CN110068774B - Lithium battery health state estimation method and device and storage medium - Google Patents
Lithium battery health state estimation method and device and storage medium Download PDFInfo
- Publication number
- CN110068774B CN110068774B CN201910371690.4A CN201910371690A CN110068774B CN 110068774 B CN110068774 B CN 110068774B CN 201910371690 A CN201910371690 A CN 201910371690A CN 110068774 B CN110068774 B CN 110068774B
- Authority
- CN
- China
- Prior art keywords
- data
- health
- lithium battery
- soh
- cycles
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 118
- 230000036541 health Effects 0.000 title claims abstract description 113
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 title claims abstract description 85
- 229910052744 lithium Inorganic materials 0.000 title claims abstract description 85
- 238000003860 storage Methods 0.000 title claims abstract description 11
- 238000007600 charging Methods 0.000 claims abstract description 147
- 238000012549 training Methods 0.000 claims abstract description 43
- 230000008569 process Effects 0.000 claims abstract description 42
- 238000001514 detection method Methods 0.000 claims abstract description 9
- 238000010801 machine learning Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 13
- 238000010219 correlation analysis Methods 0.000 claims description 11
- 238000001914 filtration Methods 0.000 claims description 11
- 238000009499 grossing Methods 0.000 claims description 9
- 230000002159 abnormal effect Effects 0.000 claims description 7
- 238000012512 characterization method Methods 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 11
- 239000002245 particle Substances 0.000 description 21
- 238000005070 sampling Methods 0.000 description 16
- 230000006870 function Effects 0.000 description 15
- 238000005457 optimization Methods 0.000 description 10
- 238000004458 analytical method Methods 0.000 description 8
- 238000010277 constant-current charging Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 8
- 238000012360 testing method Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 238000009825 accumulation Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000005034 decoration Methods 0.000 description 2
- 239000006185 dispersion Substances 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 229910001416 lithium ion Inorganic materials 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 238000005562 fading Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000011164 primary particle Substances 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 235000021419 vinegar Nutrition 0.000 description 1
- 239000000052 vinegar Substances 0.000 description 1
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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
Abstract
The invention discloses a method and a device for estimating the health state of a lithium battery and a storage medium. Wherein, the method comprises the following steps: training by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to a plurality of effective charging cycles as output variables to obtain a Gaussian process regression GPR model, wherein the GPR model is obtained by using a plurality of groups of data through machine learning training, and each group of data in the plurality of groups of data comprises: HI for an active charge cycle and SOH for HI; and acquiring HI of the charging cycle to be detected, inputting the HI into a GPR model, and outputting SOH corresponding to the HI by the GPR model. The invention solves the technical problems that the detection of the health state of the lithium battery is complex and the lithium battery is difficult to adapt to the collected data with poor quality in the related technology.
Description
Technical Field
The invention relates to the field of batteries, in particular to a method and a device for estimating the health state of a lithium battery and a storage medium.
Background
At present, lithium batteries are widely applied in the fields of Electric Vehicles (EVs), Hybrid Electric Vehicles (HEVs) and power grid energy storage, and estimation of State of Health (SOH) is a key technology of interest in the use process of lithium batteries.
The SOH estimation method mainly includes two types: model-based and data-based driving. The model-based method mainly includes an aging empirical model, an equivalent circuit model and an electrochemical model. However, the aging experience model requires a lot of manpower, time and resources, and in the background of the rapid development of the lithium battery manufacturing technology today, the experience model established for the old-model battery may lose timeliness; the equivalent circuit model represents the health state through model element parameters, but the model is mostly only suitable for batteries of the same type of a certain brand; the electrochemical model has the problem of high computational complexity. Therefore, the data-driven method is very suitable for the use in the actual operation stage of the lithium battery.
The data-driven approach is more flexible than the model-based approach and does not take external disturbances into account. However, most of the existing researches are based on high-precision and complete experimental cycles, and the researches suitable for actual operation data with poor precision and incomplete charging are still few at present. The original data quality is poor due to the fact that the current battery acquisition system is low in precision and limited in data uploading capacity. The Capacity increment Analysis (ICA) can extract detailed features in a charging curve by combining a certain data preprocessing technology to obtain a Health Indicator (HI), so that errors caused by data quality problems are reduced.
In view of the above problems, no effective solution has been proposed.
Disclosure of Invention
The embodiment of the invention provides a method and a device for estimating the health state of a lithium battery and a storage medium, which are used for at least solving the technical problems that the detection of the health state of the lithium battery is complex and the acquired data with poor quality is difficult to adapt in the related technology.
According to an aspect of an embodiment of the present invention, there is provided a method for estimating a state of health of a lithium battery, including: training by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to the plurality of effective charging cycles as output variables to obtain a Gaussian process regression GPR model, wherein the GPR model is obtained by using a plurality of groups of data through machine learning training, and each group of data in the plurality of groups of data comprises: an HI of an active charge cycle and a SOH corresponding to the HI; acquiring HI of a charging cycle to be detected, inputting the HI into the GPR model, and outputting SOH corresponding to the HI by the GPR model.
Optionally, before training the gaussian process regression GPR model using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to the plurality of effective charging cycles as output variables, the method includes: collecting lithium battery operation data; and determining a plurality of effective charging cycles according to the lithium battery operation data, wherein the effective charging cycles are charging cycles for acquiring the covering charge state of the charging process and reaching a preset charge state threshold value.
Optionally, before training the gaussian process regression GPR model using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to the plurality of effective charging cycles as output variables, the method further includes: according to the characteristic data of the capacity increment IC curves corresponding to the effective charging cycles, combining the characteristic data to obtain a plurality of health factors HI; a plurality of states of health SOH obtained from the plurality of active charging cycles.
Optionally, before combining the characteristic data of the capacity increment IC curves corresponding to the plurality of effective charging cycles to obtain a plurality of health factors HI as input variables of a GPR model, the method includes: performing capacity increment processing on the plurality of effective charging cycles to obtain corresponding IC curves; extracting feature data of the IC curve according to the IC curve, wherein the feature data comprises at least one of the following data: peak, peak area, and voltage at which the peak is located.
Optionally, the step of combining, according to feature data of capacity increment IC curves corresponding to the plurality of effective charging cycles, the feature data to obtain a plurality of health factors HI as input variables of a GPR model includes: normalizing the characteristic data of the IC curve to obtain normalized characteristic data; calculating the grey correlation degree of the normalized feature data and the SOH by utilizing a grey correlation analysis (GRA); and combining according to the grey correlation degrees corresponding to the normalized feature data to obtain the plurality of HIs.
Optionally, determining a plurality of effective charging cycles according to the lithium battery operation data includes: preprocessing the lithium battery operation data to obtain optimized operation data, wherein the preprocessing at least comprises one of the following steps: removing abnormal values, filling missing values and smoothing filtering; determining the plurality of effective charging cycles based on the optimized operational data.
Optionally, determining a plurality of effective charging cycles according to the lithium battery operation data includes: determining a zone bit of the lithium battery operation data; intercepting the charging cycle of the lithium battery operation data according to the zone bit; acquiring a data set of the charging cycle; judging whether the data set of the charging cycle meets a preset threshold value, wherein the preset threshold value comprises a data acquisition quantity threshold value and a data acquisition duration threshold value; under the condition that the data set meets the preset threshold value, the charging cycle corresponding to the data set is an effective charging cycle; otherwise, the charging cycle corresponding to the data set is an invalid charging cycle.
Optionally, calculating the gray correlation degree of the normalized feature data and the SOH by using a gray correlation analysis GRA comprises:
Wherein ρ ∈ [0, 1] is a resolution coefficient, the smaller ρ is, the higher the resolution is, i is 1, 2, 3.. m, i is the ith x compared with y, k is time, x is SOH, and y is normalized feature data;
obtaining the plurality of HIs by combining the gray correlation degrees corresponding to the normalized feature data in the following manner includes:
wherein, FiFor the ith feature data, r is the grey correlation.
According to another aspect of the embodiments of the present invention, there is also provided an apparatus for estimating a state of health of a lithium battery, including: the training module is used for training by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to the plurality of effective charging cycles as output variables to obtain a Gaussian process regression GPR model, wherein the GPR model is obtained by using a plurality of groups of data through machine learning training, and each group of data in the plurality of groups of data comprises: an HI of an active charge cycle and a SOH corresponding to the HI; and the detection module is used for acquiring HI of the charging cycle to be detected, inputting the HI into the GPR model and outputting SOH corresponding to the HI by the GPR model.
According to another aspect of the embodiments of the present invention, there is also provided a storage medium storing a program, wherein the program, when executed by a processor, causes the processor to execute the method for estimating the state of health of a lithium battery according to any one of the above.
In the embodiment of the present invention, a plurality of health factors HI corresponding to a plurality of effective charging cycles are used as input variables, and a plurality of health states SOH corresponding to the plurality of effective charging cycles are used as output variables, and training is performed to obtain a gaussian process regression GPR model, where the GPR model is obtained by using a plurality of sets of data through machine learning training, and each set of data in the plurality of sets of data includes: an HI of an active charge cycle and a SOH corresponding to the HI; the method comprises the steps of obtaining HI of a charging cycle to be detected, inputting the HI into a GPR model, outputting SOH corresponding to the HI by the GPR model, detecting the HI of the charging cycle through the GPR model, achieving the purpose of obtaining the SOH corresponding to the HI according to the HI of the charging cycle, improving the technical effect of data adaptability while reducing the complexity of detecting the health state of the lithium battery, and further solving the technical problems that the detection of the health state of the lithium battery is complex and the data acquisition with poor quality is difficult to adapt in the related technology.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention. In the drawings:
Fig. 1 is a flowchart of a method of estimating a state of health of a lithium battery according to an embodiment of the present invention;
FIG. 2 is a flow chart of a method of estimating the state of health of a lithium battery according to a preferred embodiment of the present invention;
fig. 3 is a flowchart of step S1 of the method for estimating the state of health of a lithium battery according to the preferred embodiment of the present invention;
fig. 4 is a flowchart of step S2 of the method for estimating the state of health of a lithium battery according to the preferred embodiment of the present invention;
FIG. 5 is a schematic diagram of an IC curve in a method for estimating the state of health of a lithium battery according to a preferred embodiment of the present invention;
FIG. 6 is a diagram illustrating the graph noise processing effect and feature extraction of the IC curve in the method for estimating the state of health of a lithium battery according to the preferred embodiment of the present invention;
fig. 7 is a flowchart of step S3 of the method for estimating the state of health of a lithium battery according to the preferred embodiment of the present invention;
FIG. 8 is a flow chart of the multi-population group optimization algorithm in step S32 of the method for estimating the state of health of lithium batteries according to the preferred embodiment of the present invention;
fig. 9 is a schematic diagram of the SOH estimation result in the lithium battery state of health estimation method according to the preferred embodiment of the present invention;
fig. 10 is a schematic structural diagram of a lithium battery state of health estimation apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
According to an embodiment of the present invention, there is provided an embodiment of a method for estimating a state of health of a lithium battery, where it is noted that the steps illustrated in the flowchart of the drawings may be executed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowchart, in some cases, the steps illustrated or described may be executed in an order different from that described herein.
Fig. 1 is a flowchart of a method for estimating a state of health of a lithium battery according to an embodiment of the present invention, as shown in fig. 1, the method including the steps of:
step S102, training by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to a plurality of effective charging cycles as output variables, to obtain a Gaussian Process Regression (GPR) model, where the GPR model is obtained by using a plurality of sets of data through machine learning training, and each set of data in the plurality of sets of data includes: HI for an active charge cycle and SOH for HI;
and step S104, acquiring HI of the charging cycle to be detected, inputting the HI into a GPR model, and outputting SOH corresponding to the HI by the GPR model.
Through the steps, the purpose that the SOH corresponding to the HI is accurately obtained according to the HI of the charging cycle through the GPR model can be achieved, the technical effect of improving the adaptability to data is improved while the complexity of detecting the health state of the lithium battery is reduced, and the technical problems that the health state of the lithium battery is complex and the acquired data with poor quality are difficult to adapt in the related technology are solved.
As an optional embodiment, the plurality of health factor HI and the plurality of health state SOH may respectively generate an HI sequence and a SOH sequence according to a preset rule, and further, the HI sequence may be used as an input variable, and the SOH sequence may be used as an output variable. It should be noted that the health factor HI in the HI sequence and the health state SOH in the SOH sequence have a corresponding relationship, that is, each health factor HI and the health state SOH exist correspondingly.
As an alternative embodiment, the above-mentioned gaussian process regresses the GPR model, wherein the model is obtained by selecting an appropriate kernel function and training the constructed GPR model, it should be noted that the optimal hyperparameter and the optimal GPR model can be obtained in actual training. For example, a plurality of HIs are used as model input, a plurality of SOHs are used as model output to obtain a data set (X, Y), and the data set is divided into a training set (X, Y) and a testing set (X) according to a certain proportion *，y*). A Radial Basis Function (RBF) is selected as a kernel Function of the GPR model, and the model training is carried out by using a training set. Optimizing by maximizing marginal likelihood function in training processAnd the optimization process adopts a multi-population particle swarm optimization algorithm, then an optimal GPR model is determined according to the optimal RBF kernel function, and meanwhile, the test set can be used for carrying out optimal GPR precision verification, so that the optimal superparameter and the optimal GPR model are more accurate.
As an alternative embodiment, the corresponding SOH can be determined quickly and accurately by the GPR model based on the input HI.
Optionally, before training the gaussian process regression GPR model using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to a plurality of effective charging cycles as output variables, the method includes: collecting lithium battery operation data; and determining a plurality of effective charging cycles according to the operation data of the lithium battery, wherein the effective charging cycles are charging cycles for acquiring the covering charge state of the charging process to reach a preset charge state threshold value.
As an optional embodiment, the operation data of the lithium battery may include basic information such as a battery model, a rated capacity, factory information, and an application scenario.
As an alternative embodiment, the effective charging cycle is used to obtain a charging cycle with a charging process covering a wider State of Charge (SOC), and in actual operation, the number of sampling points or the charging duration is selected according to different charging rates.
Optionally, before training the gaussian process regression GPR model using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to a plurality of effective charging cycles as output variables, the method further includes: combining the characteristic data to obtain a plurality of health factors HI according to the characteristic data of the capacity increment IC curves corresponding to the plurality of effective charging cycles; a plurality of states of health SOH obtained from a plurality of active charging cycles.
As an alternative embodiment, for the feature data of the incremental IC curve corresponding to the incomplete charging cycle, the partial available feature data may be combined to obtain the health factor.
As an alternative embodiment, a capacity increment analysis method may be adopted to calculate each charging cycle, extract the relevant features of the capacity increment IC curve, and combine the features to obtain the indirect health factor.
As an alternative embodiment, after determining the effective charge cycle, the state of health may be obtained from the effective charge cycle.
As an alternative embodiment, not only the plurality of SOHs and the plurality of HIs corresponding to the plurality of effective charging cycles, but also one SOH and one HI corresponding to one effective charging cycle may be obtained.
Optionally, before combining the characteristic data according to the characteristic data of the capacity increment IC curves corresponding to the plurality of effective charging cycles to obtain a plurality of health factors HI as the input variables of the GPR model, the method includes: performing capacity increment processing on the plurality of effective charging cycles to obtain corresponding IC curves; extracting characteristic data of the IC curve according to the IC curve, wherein the characteristic data comprises at least one of the following data: peak, peak area, and voltage at which the peak is located.
As an alternative embodiment, for an incomplete capacity increment IC curve, feature data of a partially available IC curve may be extracted.
As an alternative embodiment, before extracting the feature data of the IC curve according to the IC curve, the method further includes: carrying out noise processing on the obtained IC curve to obtain a denoised IC curve; extracting the characteristic data of the denoised IC curve according to the denoised IC curve; the characteristic data includes, but is not limited to, a peak value, a peak area, and a voltage at which the peak value is located.
As an alternative embodiment, the feature data may be continuously obtained by using a moving window used by the smoothing filter, and a high-order polynomial of least square fitting may be used for the feature data in the current window.
The capacity increment analysis method is an important method for researching the characteristic decay mechanism of the lithium ion power battery material, and takes a constant current charging process as an example, the current I is constant, so that the capacity change dQ is I dt, the dQ/dV is I dt/dV, and the smaller the voltage change in unit time is, the larger dt/dV is, and the platform of the charging curve is presented in a peak value mode.
And for the actual operation data of the lithium battery, an IC curve is obtained by adopting a method of counting the number of sampling points. Generally, the sampling frequency of the battery terminal voltage is constant, and the electric quantity charged by the battery is directly proportional to the number of sampling points in the charging process. When the voltage rises rapidly, the number of sampling points is small, and when the voltage rises slowly, the number of sampling points is large, and the voltage appears flat and has a peak value corresponding to an IC curve.
Wherein n is the number of sampling points counted in the corresponding interval, I is the current of constant current charging, f is the sampling frequency, and δ V is the voltage interval width.
Optionally, the step of combining the characteristic data according to the characteristic data of the capacity increment IC curves corresponding to the plurality of effective charging cycles to obtain a plurality of health factors HI as the input variables of the GPR model includes: normalizing the characteristic data of the IC curve to obtain normalized characteristic data; calculating the gray correlation degree of the normalized characteristic data and the SOH by utilizing gray correlation Analysis (GRA); and combining according to the grey correlation degrees corresponding to the normalized feature data to obtain a plurality of HIs.
The normalization processing of the feature data of the pair of IC curves is to distribute the feature data in a probability form within a certain probability interval, for example, the probability interval may be [0, 1], and of course, may be other probability intervals. At this time, the normalized feature data is regularly distributed.
The gray correlation analysis GRA may obtain a gray correlation degree between the normalized feature data and the SOH.
In a specific implementation, each HI may be obtained by combining the gray correlation degrees corresponding to the normalized feature data. Take Peak and Area combination as an example:
The method can obtain corresponding HI by using available characteristic data combination when certain characteristic data is missing.
Optionally, determining a plurality of effective charging cycles according to the lithium battery operation data comprises: preprocessing the lithium battery operation data to obtain optimized operation data, wherein the preprocessing at least comprises one of the following steps: removing abnormal values, filling missing values and smoothing filtering; a plurality of effective charging cycles is determined based on the optimized operational data.
As an alternative embodiment, the above-mentioned abnormal value elimination processing includes eliminating the abnormal value by using a threshold setting method, a correlation analysis, a repeated redundant data filtering, and the like; the missing value filling processing comprises the steps of using regression, Bayesian formalization, fixed value filling and the like to fill missing values; the smoothing filter processing described above includes smoothing filtering using least square filtering, cubic spline interpolation, or the like. The preprocessing method is used for optimizing the operation data of the lithium battery, and one method or a combination of several methods can be adopted in the specific implementation process. Moreover, the optimized operation data can be obtained through the preprocessing method, so that a plurality of finally obtained effective charging cycles are more accurate and have reference significance.
Optionally, determining a plurality of effective charging cycles according to the lithium battery operation data comprises: determining a zone bit of lithium battery operation data; intercepting the charging cycle of the lithium battery operation data according to the zone bit; acquiring a data set of a charging cycle; judging whether the data set of the charging cycle meets a preset threshold value or not, wherein the preset threshold value comprises a data acquisition quantity threshold value and a data acquisition duration threshold value; under the condition that the data set meets a preset threshold value, the charging cycle corresponding to the data set is an effective charging cycle; otherwise, the charging cycle corresponding to the data set is an invalid charging cycle.
As an alternative embodiment, the flag bits may include a charging state open flag, a cycle number accumulation, a charger (pile) connection state, and the like according to the source of the lithium battery operation data. Under the condition of no flag bit, whether the current fluctuates slightly in a certain negative value for a long time is judged, and then the constant current charging state can be judged. In the case of the flag bit, the data of the flag bit representing the constant current charging state is intercepted and recorded as a charging cycle, and the acquired data set of the charging cycle includes, but is not limited to, voltage, current, temperature, mileage accumulation value and the like.
As an alternative embodiment, the preset threshold needs to be set in consideration of many factors, such as data dispersion and charge-discharge rate. Further, a number threshold or a duration threshold may be set according to the above factors, wherein the setting of the preset threshold should be set according to a specific application scenario.
As an optional embodiment, when a data set of a charging cycle meets a preset threshold, the charging cycle corresponding to the data set is an effective charging cycle; otherwise, the charging cycle corresponding to the data set is an invalid charging cycle. By the method, the original charging cycles can be classified quickly and accurately.
Optionally, calculating the gray correlation degree of the normalized feature data and the SOH by using the gray correlation analysis GRA includes:
wherein ρ ∈ [0, 1] is a resolution coefficient, the smaller ρ is, the higher the resolution is, i is 1, 2, 3.. m, i is the ith x compared with y, k is time, x is SOH, and y is normalized feature data;
obtaining a plurality of HIs by combining the gray correlation degrees corresponding to the normalized feature data in the following manner:
wherein, FiFor the ith feature data, r is the grey correlation.
The feature data may be normalized feature data or non-normalized feature data, and the feature data includes: peak, peak area, peak voltage, etc., and the value of i may be any one of the characteristic data.
The following describes preferred embodiments of the present invention.
The invention provides a method for estimating the state of health of a lithium battery, which is based on a capacity increment method and Gaussian process regression to realize the detection of the state of health of the lithium battery, namely, the on-line estimation of the SOH of the lithium battery can be realized.
The implementation steps of the method are as follows:
s1: collecting lithium battery operation data, performing data preprocessing, extracting effective charging cycle, and obtaining an SOH sequence as a GPR output variable;
s2: calculating each charging cycle by adopting a capacity increment method, extracting relevant characteristics of a capacity increment curve, and combining the characteristics to obtain a health factor serving as a GPR input variable;
s3: selecting a proper kernel function, and performing GPR model training to obtain an optimal hyperparameter and an optimal GPR model;
s4: and acquiring a health factor of the latest charging cycle, and inputting the trained GPR model to obtain the current SOH.
Further, in step S1, the data preprocessing method includes one or more of outlier rejection, missing value padding, and smooth filtering. The method for removing the abnormal value comprises a threshold setting method, correlation analysis, repeated redundant data screening and the like; the method for filling the missing value comprises regression, Bayesian formalization, fixed value filling and the like; the smoothing filtering method includes least square filtering, cubic spline interpolation and the like. The effective charging cycle is used for acquiring a charging cycle with a wider charging process coverage State of Charge (SOC), and the number of sampling points or the charging duration is used as a screening basis according to different charging multiplying powers in actual operation. In the health state sequence, in order to obtain the charge amount of each charge cycle as the current capacity, the sequence obtained by adopting a normalization method is an SOH sequence.
Further, step S2 includes:
s21: extracting the Peak value (Peak) of the IC curve, the Peak Area (Area) and the voltage (V _ Peak) of the Peak value, and normalizing each characteristic;
s22: calculating the grey correlation degree of each feature and the SOH by adopting grey correlation analysis;
s23: combining the plurality of indirect features to obtain HI.
Further, step S3 includes:
S31: dividing an input variable and an output variable into a training set and a test set according to a certain proportion;
s32: selecting RBFs as kernel functions, and selecting optimal RBF kernel function parameter combinations by using a multi-population particle swarm algorithm;
s33: determining an optimal GPR model according to the optimal RBF kernel function;
s34: and carrying out model precision verification by using the test set.
Further, the gray correlation calculation method in S22 is:
sequence xi(k) The correlation coefficient calculation formula of (a) and (y), (k) is as follows:
where ρ ∈ [0, 1] is the resolution factor, the smaller ρ, the greater the resolution. i is 1, 2, 3.. m, i is the ith x sequence compared with y, and k is the sequence value at time k.
The relevance is the integration of relevance coefficients of all points of the curve:
further, the combination manner in step S23 is:
wherein FiThe ith characteristic value is r, the grey relevance degree is r, and i and j are ordinal numbers.
Further, the optimal RBF kernel function parameter selection method adopted in step S32 is a multi-population particle swarm algorithm, but is not limited to this optimal estimation method.
According to the method, the characteristics of the lithium battery charging curve are obtained through a pretreatment technology and a capacity increment analysis method, the characteristics are combined through grey correlation degree analysis to obtain a health factor, a health factor sequence and a historical SOH sequence data set are used for training a Gaussian process regression model, and further SOH estimation can be carried out through the model when new charging data are collected.
(1) The result obtained by the gaussian process regression model includes a mean value and a confidence interval, and the estimation result is more descriptive.
(2) The algorithm is not limited to a certain type of battery, and the universality is high.
(3) The quality problem of the original data is processed by a data preprocessing technology and a capacity increment analysis method, and the influence on the estimation result is small.
(4) For an incomplete charging cycle curve, part of available features can be extracted to construct a health factor, and online estimation of SOH for the partial charging curve is achieved.
(5) The algorithm can synchronously acquire data to optimize the model in the using process of the lithium battery, and the robustness is strong.
The above embodiments are described in further detail below:
fig. 2 is a flowchart of a method for estimating a state of health of a lithium battery according to a preferred embodiment of the present invention, which may include the steps of, as shown in fig. 2:
s1: collecting lithium battery operation data, performing data preprocessing, extracting effective charging cycle, and obtaining an SOH sequence as a GPR output variable;
s2: calculating each charging cycle by adopting ICA, extracting relevant characteristics of an IC curve, and combining the characteristics to obtain a health factor serving as a GPR input variable;
s3: selecting a proper kernel function, and performing GPR model training to obtain an optimal hyperparameter and an optimal GPR model;
S4: and acquiring a health factor of the latest charging cycle, and inputting the trained GPR model to obtain the current SOH.
Fig. 3 is a flowchart of step S1 in the method for estimating the state of health of a lithium battery according to the preferred embodiment of the invention, and as shown in fig. 3, step S1 specifically includes steps S11 to S17.
S11: collecting operation data of a lithium battery, and determining basic information such as a battery model, rated capacity, factory information, application scenes and the like;
s12: and carrying out data preprocessing. Removing abnormal values by using a threshold setting method, correlation analysis, repeated redundant data screening and the like; completing missing values by using regression, Bayesian formalization, fixed value filling and the like; smoothing filtering using least squares filtering, cubic spline interpolation, or the like.
S13: the selected flag bits, depending on the source of the battery data, may be: the charging state opening mark, the cycle number accumulation, the connection state of a charger (pile), and the like. If no flag bit exists, whether the current fluctuates slightly in a certain negative value for a long time or not is judged, and then the constant current charging state can be judged. And intercepting the data of the zone bit representing the constant current charging state, recording the data as the current cycle, and extracting data columns including but not limited to voltage, current, temperature, mileage accumulation value and the like.
S14: and judging whether the intercepted current cycle data set is effective, if the data acquisition quantity is too small or the acquisition time is too short (the specific threshold value is flexibly selected according to the data dispersion and the charge-discharge multiplying power), entering S15, and if not, entering S16.
S15: the data record is invalid and discarded.
S16: storing the data as an effective cycle, recording the charge capacity of the cycle, calculating the capacity when the current battery SOC is 100% according to the proportion, and recording the capacity as the available capacity of the battery in the current cycle.
S17: and normalizing the capacity sequences obtained by all cycles to obtain the SOH sequence.
Fig. 4 is a flowchart of step S2 in the method for estimating the state of health of a lithium battery according to the preferred embodiment of the invention, and as shown in fig. 4, step S2 specifically includes steps S21 to S24.
S21: and (4) carrying out capacity increment analysis on a single cycle to obtain an IC curve.
The capacity increment analysis method is an important method for researching the characteristic decay mechanism of the lithium ion power battery material, taking a constant current charging process as an example, the current I is a constant, so that the capacity variation dQ is I dt, the dQ/dV is I dt/dV, the smaller the voltage variation in unit time is, the larger dt/dV is, and the platform of the charging curve is presented in a peak value mode.
And for the actual operation data of the lithium battery, an IC curve is obtained by adopting a method of counting the number of sampling points. Generally, the sampling frequency of the battery terminal voltage is constant, and the electric quantity charged by the battery is directly proportional to the number of sampling points in the charging process. When the voltage rises rapidly, the number of sampling points is small, and when the voltage rises slowly, the number of sampling points is large, and the voltage appears flat and has a peak value corresponding to an IC curve.
Wherein n is the number of sampling points counted in the corresponding interval, I is the current of constant current charging, f is the sampling frequency, and δ V is the voltage interval width.
Fig. 5 is a schematic diagram of an IC curve in the method for estimating the state of health of a lithium battery according to the preferred embodiment of the present invention, and the initially acquired IC curve is shown in fig. 5.
S22: and carrying out noise processing on the obtained IC curve and extracting curve characteristics. Here, a Savitzky-Golay smoothing filter is used, which continuously acquires data using a moving window, and performs least square fitting on the data in the current window to obtain a high order polynomial.
Fig. 6 is a schematic diagram of graph noise processing effect and feature extraction of an IC curve in the method for estimating the state of health of a lithium battery according to the preferred embodiment of the present invention, as shown in fig. 6. The A point corresponds to the Peak value (Peak) of the IC curve, the shadow Area of the B point is the Peak Area (Area), the C point corresponds to the Peak voltage (V _ Peak), and normalization processing is carried out on each characteristic sequence.
S23: calculating gray correlation degree of single characteristic sequence and SOH sequence
S24: the HI estimation result obtained by combining the gray correlation r is the best, and taking Peak and Area as an example, the combination mode is as follows:
and if a certain characteristic value of the current cycle is missing, combining the available characteristic values to obtain HI.
Fig. 7 is a flowchart of step S3 in the method for estimating the state of health of a lithium battery according to the preferred embodiment of the invention, and as shown in fig. 7, step S3 specifically includes steps S31 to S34.
S31: taking the HI sequence as model input and the SOH sequence as model output to obtain a data set (X, Y), and dividing the data set into a training set (X, Y) and a testing set (X) according to a certain proportion*，y*) Wherein the HI sequence consists of a plurality of HIs, and the SOH sequence consists of a plurality of SOHs.
S32: RBFs are selected as kernel functions of GPR, and model training is carried out by using a training set. In the training process, the hyper-parameters are optimized through the maximized marginal likelihood function, and the optimization process adopts a multi-swarm particle swarm optimization algorithm.
Specifically, the gaussian process regression algorithm is as follows:
the GPR does not need to specify the specific form of the process f (x), but only needs to assume that it obeys joint gaussian distribution, then y is the observed value interfered by noise, that is:
var (x) is an nxn covariance matrix, and the (i, j) th element is a covariance function, i.e., Radial Basis Function (RBF):
wherein σfAnd λ is a constant. When there is a new test set S*＝(x*，y*) When the same obeys Gaussian distribution, then the vinegar is predicted
And
comprises the following steps:
wherein, k (X)*)＝[Var(x*，x1)，...，Var(x*，xn)]T
The hyperparameter θ ═ σn，σfLambda) whether the prediction result is accurately influenced, and the optimal hyper-parameter is obtained by iteration of a multi-population particle swarm optimization algorithm. θ is obtained by maximizing the marginal likelihood according to the following equation:
fig. 8 is a flowchart of the multi-population optimization algorithm in step S32 of the method for estimating the state of health of lithium battery according to the preferred embodiment of the present invention, as shown in fig. 8, the multi-population optimization algorithm in step S32 includes steps S321 to S324, specifically:
s321: determining population scale, maxT, initial omega, acceleration factor and the like; the particles are averagely divided into 3 subgroups, wherein subgroup 1 iterates according to PSO, subgroup 2 iterates according to omega self-linear adjustment particle swarm algorithm, and subgroup 3 iterates according to cloud self-adaptive particle swarm algorithm.
Specifically, the method comprises the following steps:
the PSO is mainly subjected to evolution calculation according to the following two formulas:
vi(t+1)＝ωvi(t)+c1r1(pi(t)-xi(t))+c2r2(pg(t)-xi(t))
xi(t+1)＝xi(t)+vi(t+1)
Where ω is the inertia weight, c1And c2For the acceleration factor, these three coefficients are fixed constants, r1And r2Is [0, 1 ]]Random numbers within a range. v. ofi(t) is the velocity of particle i at time t, xi(t) is the current position, pi(t) is the optimum position of the particle i, pg(t) is the optimal position for all particles.
Omega is adjusted by a self-linear adjustment particle swarm algorithm, is larger at the initial stage of iteration and is beneficial to quick search, and is reduced at the later stage of iteration, so that the optimal position is positioned. The ω adjustment is as follows:
maxT is the maximum number of iterations
The cloud self-adaptive Particle Swarm Optimization combines a cloud model, a Particle Swarm Optimization (PSO for short) and a self-adaptive idea, and a weight adjustment strategy is carried out according to the following formula:
Ex＝gbest
E′n＝rand(En，He)
ωi(t+1)＝ωi(t)×pi(t)
wherein, c1And c2To control the parameters, the morphology of the cloud model is determined. k is an initial adjustment factor, typically 0.9. p is a radical ofi(t) is the weighted inertial probability of the ith particle at time t. When the global optimal difference value of the current particle distance is larger, the weight omega is larger, otherwise, the weight omega is smaller, and the value of omega has certain ambiguity, thereby greatly increasing the particle abundance, promoting the optimizing rate and reducing the probability of falling into the local optimal.
S322: and randomly generating the position and the speed of the primary particles, selecting optimal particles for the three subgroups respectively, comparing to obtain global optimal particles and calculating the current optimal adaptive value.
S323: and calculating new weights according to the weight calculation strategies of the respective populations, iterating, updating the sub-population optimal particles and the global optimal particles, and calculating an adaptive value.
S324: and judging whether the iteration termination condition is met, and if not, repeating the step S323.
S33: and (3) realizing SOH estimation by adopting a trained GPR model, judging the error between the output of the model and a true value, and adopting a root mean square error as a judgment basis. If the root mean square error meets the convergence condition, the convergence condition can be that the iteration number reaches the upper limit or the precision meets the condition, and then the operation is finished; otherwise, the process proceeds to S32.
Fig. 9 is a schematic diagram of an SOH estimation result in the estimation method of the state of health of a lithium battery according to the preferred embodiment of the present invention, as shown in fig. 9, the SOH estimation result is obtained based on the operation data of a vehicle-mounted lithium battery of a certain company, and when a vehicle is actually operated, because there are many shelves, the capacity recovery phenomenon is obvious, and the capacity fading curve has obvious fluctuation, but the whole shows a slow-down trend.
The solid line is the true SOH (SOH _ real), the dashed line is the estimated SOH (SOH _ pre) based on the trained GPR model extracted HIs in the current cycle, the first and fourth curves are the upper and lower limits of the 95% confidence interval of the estimate, respectively, and it can be seen that substantially all true values are within the upper and lower bounds of the confidence interval. The estimation algorithm uses 56.6% of the data set as training data, and the root mean square error of the estimated value sequence and the real value sequence obtained on the test set is 0.00374. The error is within an acceptable range in consideration of the influence caused by the vehicle data acquisition precision and the complex operation condition.
Fig. 10 is a schematic structural diagram of an apparatus for estimating a state of health of a lithium battery according to an embodiment of the present invention, and as shown in fig. 10, the apparatus for estimating a state of health of a lithium battery includes: a training module 1002 and a detection module 1004. The following describes the estimation device of the state of health of the lithium battery in detail.
A training module 1002, configured to train using multiple health factors HI corresponding to multiple effective charging cycles as input variables and multiple health states SOH corresponding to multiple effective charging cycles as output variables, to obtain a gaussian process regression GPR model, where the GPR model is obtained by using multiple sets of data through machine learning training, and each set of data in the multiple sets of data includes: HI for an active charge cycle and SOH for HI; and the detection module 1004 is connected to the training module 1002, and is configured to acquire HI of a charging cycle to be detected, input the HI into a GPR model, and output SOH corresponding to the HI by the GPR model.
According to the embodiment, the HI of the charging cycle can be detected through the GPR model, the purpose of obtaining the SOH corresponding to the HI according to the HI of the charging cycle is achieved, the technical effect of improving the adaptability to data is improved while the complexity of detecting the health state of the lithium battery is reduced, and the technical problems that the detection of the health state of the lithium battery is complex and the data acquisition with poor quality is difficult to adapt in the related technology are solved.
According to another aspect of the embodiments of the present invention, there is also provided a storage medium storing a program, wherein the program, when executed by a processor, causes the processor to execute the method for estimating the state of health of a lithium battery according to any one of the above.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
In the above embodiments of the present invention, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
In the embodiments provided in the present application, it should be understood that the disclosed technology can be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units may be a logical division, and in actual implementation, there may be another division, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, units or modules, and may be in an electrical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (8)
1. A method for estimating the state of health of a lithium battery, comprising:
training by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to the plurality of effective charging cycles as output variables to obtain a Gaussian process regression GPR model, wherein the GPR model is obtained by using a plurality of groups of data through machine learning training, and each group of data in the plurality of groups of data comprises: an HI of an active charge cycle and a SOH corresponding to the HI;
acquiring HI of a charging cycle to be detected, inputting the HI into the GPR model, and outputting SOH corresponding to the HI by the GPR model;
before training a Gaussian process regression GPR model by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to the plurality of effective charging cycles as output variables, the method comprises the following steps: collecting lithium battery operation data; determining a plurality of effective charging cycles according to the lithium battery operation data, wherein the effective charging cycles are charging cycles for acquiring the covering charge state of the charging process and reaching a preset charge state threshold;
Before training a gaussian process regression GPR model by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to the plurality of effective charging cycles as output variables, the method further comprises: according to the characteristic data of the capacity increment IC curves corresponding to the effective charging cycles, combining the characteristic data to obtain a plurality of health factors HI; and a plurality of state of health (SOH) obtained from the plurality of effective charging cycles, wherein the health factor is obtained by combining available characteristic data for the characteristic data of the incremental IC curve corresponding to the incomplete charging cycle.
2. The method of claim 1, wherein before combining the characterization data from capacity increment IC curves corresponding to the plurality of active charge cycles to obtain a plurality of health factor HI as input variables for a GPR model, comprising:
performing capacity increment processing on the plurality of effective charging cycles to obtain corresponding IC curves;
extracting feature data of the IC curve according to the IC curve, wherein the feature data comprises at least one of the following data: peak, peak area, and voltage at which the peak is located.
3. The method of claim 1, wherein combining the characteristic data of the capacity increment IC curves corresponding to the plurality of effective charge cycles to obtain a plurality of health factor HI as input variables of a GPR model comprises:
normalizing the characteristic data of the IC curve to obtain normalized characteristic data;
calculating the grey correlation degree of the normalized feature data and the SOH by utilizing a grey correlation analysis (GRA);
and combining according to the grey correlation degrees corresponding to the normalized feature data to obtain the plurality of HIs.
4. The method of claim 1, wherein determining a plurality of effective charge cycles based on the lithium battery operational data comprises:
preprocessing the lithium battery operation data to obtain optimized operation data, wherein the preprocessing at least comprises one of the following steps: removing abnormal values, filling missing values and smoothing filtering;
determining the plurality of effective charging cycles based on the optimized operational data.
5. The method of claim 1, wherein determining a plurality of effective charge cycles based on the lithium battery operational data comprises:
Determining a zone bit of the lithium battery operation data;
intercepting the charging cycle of the lithium battery operation data according to the zone bit;
acquiring a data set of the charging cycle;
judging whether the data set of the charging cycle meets a preset threshold value, wherein the preset threshold value comprises a data acquisition quantity threshold value and a data acquisition duration threshold value;
under the condition that the data set meets the preset threshold value, the charging cycle corresponding to the data set is an effective charging cycle; otherwise, the charging cycle corresponding to the data set is an invalid charging cycle.
6. The method of claim 3,
calculating a gray correlation degree of the normalized feature data and the SOH by using a gray correlation analysis (GRA) according to the following modes:
wherein ρ ∈ [0, 1] is a resolution coefficient, the smaller ρ is, the higher the resolution is, i is 1, 2, 3.. m, i is the ith x compared with y, k is time, x is SOH, and y is normalized feature data;
obtaining the plurality of HIs by combining the gray correlation degrees corresponding to the normalized feature data in the following manner includes:
wherein, FiFor the ith feature data, r is the grey correlation.
7. An apparatus for estimating a state of health of a lithium battery, comprising:
the training module is used for training by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to the plurality of effective charging cycles as output variables to obtain a Gaussian process regression GPR model, wherein the GPR model is obtained by using a plurality of groups of data through machine learning training, and each group of data in the plurality of groups of data comprises: an HI of an active charge cycle and a SOH corresponding to the HI;
the detection module is used for acquiring HI of a charging cycle to be detected, inputting the HI into the GPR model and outputting SOH corresponding to the HI by the GPR model;
the device is used for collecting the operation data of the lithium battery before training a Gaussian process regression GPR model by using a plurality of health factors HI corresponding to a plurality of effective charging cycles as input variables and a plurality of health states SOH corresponding to a plurality of effective charging cycles as output variables; determining a plurality of effective charging cycles according to the lithium battery operation data, wherein the effective charging cycles are charging cycles for acquiring the covering charge state of the charging process and reaching a preset charge state threshold;
The device is further used for combining the characteristic data according to the characteristic data of the capacity increment IC curves corresponding to the effective charging cycles to obtain a plurality of health factors HI before training a Gaussian process regression GPR model by using the health factors HI corresponding to the effective charging cycles as input variables and the health states SOH corresponding to the effective charging cycles as output variables; and a plurality of state of health (SOH) obtained from the plurality of effective charging cycles, wherein the health factor is obtained by combining available characteristic data for the characteristic data of the incremental IC curve corresponding to the incomplete charging cycle.
8. A storage medium storing a program, wherein the program, when executed by a processor, causes the processor to execute the method for estimating a state of health of a lithium battery according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910371690.4A CN110068774B (en) | 2019-05-06 | 2019-05-06 | Lithium battery health state estimation method and device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910371690.4A CN110068774B (en) | 2019-05-06 | 2019-05-06 | Lithium battery health state estimation method and device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110068774A CN110068774A (en) | 2019-07-30 |
| CN110068774B true CN110068774B (en) | 2021-08-06 |
Family
ID=67370029
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910371690.4A Active CN110068774B (en) | 2019-05-06 | 2019-05-06 | Lithium battery health state estimation method and device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110068774B (en) |
Families Citing this family (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110703112A (en) * | 2019-10-14 | 2020-01-17 | 重庆大学 | Online estimation method of battery pack state based on local charging data |
| CN110927606B (en) * | 2019-11-05 | 2022-01-25 | 腾讯科技（深圳）有限公司 | Battery state monitoring method and device |
| CN112834942A (en) * | 2019-11-25 | 2021-05-25 | 郑州宇通集团有限公司 | Battery management system service life testing method and device based on temperature alternation test |
| CN112834943A (en) * | 2019-11-25 | 2021-05-25 | 郑州宇通集团有限公司 | Method and device for testing service life of battery management system based on high-temperature endurance test |
| CN111007417A (en) * | 2019-12-06 | 2020-04-14 | 重庆大学 | Battery pack SOH and RUL prediction method and system based on inconsistency evaluation |
| CN110927591B (en) * | 2019-12-11 | 2020-10-02 | 北京理工大学 | Battery capacity estimation method, computer readable medium and vehicle |
| CN111308378B (en) * | 2020-03-10 | 2021-05-18 | 北京理工大学 | Battery pack health state detection method and system based on capacity increment curve |
| CN111398833B (en) * | 2020-03-13 | 2021-08-31 | 浙江大学 | Battery health state assessment method |
| CN111323719A (en) * | 2020-03-18 | 2020-06-23 | 北京理工大学 | Method and system for online determination of health state of power battery pack of electric automobile |
| CN111460380B (en) * | 2020-03-30 | 2022-03-18 | 上海交通大学 | Multi-working-condition driving range prediction method and system based on Gaussian process regression |
| CN111398837A (en) * | 2020-04-01 | 2020-07-10 | 重庆大学 | Vehicle battery health state estimation method based on data driving |
| CN111458649A (en) * | 2020-04-23 | 2020-07-28 | 国网陕西省电力公司汉中供电公司 | Rapid detection method for health degree of battery module |
| CN111624495B (en) * | 2020-04-28 | 2022-08-02 | 合肥工业大学 | Lithium battery SOC interval estimation method and system based on EKF (extended Kalman Filter) optimization by deep belief network |
| CN111458643B (en) * | 2020-05-22 | 2022-04-15 | 清华四川能源互联网研究院 | Abnormal battery screening method and device, electronic equipment and readable storage medium |
| JP7457575B2 (en) * | 2020-05-25 | 2024-03-28 | 株式会社Ａｅｓｃジャパン | Deterioration estimation device, model generation device, deterioration estimation method, model generation method, and program |
| CN111896873B (en) * | 2020-06-28 | 2023-03-14 | 中国电力科学研究院有限公司 | Method and system for evaluating health state of energy storage power station |
| CN111917840B (en) * | 2020-07-15 | 2023-02-03 | 唐山学院 | Remote communication information processing method and system for lithium battery |
| CN112083337B (en) * | 2020-10-22 | 2023-06-16 | 重庆大学 | Predictive operation and maintenance-oriented power battery health prediction method |
| CN113065693B (en) * | 2021-03-22 | 2022-07-15 | 哈尔滨工程大学 | Traffic flow prediction method based on radial basis function neural network |
| CN113406520B (en) * | 2021-05-21 | 2022-11-15 | 北京理工大学 | Battery health state estimation method for real new energy automobile |
| CN113567875A (en) * | 2021-06-11 | 2021-10-29 | 国电南瑞科技股份有限公司 | Ternary lithium ion battery health state estimation method based on K nearest neighbor regression |
| CN113640674B (en) * | 2021-06-30 | 2022-09-06 | 昆明理工大学 | Lithium ion battery available capacity estimation method based on optimized Gaussian process regression |
| CN113721159A (en) * | 2021-09-27 | 2021-11-30 | 重庆大学 | Data-driven battery health estimation method oriented to random local charging |
| CN113820615B (en) * | 2021-09-30 | 2024-05-07 | 国网福建省电力有限公司龙岩供电公司 | Battery health degree detection method and device |
| CN114035095B (en) * | 2021-11-02 | 2023-12-22 | 江苏博强新能源科技股份有限公司 | Lithium battery SOH estimation method, medium and equipment based on voltage curve inflection point identification |
| CN114243833A (en) * | 2021-12-16 | 2022-03-25 | 湖北亿纬动力有限公司 | Battery system control method and battery system |
| CN114720879A (en) * | 2022-03-25 | 2022-07-08 | 中国长江三峡集团有限公司 | Energy storage lithium battery pack aging mode automatic identification method based on BP neural network |
| CN115144780A (en) * | 2022-06-16 | 2022-10-04 | 中国第一汽车股份有限公司 | Battery health detection method and storage medium |
| CN116593903B (en) * | 2023-07-17 | 2023-10-20 | 中国华能集团清洁能源技术研究院有限公司 | Battery remaining life prediction method and device |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10393813B2 (en) * | 2013-08-27 | 2019-08-27 | The Regents Of The University Of Michigan | On-board state of health monitoring of batteries using incremental capacity analysis |
| CN106980091B (en) * | 2017-03-29 | 2019-09-17 | 北京理工大学 | A kind of electrokinetic cell system health status estimation method based on fractional model |
| CN109031153B (en) * | 2018-10-16 | 2020-01-24 | 北京交通大学 | Online health state estimation method for lithium ion battery |
| CN109541396B (en) * | 2018-11-27 | 2021-04-27 | 国网山东省电力公司电力科学研究院 | Method for extracting effective recording data of fault indicator |
| CN111007417A (en) * | 2019-12-06 | 2020-04-14 | 重庆大学 | Battery pack SOH and RUL prediction method and system based on inconsistency evaluation |
-
2019
- 2019-05-06 CN CN201910371690.4A patent/CN110068774B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110068774A (en) | 2019-07-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110068774B (en) | Lithium battery health state estimation method and device and storage medium | |
| CN111707951B (en) | Battery pack consistency evaluation method and system | |
| CN111007399B (en) | Lithium battery state of charge prediction method based on improved generation countermeasure network | |
| JP5944291B2 (en) | Battery parameter estimation apparatus and method | |
| CN106026260A (en) | SOC estimation method for series-wound battery pack having equalization circuit | |
| CN110146822A (en) | A kind of Vehicular dynamic battery capacity On-line Estimation method based on constant-current charge process | |
| CN109001640B (en) | Data processing method and device for power battery | |
| CN108556682B (en) | Driving range prediction method, device and equipment | |
| CN110609232B (en) | Lithium ion battery SOC estimation method and device and terminal equipment | |
| CN107132490B (en) | Method for estimating state of charge of lithium battery pack | |
| CN111581904A (en) | Lithium battery SOC and SOH collaborative estimation method considering influence of cycle number | |
| CN111426957B (en) | SOC estimation optimization method for power battery under simulated vehicle working condition | |
| CN113189490B (en) | Lithium battery health state estimation method based on feature screening and Gaussian process regression | |
| CN115902647B (en) | Intelligent battery state monitoring method | |
| CN112083346B (en) | LSTM-based parallel battery pack internal current distribution estimation method and system | |
| CN113109717B (en) | Lithium battery state of charge estimation method based on characteristic curve optimization | |
| CN114035072A (en) | Battery pack multi-state joint estimation method based on cloud edge cooperation | |
| CN114290960A (en) | Method and device for acquiring battery health degree of power battery and vehicle | |
| CN108872865B (en) | Lithium battery SOC estimation method for preventing filtering divergence | |
| CN113761726A (en) | Lithium battery parameter identification method and system | |
| CN113406503A (en) | Lithium battery SOH online estimation method based on deep neural network | |
| CN116953547A (en) | Energy storage battery health evaluation method, system, equipment and storage medium | |
| CN116298936A (en) | Intelligent lithium ion battery health state prediction method in incomplete voltage range | |
| CN113406500B (en) | Method for estimating residual electric quantity of power lithium battery | |
| CN112418564A (en) | Charging and battery replacing load prediction method for charging and battery replacing station based on LSTM and related components thereof |
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 |