CN111413622B - Lithium battery life prediction method based on stacking noise reduction automatic coding machine - Google Patents
Lithium battery life prediction method based on stacking noise reduction automatic coding machine Download PDFInfo
- Publication number
- CN111413622B CN111413622B CN202010261039.4A CN202010261039A CN111413622B CN 111413622 B CN111413622 B CN 111413622B CN 202010261039 A CN202010261039 A CN 202010261039A CN 111413622 B CN111413622 B CN 111413622B
- Authority
- CN
- China
- Prior art keywords
- lithium battery
- data
- effective capacity
- noise reduction
- residual effective
- 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
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]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- 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]
- G01R31/385—Arrangements for measuring battery or accumulator variables
-
- 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]
- G01R31/385—Arrangements for measuring battery or accumulator variables
- G01R31/387—Determining ampere-hour charge capacity or SoC
-
- 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)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
- Secondary Cells (AREA)
Abstract
The invention relates to a lithium battery service life prediction method based on a stacking noise reduction automatic coding machine, and belongs to the technical field of battery management. According to the method, rated capacity of a lithium battery provided when the lithium battery leaves a factory is used as initial data, relevant data are collected through each charge and discharge cycle, then the data are screened according to preset screening conditions, missing data and abnormal data are added and corrected through an interpolation algorithm, then a mapping relation between the relative residual effective capacity and the residual effective capacity of the lithium battery is established through a stacking noise reduction automaton, the residual effective capacity of the lithium battery is estimated, a fitting curve of the residual effective capacity of the lithium battery is obtained through polynomial fitting by combining historical data, and therefore the service life of the lithium battery is predicted. The invention has high reliability, wide applicability and high data utilization rate.
Description
Technical Field
The invention belongs to the technical field of battery management, and relates to a lithium battery service life prediction method based on a stacking noise reduction automatic coding machine.
Background
The lithium battery life prediction means that the residual effective capacity of the lithium battery is scientifically estimated and predicted from the application angle, the operation and maintenance of the lithium battery are further known, a state monitoring and health management system of the lithium battery is constructed, the overcharge and the overdischarge of the battery are prevented, the performance state of the lithium battery is estimated, the residual effective capacity evolution of the lithium battery is predicted, the important aspect of realizing the long-time reliable work of the lithium battery is realized, and the lithium battery life prediction method has great significance for system task decision and the prevention of catastrophic accidents caused by accidental battery life termination.
According to the regulation in IEEE standard 1188-. However, in the actual use process, the lithium battery does not perform the charge-discharge cycle from full discharge to full charge every time, so that the remaining effective capacity of the lithium battery cannot be accurately obtained. Therefore, the invention aims to obtain the relative residual effective capacity, and the mapping relation between the relative residual effective capacity and the residual effective capacity is established by stacking the noise reduction automatic coding machine, so that the service life of the lithium battery is predicted.
Disclosure of Invention
In view of the above, the present invention provides a method for predicting a lifetime of a lithium battery based on a stacking noise reduction automatic coding machine.
In order to achieve the purpose, the invention provides the following technical scheme:
a lithium battery life prediction method based on a stacking noise reduction automatic coding machine comprises the following steps:
1) the rated capacity of the lithium battery provided by the delivery of the lithium battery is initial data, and the number of charge and discharge cycles is recorded as zero;
2) the method comprises the steps of collecting relevant data of a lithium battery every time the lithium battery performs charge-discharge circulation, wherein the relevant data mainly comprises discharge ending time t0Discharge cutoff voltage U0Charge cutoff time t1Charging cut-off voltage U1And the charging current I, and judging whether the data is valid or not according to the preset condition;
3) if the data of the charge-discharge cycle is valid, cleaning the data according to a preset abnormal point detection method, and correcting the abnormal data by adopting an interpolation algorithm; if the data of the charge-discharge cycle is invalid, the data is regarded as missing data, the missing data is interpolated through an interpolation algorithm, and the relative residual effective capacity of the lithium battery in a certain voltage range is calculated;
4) calculating the relative residual effective capacity of the lithium battery in a certain voltage range according to the related data;
5) establishing a mapping relation between the relative residual effective capacity and the residual effective capacity through a stacking noise reduction automaton, estimating the residual effective capacity of the lithium battery, and obtaining a fitting curve of the residual effective capacity of the lithium battery through polynomial fitting by combining historical data;
6) repeating the step 2), the step 3) and the step 4) to obtain new relative residual effective capacity along with the repeated use of the lithium battery, performing reinforced training through a stacking noise reduction automatic coding machine to obtain new residual effective capacity, and then correcting the fitting curve obtained in the step 5) through polynomial fitting.
Optionally, the preset conditions specifically include:
performing a charge-discharge cycle on the lithium battery once, and if the voltage of the lithium battery, the charging of which is started, does not exceed a preset lower limit voltage value and the voltage of the lithium battery, the charging of which is finished, is not lower than a preset upper limit voltage value, the collected related data of the lithium battery is regarded as effective; and if the voltage of the lithium battery which starts to be charged exceeds a preset lower limit voltage value or the voltage of the lithium battery which finishes charging is lower than a preset upper limit voltage value, the collected related data of the lithium battery are regarded as invalid.
Optionally, when the lithium battery performs a charge-discharge cycle once, under the condition that the collected lithium battery related data is valid, the discharge cutoff time of the lithium battery is t0The lithium battery charging cut-off time is t1And calculating the relative residual effective capacity of the lithium battery from a value not exceeding a preset lower limit voltage value to a value not less than a preset upper limit voltage value during each charging through ampere integration, wherein the calculation formula is as follows:
optionally, the interpolation algorithm specifically includes:
let XcFor a collection of collected lithium battery related sample data x, x*Is a set XcWhere there are missing data samples, x is represented as a feature vector (a) in an n-dimensional space1(x)，a2(x)，...，an(x))；
Calculating x*And set XeThe Euclidean distance between all other samples x is calculated by only considering x*Without missing coordinates of the value, finally determining x*K nearest neighbors of (a):
in the formula, d (x)*，xj) Is x*And xjThe euclidean distance between; m is x*Where m is a 1-dimensional or multi-dimensional array;
interpolating x by the mean value of the data at the corresponding coordinate positions of the K nearest neighbors*The missing coordinate value of (a);
in the formula, NN represents the numbers of K nearest neighbors.
Optionally, the establishing of the mapping relationship between the relative remaining effective capacity and the remaining effective capacity by the stacked noise reduction automaton is specifically:
on the basis of a certain amount of relevant data of the lithium battery, normalization processing is carried out on the charging duration, the discharging cut-off voltage, the charging cut-off voltage and the relative remaining effective capacity data;
dividing the processed data into two groups according to a certain proportion and certain conditions, wherein one group is a training set, and the other group is a testing set;
let the training set data x be [0, 1 ]]dFor the input vector, y ∈ [0, 1 ]]d′Is the vector quantity, fθA hidden layer mapping function is formed; by randomly mapping x' qD(x '| x) changing the input vector x to x';
x' is mapped to the hidden layer output by the automatic coding machine:
y＝fθ(x)＝s(Wx+b)
where, s is a Sigmoid function,
w is a weight matrix of d x d'; b is a bias vector; f ofThe network parameter is θ, θ ═ W, b };
each training data x(i)Mapping to corresponding y(i)Reconstructed into a vector z(i)(z∈[0，1]d):
z＝gθ′(x)＝s(W′x+b′)
Wherein, the network parameter of g is θ ', θ' ═ { W ', b' }; the weight matrix W 'satisfies W' WT by definition;
inputting training data x through a first-layer noise reduction automatic coding machine to obtain reconstructed data alpha and a mapping function fθNetwork parameter θ ═ { W, b }; then the reconstructed data alpha is used as the input data of the second layer automatic coding machine, and the same training is carried out, so that new reconstructed data alpha can be reconstructed and obtained1And a mapping function fθ (1)Network parameter θ of(1)＝{W,b}；
Repeating the steps, and constructing a stacking noise reduction automatic coding machine to obtain a multilayer mapping network weight;
initializing a deep neural network by using the obtained multilayer mapping network weight, and iterating the network weight by a small-batch gradient descent method in a BP algorithm until convergence;
inputting test set data, generating network output through feedforward calculation, carrying out anti-standardization processing to obtain a predicted value of the residual effective capacity of the lithium battery, and testing the prediction effect of the residual effective capacity of the model lithium battery through an evaluation criterion.
Optionally, the residual effective capacity prediction model obtained by the training set is used for predicting the test set data, the prediction value of the residual effective capacity of the lithium battery is obtained through anti-standardization processing, and validity and accuracy are verified through an evaluation criterion.
The invention has the beneficial effects that:
1. according to the invention, through the fitting curve of the effective battery capacity, not only can the historical health state and the current health state of the lithium battery be known, but also the service life of the lithium battery can be estimated and predicted.
2. The invention can utilize the relevant data of each charge and discharge process in charge and discharge under the condition of considering the actual use condition of the lithium battery, thereby estimating the effective battery capacity of the lithium battery more accurately.
3. With the continuous use of the lithium battery, the fitting curve of the effective battery capacity can be further corrected, so that the estimation and prediction accuracy of the service life of the lithium battery is improved; in addition, the training effect of the stacking noise reduction automatic coding machine is enhanced, so that the preset conditions can be properly adjusted, and the utilization rate of data is improved.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the means of the instrumentalities and combinations particularly pointed out hereinafter.
Drawings
For the purposes of promoting a better understanding of the objects, aspects and advantages of the invention, reference will now be made to the following detailed description taken in conjunction with the accompanying drawings in which:
FIG. 1 is a block flow diagram of the present invention;
FIG. 2 is a model of estimation of remaining effective capacity based on a stacked noise reduction auto-encoder.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention in a schematic way, and the features in the following embodiments and examples may be combined with each other without conflict.
Wherein the showings are for the purpose of illustrating the invention only and not for the purpose of limiting the same, and in which there is shown by way of illustration only and not in the drawings in which there is no intention to limit the invention thereto; to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The same or similar reference numerals in the drawings of the embodiments of the present invention correspond to the same or similar components; in the description of the present invention, it should be understood that if there is an orientation or positional relationship indicated by terms such as "upper", "lower", "left", "right", "front", "rear", etc., based on the orientation or positional relationship shown in the drawings, it is only for convenience of description and simplification of description, but it is not an indication or suggestion that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes, and are not to be construed as limiting the present invention, and the specific meaning of the terms may be understood by those skilled in the art according to specific situations.
As shown in fig. 1 to 2, a lithium battery life prediction method based on a stack noise reduction automatic coding machine includes the following steps:
1) the rated capacity of the lithium battery provided by the delivery of the lithium battery is initial data, and the number of charge and discharge cycles is recorded as zero;
2) the method comprises the steps of collecting relevant data of a lithium battery every time the lithium battery performs charge-discharge circulation, wherein the relevant data mainly comprises discharge ending time t0Discharge cutoff voltage U0Charge cutoff time t1Charging cut-off voltage U1The charging current I and the like, and the data is subjected to relevant judgment to determine whether the data is valid according to preset conditions;
3) if the data of the charge-discharge cycle is valid, cleaning the data according to a preset abnormal point detection method, and correcting the abnormal data by adopting an interpolation algorithm; if the data of the charge-discharge cycle is invalid, the data is regarded as missing data, and the missing data is interpolated through an interpolation algorithm;
4) calculating the relative residual effective capacity of the lithium battery in a certain voltage range according to the related data;
5) establishing a mapping relation between the relative residual effective capacity and the residual effective capacity through a stacking noise reduction automaton, estimating the residual effective capacity of the lithium battery, and obtaining a fitting curve of the residual effective capacity of the lithium battery through polynomial fitting by combining historical data;
6) with the repeated use of the lithium battery, new relative residual effective capacity can be obtained by repeating the step 2), the step 3) and the step 4), the training is enhanced by stacking the noise reduction automatic coding machine to obtain new residual effective capacity, and then the fitting curve obtained in the step 5) is corrected by polynomial fitting.
Preferably, the preset screening conditions are as follows: performing a charge-discharge cycle on the lithium battery once, and if the voltage of the lithium battery, the charging of which is started, does not exceed a preset lower limit voltage value and the voltage of the lithium battery, the charging of which is finished, is not lower than a preset upper limit voltage value, the collected related data of the lithium battery is regarded as effective; and if the voltage of the lithium battery which starts to be charged exceeds a preset lower limit voltage value or the voltage of the lithium battery which finishes charging is lower than a preset upper limit voltage value, the collected related data of the lithium battery are regarded as invalid.
Preferably, the relative remaining battery capacity of the lithium battery is calculated as follows: when the lithium battery is subjected to one-time charge-discharge cycle and the collected relevant data of the lithium battery is effective, the moment when the lithium battery starts to be charged is taken as t0The moment when the lithium battery finishes charging is taken as t1And calculating the relative residual battery capacity of the lithium battery from a value not exceeding a preset lower limit voltage value to a value not less than a preset upper limit voltage value when charging every time through ampere integration, wherein the calculation formula is as follows:
preferably, the interpolation algorithm adopts a K nearest neighbor algorithm, and the specific implementation steps thereof are as follows:
let XcFor the collection of the collected lithium battery related sample data x, x*Is a set XcSamples with missing data in them, x is represented as n-dimensional spaceCharacteristic vector (a) of1(x)，a2(x)，...，an(x))。
Further, x is calculated*And set XcThe Euclidean distance between all other samples x is calculated by only considering x*Without missing coordinates of the value, finally determining x*K nearest neighbors.
In the formula, d (x)*，xj) Is x*And xjThe euclidean distance between; m is x*Where the feature is missing, m is a 1-dimensional or multi-dimensional array.
Further, the missing coordinate value of x is interpolated by the mean value of the data at the corresponding coordinate positions of the K nearest neighbors obtained in the previous step.
In the formula, NN represents the numbers of K nearest neighbors.
Preferably, the process of establishing the mapping relationship between the relative remaining effective capacity and the remaining effective capacity by the stacking noise reduction automaton is as follows: on the basis of a certain amount of relevant data of the lithium battery, the data such as charging duration, discharging cut-off voltage, charging cut-off voltage, relative residual effective capacity, cycle number and the like are subjected to normalization processing.
Further, the processed data is divided into two groups according to a certain proportion and certain conditions, wherein one group is a training set, and the other group is a testing set.
Further, let training set data x ∈ [0, 1 ∈ [ ]]dFor the input vector, y ∈ [0, 1 ]]d′Is the vector quantity, fθA function is mapped for hidden layers. By randomly mapping x' qD(x '| x) changes the input vector x to x'.
Further, x' is mapped to the hidden layer output by the auto-encoder:
y＝fθ(x)＝s(Wx+b)
where, s is a Sigmoid function,
w is a weight matrix of d x d'; b is a bias vector; the network parameter of f is θ, θ ═ W, b.
Further, each training data x(i)Mapping to corresponding y(i)Reconstructed into a vector z(i)(z∈[0，1]d):
z＝gθ′(x)＝s(W′x+b′)
Wherein, the network parameter of g is θ ', θ' ═ { W ', b' }; the weight matrix W 'satisfies W' by defining W ═ WT。
Further, training data x is input through a first-layer noise reduction automatic coding machine to obtain reconstruction data alpha and a mapping function fθIs equal to { W, b }. Then the reconstructed data alpha is used as the input data of the second layer automatic coding machine, and the same training is carried out, so that new reconstructed data alpha can be reconstructed and obtained1And a mapping function fθ (1)Network parameter θ of(1)＝{W,b}。
Further, the above steps are repeatedly implemented, and then the stacking noise reduction automatic coding machine can be constructed, so that the multilayer mapping network weight is obtained.
Further, a deep neural network is initialized by using the multilayer mapping network weight obtained in the previous step, and the network weight is iterated through a small batch gradient descent method in a BP algorithm until convergence.
Furthermore, inputting test set data, generating network output through feedforward calculation, carrying out anti-standardization processing to obtain a predicted value of the residual effective capacity of the lithium battery, and testing the prediction effect of the residual effective capacity of the model lithium battery through an evaluation criterion.
Finally, the above embodiments are only intended to illustrate the technical solutions of the present invention and not to limit the present invention, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions, and all of them should be covered by the claims of the present invention.
Claims (6)
1. A lithium battery life prediction method based on a stacking noise reduction automatic coding machine is characterized by comprising the following steps: the method comprises the following steps:
1) the rated capacity of the lithium battery provided by the delivery of the lithium battery is initial data, and the number of charge and discharge cycles is recorded as zero;
2) the method comprises the steps of collecting relevant data of a lithium battery every time the lithium battery performs charge-discharge circulation, wherein the relevant data mainly comprises discharge ending time t0Discharge cutoff voltage U0Charge cutoff time t1Charging cut-off voltage U1And the charging current I, and judging whether the data is valid or not according to the preset condition;
3) if the data of the charge-discharge cycle is valid, cleaning the data according to a preset abnormal point detection method, and correcting the abnormal data by adopting an interpolation algorithm; if the data of the charge-discharge cycle is invalid, the data is regarded as missing data, the missing data is interpolated through an interpolation algorithm, and the relative residual effective capacity of the lithium battery in a certain voltage range is calculated;
4) calculating the relative residual effective capacity of the lithium battery in a certain voltage range according to the related data;
5) establishing a mapping relation between the relative residual effective capacity and the residual effective capacity through a stacking noise reduction automatic coding machine, estimating the residual effective capacity of the lithium battery, and obtaining a fitting curve of the residual effective capacity of the lithium battery through polynomial fitting by combining historical data;
6) repeating the step 2), the step 3) and the step 4) to obtain new relative residual effective capacity along with the repeated use of the lithium battery, performing reinforced training through a stacking noise reduction automatic coding machine to obtain new residual effective capacity, and then correcting the fitting curve obtained in the step 5) through polynomial fitting.
2. The lithium battery life prediction method based on the stacking noise reduction automatic coding machine according to claim 1, characterized in that: the preset conditions are specifically as follows:
performing a charge-discharge cycle on the lithium battery once, and if the voltage of the lithium battery, the charging of which is started, does not exceed a preset lower limit voltage value and the voltage of the lithium battery, the charging of which is finished, is not lower than a preset upper limit voltage value, the collected related data of the lithium battery is regarded as effective; and if the voltage of the lithium battery which starts to be charged exceeds a preset lower limit voltage value or the voltage of the lithium battery which finishes charging is lower than a preset upper limit voltage value, the collected related data of the lithium battery are regarded as invalid.
3. The lithium battery life prediction method based on the stacking noise reduction automatic coding machine according to claim 2, characterized in that: when the lithium battery is subjected to one-time charge-discharge cycle, under the condition that the collected lithium battery related data is effective, the discharge ending moment of the lithium battery is taken as t0The lithium battery charging cut-off time is t1And calculating the relative residual effective capacity of the lithium battery from a value not exceeding a preset lower limit voltage value to a value not less than a preset upper limit voltage value during each charging through ampere integration, wherein the calculation formula is as follows:
4. the lithium battery life prediction method based on the stacking noise reduction automatic coding machine according to claim 2, characterized in that: the interpolation algorithm specifically comprises the following steps:
let XcFor a collection of collected lithium battery related sample data x, x*Is a set XcWhere there are missing data samples, x is represented as a feature vector (a) in an n-dimensional space1(x)，a2(x)，...，an(x)；
Calculating x*And set XcThe Euclidean distance between all other samples x is calculated by only considering x*Without missing coordinates of the value, finally determining x*K nearest neighbors of (a):
in the formula, d (x)*，xj) Is x*And xjThe euclidean distance between; m is x*Where m is a 1-dimensional or multi-dimensional array;
interpolating x by the mean value of the data at the corresponding coordinate positions of the K nearest neighbors*The missing coordinate value of (a);
in the formula, NN represents the numbers of K nearest neighbors.
5. The lithium battery life prediction method based on the stacking noise reduction automatic coding machine according to claim 1, characterized in that: the establishing of the mapping relation between the relative residual effective capacity and the residual effective capacity through the stacking noise reduction automatic coding machine is specifically as follows:
on the basis of a certain amount of relevant data of the lithium battery, normalization processing is carried out on the charging duration, the discharging cut-off voltage, the charging cut-off voltage and the relative remaining effective capacity data;
dividing the processed data into two groups according to a certain proportion and certain conditions, wherein one group is a training set, and the other group is a testing set;
let the training set data x be [0, 1 ]]dFor the input vector, y ∈ [0, 1 ]]d′As an output vector, fθMapping a function for the hidden layer; by randomly mapping x' qD(x '| x) changing the input vector x to x';
x' is mapped to the hidden layer output by the automatic coding machine:
y＝fθ(x)＝s(Wx+b)
wherein s is Sigmoid functionThe number of the first and second groups is,
w is a weight matrix of d x d'; b is a bias vector; the network parameter of f is theta, theta is { W, b };
each training data x(i)Mapping to corresponding y(i)Reconstructed into a vector z(i)(z∈[0，1]d)：
z＝gθ′(x)＝s(W′x+b′)
Wherein, the network parameter of g is θ ', θ' ═ { W ', b' }; the weight matrix W 'satisfies W' by defining W ═ WT；
Inputting training data x through a first-layer noise reduction automatic coding machine to obtain reconstructed data alpha and a mapping function fθNetwork parameter θ ═ { W, b }; then the reconstructed data alpha is used as the input data of the second layer automatic coding machine, and the same training is carried out, so that new reconstructed data alpha can be reconstructed and obtained1And a mapping function fθ (1)Network parameter θ of(1)＝{W，b}；
Repeating the steps, and constructing a stacking noise reduction automatic coding machine to obtain a multilayer mapping network weight;
initializing a deep neural network by using the obtained multilayer mapping network weight, and iterating the network weight by a small-batch gradient descent method in a BP algorithm until convergence;
inputting test set data, generating network output through feedforward calculation, carrying out anti-standardization processing to obtain a predicted value of the residual effective capacity of the lithium battery, and testing the prediction effect of the residual effective capacity of the model lithium battery through an evaluation criterion.
6. The lithium battery life prediction method based on the stacking noise reduction automatic coding machine according to claim 5, characterized in that: and predicting the data of the test set by using the residual effective capacity prediction model obtained by the training set, carrying out anti-standardization processing to obtain a predicted value of the residual effective capacity of the lithium battery, and verifying the effectiveness and the accuracy through an evaluation criterion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010261039.4A CN111413622B (en) | 2020-04-03 | 2020-04-03 | Lithium battery life prediction method based on stacking noise reduction automatic coding machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010261039.4A CN111413622B (en) | 2020-04-03 | 2020-04-03 | Lithium battery life prediction method based on stacking noise reduction automatic coding machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111413622A CN111413622A (en) | 2020-07-14 |
| CN111413622B true CN111413622B (en) | 2022-04-15 |
Family
ID=71493383
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010261039.4A Active CN111413622B (en) | 2020-04-03 | 2020-04-03 | Lithium battery life prediction method based on stacking noise reduction automatic coding machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111413622B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116449218B (en) * | 2023-04-26 | 2024-03-12 | 淮阴工学院 | Lithium battery health state estimation method |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE10347110B3 (en) * | 2003-10-10 | 2005-01-13 | Siemens Ag | Cell voltage measuring device for cell stack in motor vehicle power network, has diode pairs connected in parallel with each cell, and switch connected to differential amplifier |
| JP6473924B2 (en) * | 2014-12-03 | 2019-02-27 | 東芝Ｉｔコントロールシステム株式会社 | Battery inspection device |
| CN107765190B (en) * | 2017-12-11 | 2020-02-11 | 太原理工大学 | Method for predicting service life of long-life fast-charging lithium iron phosphate battery |
| CN108510191B (en) * | 2018-04-02 | 2021-10-08 | 重庆大学 | Mangrove ecological health evaluation method based on stacking noise reduction automatic coding algorithm |
| CN110222784B (en) * | 2019-06-13 | 2020-05-26 | 郑州轻工业学院 | Solar cell defect detection method integrating short-term and long-term depth features |
| CN110568359B (en) * | 2019-09-04 | 2021-11-23 | 太原理工大学 | Lithium battery residual life prediction method |
| CN110658459B (en) * | 2019-09-12 | 2021-10-15 | 北京航空航天大学 | Lithium ion battery state of charge estimation method based on bidirectional cyclic neural network |
-
2020
- 2020-04-03 CN CN202010261039.4A patent/CN111413622B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111413622A (en) | 2020-07-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109633474B (en) | Lithium ion battery residual life prediction method | |
| CN111398832A (en) | Bus battery SOC prediction method based on ANFIS model | |
| CN108872869B (en) | Lithium ion battery degradation classification method based on BP neural network | |
| CN109598449B (en) | Lithium ion battery pack performance evaluation method based on charging data | |
| CN113030763A (en) | Potential risk diagnosis system for consistency difference of power batteries for vehicles | |
| KR20180115124A (en) | Apparatus and method for calculating soc | |
| CN112904219B (en) | Big data-based power battery health state prediction method | |
| CN110995153B (en) | Abnormal data detection method and device for photovoltaic power station and electronic equipment | |
| CN111967194B (en) | Battery classification method based on cloud historical data | |
| CN110609233A (en) | Method for predicting SOH of energy storage battery based on big data | |
| CN111413622B (en) | Lithium battery life prediction method based on stacking noise reduction automatic coding machine | |
| US11714134B2 (en) | Device and method for predicting state of battery | |
| CN113030758B (en) | Aging early warning method and system based on lithium ion battery capacity water jump point, automobile and computer storage medium | |
| CN115166533A (en) | Lithium ion battery fault diagnosis method based on segmented dimensionality reduction and outlier identification | |
| CN104882914B (en) | Multi-battery cell balancing method | |
| CN110516813A (en) | A method of batteries of electric automobile RDR prediction is carried out based on big data machine learning | |
| CN111325402A (en) | Method for predicting charging behavior of electric vehicle user based on BP neural network | |
| CN113253128A (en) | Battery system SOC consistency evaluation method and internal resistance consistency evaluation method | |
| CN111983464A (en) | Lithium battery lithium separation testing method based on pure electric vehicle | |
| CN109507594A (en) | Indirect health factor selection method for lithium battery capacity estimation | |
| CN113253113B (en) | Method for predicting consistency fault of power battery | |
| CN117067920B (en) | Fault detection method and device of power battery, electronic equipment and electric automobile | |
| US11953860B2 (en) | System and method of estimating vehicle battery charging time using big data | |
| CN117517966A (en) | Fault diagnosis method for battery pack | |
| CN116442786A (en) | Power battery differential pressure abnormality identification method, device, server and storage medium |
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 |