic internal resistance value of the power battery when the power battery leaves the factory is obtained.
6. The method according to claim 1, characterized in that it comprises:
acquiring cyclic charging voltage data of the power battery;
dividing each cyclic voltage interval in the cyclic charging voltage data into a plurality of voltage values;
respectively acquiring the capacity value of the corresponding power battery according to each voltage value;
for each cycle, combining the plurality of capacity values and the corresponding maximum capacity to form a single data sample;
combining a plurality of the single data samples to form a data set for training the random forest algorithm model.
7. The method of claim 6, wherein the method comprises:
dividing the data set by a preset probability value to form a training set and a test set;
training the random forest algorithm model by adopting the training set;
testing the random forest algorithm model by adopting the test set;
and calculating the accuracy of the random forest algorithm model by adopting a root mean square error method.
8. The method according to claim 1, wherein the weighting the first SOH and the second SOH to obtain the SOH of the power battery specifically comprises:
calculating the SOH of the power battery according to the formula (10),
SOHfruit of Chinese wolfberry＝x1SOH1+x2SOH2，(10)
wherein ,SOHFruit of Chinese wolfberryIs the SOH (true value), SOH, of the power cell1Is the first SOH, SOH2Is a second SOH, x1、x2Are respectively corresponding weighting coefficients and x1+x2＝1。
9. A system for predicting a state of health of a power battery, the system comprising a processor configured to perform the method of any of claims 1 to 8.
10. A storage medium storing instructions for reading by a machine to cause the machine to perform a method according to any one of claims 1 to 8.
CN202011050925.9A 2020-09-29 2020-09-29 Method, system and storage medium for predicting power battery health status Active CN112379295B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011050925.9A CN112379295B (en) 2020-09-29 2020-09-29 Method, system and storage medium for predicting power battery health status

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011050925.9A CN112379295B (en) 2020-09-29 2020-09-29 Method, system and storage medium for predicting power battery health status

Publications (2)

Publication Number Publication Date
CN112379295A true CN112379295A (en) 2021-02-19
CN112379295B CN112379295B (en) 2023-08-04

Family

ID=74580609

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011050925.9A Active CN112379295B (en) 2020-09-29 2020-09-29 Method, system and storage medium for predicting power battery health status

Country Status (1)

Country Link
CN (1) CN112379295B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114035049A (en) * 2021-11-08 2022-02-11 东软睿驰汽车技术（沈阳）有限公司 SOH precision calculation method and device and electronic equipment

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2018096953A (en) * 2016-12-16 2018-06-21 三菱自動車工業株式会社 Battery state estimation device
CN108846227A (en) * 2017-12-05 2018-11-20 北京航空航天大学 A kind of capacity of lithium ion battery degradation prediction appraisal procedure based on random forest and capacity self- recoverage effect analysis
CN108896926A (en) * 2018-07-18 2018-11-27 湖南宏迅亿安新能源科技有限公司 A kind of appraisal procedure, assessment system and the associated component of lithium battery health status
CN110888058A (en) * 2019-12-02 2020-03-17 西安科技大学 Algorithm based on power battery SOC and SOH joint estimation
CN111044928A (en) * 2019-12-31 2020-04-21 福州大学 Lithium battery health state estimation method
CN111060834A (en) * 2019-12-19 2020-04-24 中国汽车技术研究中心有限公司 Power battery state of health estimation method
CN111337832A (en) * 2019-12-30 2020-06-26 南京航空航天大学 Power battery multidimensional fusion SOC and SOH online joint estimation method

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2018096953A (en) * 2016-12-16 2018-06-21 三菱自動車工業株式会社 Battery state estimation device
CN108846227A (en) * 2017-12-05 2018-11-20 北京航空航天大学 A kind of capacity of lithium ion battery degradation prediction appraisal procedure based on random forest and capacity self- recoverage effect analysis
CN108896926A (en) * 2018-07-18 2018-11-27 湖南宏迅亿安新能源科技有限公司 A kind of appraisal procedure, assessment system and the associated component of lithium battery health status
CN110888058A (en) * 2019-12-02 2020-03-17 西安科技大学 Algorithm based on power battery SOC and SOH joint estimation
CN111060834A (en) * 2019-12-19 2020-04-24 中国汽车技术研究中心有限公司 Power battery state of health estimation method
CN111337832A (en) * 2019-12-30 2020-06-26 南京航空航天大学 Power battery multidimensional fusion SOC and SOH online joint estimation method
CN111044928A (en) * 2019-12-31 2020-04-21 福州大学 Lithium battery health state estimation method

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114035049A (en) * 2021-11-08 2022-02-11 东软睿驰汽车技术（沈阳）有限公司 SOH precision calculation method and device and electronic equipment

Also Published As

Publication number Publication date
CN112379295B (en) 2023-08-04

Similar Documents

Publication Publication Date Title
CN108896914B (en) Gradient lifting tree modeling and prediction method for health condition of lithium battery
CN111007400A (en) Lithium battery SOC estimation method based on self-adaptive double-extended Kalman filtering method
CN112034349B (en) Lithium battery health state online estimation method
CN107748336B (en) On-line estimation method and system for state of charge of lithium ion battery
CN108490365B (en) Method for estimating residual life of power battery of electric automobile
JP2017538931A (en) Method for automatically estimating the capacitance of a battery cell
CN108490356A (en) A kind of lithium battery SOC estimation method for improving EKF algorithms
CN110058160B (en) Lithium battery health state prediction method based on square root extended Kalman filtering
JP2017539060A (en) Method for automatically estimating the state of charge of a battery cell
CN109839596B (en) SOC estimation method based on UD decomposition and adaptive extended Kalman filtering
CN110196395B (en) Storage battery SOC estimation method
JP2017538932A (en) Method for automatically estimating the state of charge of a battery cell
JP2017538929A (en) Method for automatically estimating the state of charge of a battery cell
CN112904208A (en) Method and system for estimating open circuit voltage of battery cell
CN112946487B (en) Parameter identification method and device, storage medium and computer equipment
CN113138344B (en) SOC estimation method based on fractional order extended Kalman filtering algorithm
CN111142025A (en) Battery SOC estimation method and device, storage medium and electric vehicle
CN113777510A (en) Lithium battery state of charge estimation method and device
CN115754741A (en) SOC estimation method and device of energy storage system, equipment and storage medium
CN114114038A (en) Lithium battery SOC and available capacity joint estimation method under full-life and full-temperature conditions
CN110412472B (en) Battery state of charge estimation method based on normal gamma filtering
CN112379295B (en) Method, system and storage medium for predicting power battery health status
CN113466725A (en) Method and device for determining state of charge of battery, storage medium and electronic equipment
CN114779089A (en) Method for calculating battery state of charge based on energy storage lithium battery equivalent circuit model
CN114091282A (en) Lithium ion battery state estimation method and system based on fractional order model

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant