CN117054887A - Internal fault diagnosis method for lithium ion battery system - Google Patents
Internal fault diagnosis method for lithium ion battery system Download PDFInfo
- Publication number
- CN117054887A CN117054887A CN202311039865.4A CN202311039865A CN117054887A CN 117054887 A CN117054887 A CN 117054887A CN 202311039865 A CN202311039865 A CN 202311039865A CN 117054887 A CN117054887 A CN 117054887A
- Authority
- CN
- China
- Prior art keywords
- data
- battery
- faults
- fault
- battery system
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 title claims abstract description 27
- 229910001416 lithium ion Inorganic materials 0.000 title claims abstract description 27
- 238000003745 diagnosis Methods 0.000 title claims abstract description 14
- 238000004458 analytical method Methods 0.000 claims abstract description 16
- 230000036541 health Effects 0.000 claims abstract description 12
- 230000008569 process Effects 0.000 claims abstract description 5
- 238000013528 artificial neural network Methods 0.000 claims abstract 3
- 238000012545 processing Methods 0.000 claims description 21
- 238000012544 monitoring process Methods 0.000 claims description 15
- 238000013136 deep learning model Methods 0.000 claims description 12
- 238000000605 extraction Methods 0.000 claims description 8
- 238000013480 data collection Methods 0.000 claims description 5
- 238000013527 convolutional neural network Methods 0.000 claims description 4
- 238000012423 maintenance Methods 0.000 claims description 4
- 238000004140 cleaning Methods 0.000 claims description 2
- 238000001816 cooling Methods 0.000 claims description 2
- 230000006403 short-term memory Effects 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 claims 6
- 230000000306 recurrent effect Effects 0.000 claims 2
- 230000005856 abnormality Effects 0.000 claims 1
- 238000013075 data extraction Methods 0.000 claims 1
- 238000013500 data storage Methods 0.000 claims 1
- 238000004146 energy storage Methods 0.000 abstract description 7
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000013135 deep learning Methods 0.000 abstract description 3
- 238000011161 development Methods 0.000 abstract description 3
- 238000010801 machine learning Methods 0.000 abstract description 3
- 238000005457 optimization Methods 0.000 abstract description 3
- 238000007599 discharging Methods 0.000 abstract description 2
- 238000001514 detection method Methods 0.000 abstract 1
- 230000002265 prevention Effects 0.000 abstract 1
- 230000001737 promoting effect Effects 0.000 abstract 1
- 230000008859 change Effects 0.000 description 7
- 238000009826 distribution Methods 0.000 description 7
- 238000012549 training Methods 0.000 description 7
- 230000000875 corresponding effect Effects 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 238000000354 decomposition reaction Methods 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000002405 diagnostic procedure Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000005259 measurement Methods 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
- 239000000178 monomer Substances 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 230000002087 whitening effect Effects 0.000 description 2
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000003487 electrochemical reaction Methods 0.000 description 1
- 230000003862 health status Effects 0.000 description 1
- 230000017525 heat dissipation Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 229910052744 lithium Inorganic materials 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000000348 solid-phase epitaxy Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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]
- G01R31/378—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC] specially adapted for the type of battery or accumulator
-
- 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
-
- 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]
- G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Biology (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Algebra (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The invention provides a method for diagnosing internal faults of a lithium ion battery system, which realizes the accurate detection and prevention of potential faults in the battery by using intelligent diagnosis equipment with advanced machine learning technology and analysis algorithm. The data such as voltage, current, internal resistance and temperature in the battery charging and discharging process are collected, the data are analyzed by utilizing the deep learning network model, the health state of the battery is monitored, and potential faults are identified in advance. In addition, a reverse neural network is introduced to analyze the historical working state of the failed battery, and the potential cause of the failure is identified. On the basis of a machine learning prediction model, the fault prediction precision and efficiency are improved by using a mutual information optimization strategy. The method can effectively discover and diagnose the internal faults of the lithium ion battery in advance, realize the health management of the battery, improve the safety and reliability of a battery system, and have important significance for promoting the development of electric automobiles and renewable energy storage systems.
Description
Technical Field
The invention belongs to the field of energy sources, and particularly relates to a method for diagnosing internal faults of a lithium ion battery system.
Background
In recent years, lithium ion batteries have been widely focused by academia and industry because of the advantages of high power and energy density, long cycle life, low self-discharge rate, and the like, and have been applied to different scenes such as electronic consumer products, electric automobiles, distributed energy storage, large-scale energy storage, and the like. As a typical energy storage method involving complex electrochemical reaction and transmission mechanism, the lithium ion battery has higher potential safety hazard, and needs to adopt the theory and method of system engineering, pay attention to various layers of battery material system research and development, battery management system design, energy storage system structure optimization and the like, so as to ensure safe, stable and reliable operation in practical use.
However, on one hand, in the practical operation process of the lithium ion battery, the lithium ion battery is limited by the current technical development level of the electric, thermal and safety management system, and mechanical, electric and thermal abuse, such as overcharge, overdischarge, overheat and the like, can occur in some cases, so that the battery performance is easily degraded rapidly, and even internal short circuit occurs to cause safety problems. On the other hand, in the application of electric vehicles, distributed energy storage and large-scale energy storage fields, in order to meet the requirements of current, voltage, power and energy, a large number of monomers are often required to be combined into a battery pack, a battery pack and even a battery cluster through a serial-parallel connection method, a large number of connecting components exist, the complexity of the system is greatly increased, the probability of various faults is increased, and potential safety hazards are increased.
Therefore, on the basis of recognizing the fault triggering mechanism of the battery system, fault diagnosis needs to be implemented on two layers of the lithium ion battery monomer and the system, so that various faults such as gradual change, sudden and the like can be effectively recognized, early warning is realized, and the safety, stability and reliability of the actual operation of the battery system are improved.
Disclosure of Invention
The invention provides a comprehensive method for diagnosing internal faults of a lithium ion battery system. Firstly, the method collects a large amount of operation data of the lithium ion battery, including battery voltage, current, internal resistance, temperature, charging and discharging states of the battery and the like. The data collected is not only from the battery itself, but also includes external environmental conditions such as temperature, humidity, etc. Therefore, the running state of the battery can be monitored in an omnibearing way, and the health condition of the battery can be better estimated.
Secondly, we transform the large amount of data collected above into useful information through feature selection and feature extraction. A feature selection part for selecting features most likely to reflect the health condition of the battery, such as the voltage, the current, the internal resistance and the like of the battery, from a plurality of battery operation parameters; and a feature extraction part for extracting key indexes of the features from the original data, such as voltage fluctuation amplitude, current peak value, internal resistance change trend and the like.
Again, these features will be input into the deep learning model for analysis. The model adopts the most advanced deep learning technology, can carry out self-learning on a large amount of data, and improves the accuracy of fault prediction. The input of the model not only comprises the battery operation characteristics, but also comprises information such as battery historical operation data, battery types, battery use environments and the like, so that the prediction of the model is more accurate.
Finally, when the model predicts that the battery has internal faults, the system immediately gives a warning. The warning modes are various, and can be physical signals such as sound, light and the like, or electronic notification sent out through a network. Meanwhile, the system can automatically take a series of countermeasures according to the type and severity of the fault, such as cutting off the power supply, starting the cooling system, notifying maintenance personnel and the like, so that the influence of the fault on the battery system is reduced as much as possible.
The invention has the advantages that the deep learning technology is introduced into the fault diagnosis of the lithium ion battery for the first time, the accuracy of fault prediction is greatly improved, and the possibility of false alarm is reduced. In addition, the method has comprehensiveness and instantaneity, can monitor the running state of the battery in real time, comprehensively evaluate the health condition of the battery, and has important significance for long-term safe running of the battery.
In addition, the invention also comprises a corresponding fault diagnosis method. The modules of the method respectively correspond to the functions of data acquisition, data processing, health state monitoring, fault early warning and the like. The modules are designed to be very compact, so that the method is small in size and convenient to install.
The technical scheme adopted by the invention is as follows:
the invention provides a method for diagnosing internal faults of a lithium ion battery system.
Step one, extraction and parallel processing of voltage dependence
The use of voltage correlation rather than voltage measurement can effectively avoid problems caused by inconsistencies between cells, and the correlation coefficient is a very suitable measure for this purpose, the invention adjusts the original correlation coefficient to a recursive form suitable for on-line calculation based on the actual cell parameters, the specific adjustment formula being as follows:
wherein P is k ＝P k-1 +x i y i -x i-w y i-w ，Q k ＝Q k-1 +x i -x i-w ，R k ＝R k-1 +y i -y i-w ，x and y represent the measured values of the two voltage sensors, k is the sampling instant and w is the window width.
In this work, all relevant signals are collected as a vector x= [ C ] 1，2 ，C 2，3 ，...C m，1 ] T For parallel processing and x is normalized before use, i.e. processed as vector samples with unity variance and zero mean. If x is a strong non-gaussian distribution, it is first input to a non-metric multidimensionalThe model is scale analyzed to extract non-gaussian portions thereof. Through non-metric multidimensional scale analysis, x can be expressed as an unknown independent component s of a linear combination r (r.ltoreq.m) 1 ，s 2 ，...s γ The formula can be expressed as:
x＝As+e
wherein s= [ s ] 1 ，s 2 ，...s γ ] T ∈R γ ，A＝[a 1 ，a 2 ，...a γ ] T ∈R m×γ Is a mixed matrix, e.epsilon.R m Is a residual matrix.
The objective of the non-metric multidimensional scaling algorithm is to estimate a and s from the observed data x, or to find a separation matrix W to reconstruct the dataSo that the components are independent of each other as much as possible. To eliminate cross-correlation between random variables, a non-metric multidimensional scaling analysis algorithm should be initialized by a whitening process. For the random vector x, the eigenvalue decomposition R is performed on the random vector x x ＝E(xx T ) The specific eigenvalue decomposition and whitening transformation formulas are as follows:
from the nature of eigenvalue decomposition, the following formula can be derived:
from x=as+e, the following formula can be obtained, s without loss of generality k Can be set as unit variance according to E (zz) T )＝BE(ss T )B T ＝BB T The result of I is that B is actually an orthogonal matrix.
Therefore, according to the formula, a non-metric multidimensional scale analysis algorithm can be obtained Separation matrix->Finally, the goal of non-metric multidimensional scaling is reduced to the simpler problem of finding the orthogonal matrix B.
After non-Gaussian part is extracted by adopting a non-metric multidimensional scale analysis algorithm, the residual part (e) is approximately Gaussian distribution, the non-Gaussian part and residual Gaussian noise part can be further fed into a redundancy analysis model to be extracted, the data extracted by the redundancy analysis part is approximately Gaussian distribution, and the residual part can be obtained by adopting redundancy analysis to directly processWherein (1)>
Step two, threshold setting and window width selection
Step 2.1: threshold setting: during the on-line operation of the voltage sensor, a new measurement value is the related signal vector x new After that, the sample data needs to be normalized according to the mean value and standard deviation of the training sample, and then decomposed into three parts through non-metric multidimensional scale analysis and redundancy analysis: non-gaussian part Gaussian partAnd noise part->Wherein,the non-Gaussian fraction is usually composed of +.>Calculated, the Gaussian part is usually composed of +.> Calculating, noise part is usually sub-o>Calculation, I 2 ，T 2 And SPE, can be statistically derived from the sample data. Image T 2 And SPEs satisfy the assumption of a Gaussian distribution, their threshold may be either F-assignment or x 2 Distribution is obtained, and I 2 The calculation cannot be done in the same way as it is not a gaussian distribution, however, as an alternative the invention estimates their thresholds in a uniform way by using an algorithm based on mean shift estimation.
First a sufficient number of statistics I are given 2 The mean shift estimation algorithm is used to estimate the probability density function f in these statistics p (I 2 ). Based on the principle of mean shift estimation algorithm Threshold->For a given confidence limit α, the magnitude of the threshold can be determined by determining the confidence limit, which can be calculated by combining the statistics with the following formula:
step 2.2: window width selection: the window width w has a large effect on the correlation signal, and a small w introduces excessive noise and interference, thereby reducing the sensitivity of the correlation signal to anomalies with a large w. However, the specific value of w is still a pending problem which has not been effectively solved, and only one empirical value can be provided according to the statistical result. To better address this problem, the present invention uses a threshold based on mean shift estimation to coarsely quantify the change in correlation coefficients under fault-free conditions for different window widths, and determines the optimal threshold, referred to as the semi-quantitative preferential criterion, from the evolution trend curve of the threshold with window width. Since the threshold based on the mean shift estimation can be automatically adjusted according to the distribution of the training data, the false alarm rate under the condition of no fault does not exceed the limit (1-alpha) multiplied by 100 percent. Thus, once α is in a given situation, a threshold based on the mean shift estimate can be used to reflect the change in correlation coefficient for different w cases, and w corresponds to the minimum threshold meaning the best suppression of noise and interference by the correlation coefficient at that window width. Thus, the optimal w can be searched for using a threshold versus window width trend curve based on the mean shift estimate. In finding the optimal w, two points need to be noted: (1) In general, the threshold based on mean shift estimation shows a tendency to become smaller as w increases, but when w is very small, the trend curve is not strictly monotonic; (2) As w continuously increases, the trend curve becomes flatter, and in order to avoid an excessive window width decreasing the sensitivity of the correlation coefficient to anomalies, the window width should not be taken too large.
Therefore, the inflection point corresponding to the trend curve no longer oscillates strongly and starts to flatten out should be selected as the optimal window width w. It is worth emphasizing that although the trend curve is not strictly monotonic, the effect of noise and interference certainly gets weaker as the window width increases, and the threshold based on mean shift estimation necessarily gets smaller, so it is always possible to find an inflection point as the choice of the optimal window width.
Third, related coefficient isolation and fault identification
Once the complex statistics calculated in the first two steps are problematic, it can be demonstrated that a fault is detected, and in order to identify the type of fault, the problematic correlation signal is first isolated by improving the contribution graph, the j-th correlation signal being defined asWherein ε j A vector with the j-th element being 1, the other vector elements being 0,/and a method for generating the vector>For the comprehensive statistics of the samples, it is obvious +.>The larger the value of (c), the more likely the jth signal is a problematic signal.
Step 3.1: first, based on training data x, data of all relevant signals are normalized into the synthetic statistic data by the following formula:
step 3.2: calculating the contribution rate of each relevant signal:
step 3.3:after isolating the problematic correlation signals, the type of current fault may be identified according to the following fault identification logic: (1) short circuit fault: if the jth (j=1, 2 …, m) cell experiences an ISC failure, the jth and jth-1 voltage sensors will change simultaneously, resulting in C j-2,j-1 And C j,j+1 For anomalies, while others remain unchanged, the contribution graph intuitively displays the two largest correlation coefficient contribution values of the interval; (2) sensor failure: if the jth (j=1, 2, m) sensor fails, the jth voltage sensor will change, resulting in C j-1,j And C j,j+1 For anomalies, while other values remain unchanged, the contribution graph will visually display the two adjacent maximum correlation coefficient contribution values.
The beneficial effects of the invention are as follows:
the beneficial effects of the invention are as follows:
1. high-precision prediction: by collecting a large amount of battery operation data and applying an advanced deep learning model to conduct fault prediction, the invention can improve the accuracy of fault prediction and reduce the possibility of false alarm.
2. And (3) comprehensively monitoring: the invention not only monitors the running state of the battery, but also considers the influence of the external environment of the battery, thereby being capable of comprehensively evaluating the health condition of the battery and preventing possible problems in advance.
3. Timeliness processing: when the model predicts that the internal fault exists in the battery, the system immediately gives a warning, and a series of countermeasures are automatically taken, so that the influence of the fault on the battery system is reduced as much as possible.
4. Easy implementation: the invention also provides a corresponding fault diagnosis method, and each module of the method is designed to be very compact, so that the method is convenient to implement.
5. The maintainability is strong: because the prediction model can learn by oneself, the prediction capability of the prediction model can be continuously improved along with the time, so that the prediction model can be better suitable for the development and the change of battery technology, and has strong maintainability and adaptability.
In general, the internal fault diagnosis method of the lithium ion battery system can comprehensively, accurately and timely perform battery fault diagnosis, and has important values in the aspects of improving the safety and stability of the battery, prolonging the service life of the battery, reducing the maintenance cost of the battery and the like.
The innovation points of the invention are as follows:
1. a lithium battery pack fault diagnosis model based on non-metric multidimensional scaling analysis (NMDS) and redundancy analysis (RDA) is designed and used for parallel processing of related signals;
2. a semi-quantitative preference criterion based on mean shift estimation is proposed to determine an optimal window width;
3. a multi-fault recognition logic based on an improved contribution graph and a cross-cell sensor topology is presented for intuitively identifying and locating shorts and sensor faults.
Drawings
FIG. 1 is a system flow diagram of the present invention;
FIG. 2 is a flow chart of the data type check of the present invention;
FIG. 3 is a flow chart of health status monitoring in accordance with the present invention;
FIG. 4 is a flow chart of latent fault identification in accordance with the present invention;
fig. 5 is a flowchart of battery data acquisition according to the present invention.
Detailed Description
The present invention will be further described in detail with reference to the following examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Embodiment one:
in one embodiment of a specific method of diagnosing internal faults of a lithium-ion battery system, first, internal operating data and external environmental data of a lithium-ion battery are collected by a series of battery monitoring devices. Such data includes, but is not limited to, the voltage, current, temperature, impedance of the battery, and the ambient temperature, humidity, etc., outside of the battery.
The collected data is then subjected to preliminary processing by a preprocessing module, including steps of cleaning, normalization, feature extraction, etc., to facilitate model input and training. In particular, in the feature extraction step, key features closely related to the battery state and performance are extracted from the raw data.
The preprocessed data is then input into a deep learning model for training and prediction. The model may be a neural network model or other deep learning model. The model can learn and train according to the input data, and predict whether the battery has internal faults according to the learning result.
When the model predicts that the battery has internal faults, the warning module immediately gives a warning, and a series of corresponding measures are taken by the corresponding module, including but not limited to reducing the working current and voltage of the battery, increasing the working strength of the heat dissipation device, or directly stopping the operation of the battery.
Finally, the model feeds back the processing result to the data collection module so as to facilitate further learning and training of the model, thereby realizing self-learning and optimization.
This is just one specific embodiment of the present invention, and in practical application, the specific implementation manner, workflow, and data collection, processing, analysis manner of each module may be adjusted according to specific requirements. .
Embodiment two:
in this embodiment, we apply an improved method of diagnosing internal faults of a lithium ion battery system. First, detailed operating parameters including voltage, current, temperature, impedance, and ambient humidity, etc., inside and outside the battery system are collected by a plurality of sensors and measuring devices. Meanwhile, a data acquisition method is introduced to perform continuous data acquisition.
These collected data are then sent to a preprocessing module for processing, including data cleansing, formatting, normalization, and feature extraction. At this point, feature extraction will rely on a priori knowledge and machine learning techniques to extract key features highly correlated to battery state and performance from raw data.
After the preprocessing is completed, the data is sent to a deep learning model for training and prediction. In this embodiment we use an improved deep learning model that combines convolutional neural networks and long-term short-term memory networks to more accurately process time-series data.
When the model predicts possible internal faults, the warning module will immediately send out warning signals, and then the coping module will immediately start up, and a series of measures are taken, such as reducing the working current and voltage of the battery, increasing the working strength of the heat dissipating device, or directly cutting off the working of the battery, so as to protect the safety of the battery and the system.
Furthermore, after each diagnostic procedure has ended, all of the resulting data will be sent back to the data acquisition module, enabling the model to be further learned and optimized after each diagnostic procedure. This allows for a good self-learning and adaptability of the present diagnostic system.
This is just one specific embodiment of the present invention, and in practical application, the specific implementation manner, workflow, and data collection, processing, analysis manner of each module may be adjusted according to specific requirements.
It should be noted that while the invention has been illustrated and described in terms of its preferred embodiments, it is not intended to limit the scope of the invention. Any person skilled in the art may make various possible variations and modifications without departing from the spirit and scope of the inventive design.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.
Claims (10)
1. The internal fault diagnosis method of the lithium ion battery system is characterized by comprising the following steps of: the method comprises the steps of monitoring the voltage, current, internal resistance and temperature of a battery system in real time through an internal sensor, and collecting related data; according to the collected data types, respectively performing voltage, current, internal resistance and temperature data processing, including data standardization, so as to obtain processed battery data; performing real-time monitoring on the health state of the battery by using the processed battery data, and generating a warning signal if an abnormality occurs in the monitoring result; and further identifying potential faults through a deep learning model according to the abnormal data, determining whether battery faults exist or not, and identifying fault types so as to perform fault early warning and processing.
2. The method of claim 1, wherein the collection of battery data includes initializing sensors of the battery system to enable the sensors to monitor voltage, current, internal resistance, and temperature, respectively, and the four types of data are stored in corresponding data storage areas, respectively, for subsequent processing and analysis.
3. The internal fault diagnosis method of lithium ion battery system according to claim 1 or 2, wherein the processing of the battery data comprises data extraction and data type judgment, and the voltage, current, internal resistance and temperature data are processed differently according to the data type, including data cleaning and standardization processing, so as to reduce noise and abnormal value of the data and improve quality and accuracy of the data.
4. The method for diagnosing internal faults of a lithium ion battery system according to claim 1,2 or 3, wherein the monitoring of the state of health of the battery includes initialization of system monitoring and judgment of monitoring results, and updating the state of the system if the monitoring results indicate that the state of the battery system is normal; if the monitoring result shows that the state of the battery system is abnormal, a warning signal is generated, abnormal data are stored, and data support is provided for subsequent fault identification.
5. The internal fault diagnosis method of a lithium ion battery system according to claim 1,2, 3 or 4, wherein the identification of potential faults comprises the acquisition of abnormal data and the loading of a deep learning model, the abnormal data is predicted through the model, and if the predicted fault exists, the fault type is further identified for fault early warning and processing according to the prediction result; if no faults are predicted, the data is archived for later use.
6. The method for diagnosing internal faults of a lithium ion battery system according to any of claims 1 to 5, wherein the deep learning model is a pre-trained model, including but not limited to a convolutional neural network (Convolutional Neural Networks, CNN), a recurrent neural network (Recurrent Neural Networks, RNN), or a Long Short-Term Memory (LSTM) model, so as to more accurately identify various battery faults.
7. The method for diagnosing an internal fault of a lithium ion battery system according to any one of claims 1 to 6, wherein the data processing step further comprises using feature selection and feature extraction techniques to extract and select feature data that most reflects the battery status, thereby further improving the accuracy and efficiency of fault identification.
8. The method for diagnosing internal faults of a lithium ion battery system according to any of claims 1 to 7, wherein the warning signal can be a sound, light or other perceptible signal, or a remote warning is carried out through a network, so as to inform a user or a maintenance personnel to process in time when the fault of the battery is found.
9. The method for diagnosing internal faults of a lithium ion battery system according to any of claims 1 to 8, wherein when faults are predicted to exist through a deep learning model, after further identifying the fault type, corresponding fault handling measures can be taken, including but not limited to power cut-off, battery cooling, notifying maintenance personnel and the like, so as to reduce possible losses and influences caused by the faults of the battery.
10. The internal fault diagnosis method of the lithium ion battery system is characterized by comprising a data collection module, a data processing module, a health state monitoring module and a deep learning model module, wherein the data collection module is used for collecting voltage, current, internal resistance and temperature data of the battery system; the data processing module is used for processing the collected data; the health state monitoring module is used for monitoring the health state of the battery according to the processed data; the deep learning model module is used for identifying potential faults according to monitoring results, determining whether battery faults exist or not, and identifying fault types so as to perform fault early warning and processing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311039865.4A CN117054887A (en) | 2023-08-17 | 2023-08-17 | Internal fault diagnosis method for lithium ion battery system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311039865.4A CN117054887A (en) | 2023-08-17 | 2023-08-17 | Internal fault diagnosis method for lithium ion battery system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117054887A true CN117054887A (en) | 2023-11-14 |
Family
ID=88660442
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311039865.4A Withdrawn CN117054887A (en) | 2023-08-17 | 2023-08-17 | Internal fault diagnosis method for lithium ion battery system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117054887A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117799498A (en) * | 2024-03-01 | 2024-04-02 | 湘潭南方电机车制造有限公司 | Comprehensive protection system for explosion-proof storage battery electric locomotive |
| CN117825957A (en) * | 2023-11-30 | 2024-04-05 | 重庆大学 | Fault detection method of power battery |
| CN117895119A (en) * | 2024-03-13 | 2024-04-16 | 天芯电子科技(山东)有限公司 | Lithium battery thermal runaway early warning system and method |
| CN117996243A (en) * | 2024-04-07 | 2024-05-07 | 山东浪潮数据库技术有限公司 | Lithium battery energy storage battery pack control method and system |
-
2023
- 2023-08-17 CN CN202311039865.4A patent/CN117054887A/en not_active Withdrawn
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117825957A (en) * | 2023-11-30 | 2024-04-05 | 重庆大学 | Fault detection method of power battery |
| CN117799498A (en) * | 2024-03-01 | 2024-04-02 | 湘潭南方电机车制造有限公司 | Comprehensive protection system for explosion-proof storage battery electric locomotive |
| CN117799498B (en) * | 2024-03-01 | 2024-06-11 | 湘潭南方电机车制造有限公司 | Comprehensive protection system for explosion-proof storage battery electric locomotive |
| CN117895119A (en) * | 2024-03-13 | 2024-04-16 | 天芯电子科技(山东)有限公司 | Lithium battery thermal runaway early warning system and method |
| CN117996243A (en) * | 2024-04-07 | 2024-05-07 | 山东浪潮数据库技术有限公司 | Lithium battery energy storage battery pack control method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Voltage fault diagnosis and prognosis of battery systems based on entropy and Z-score for electric vehicles | |
| CN117054887A (en) | Internal fault diagnosis method for lithium ion battery system | |
| Yao et al. | Fault detection of the connection of lithium-ion power batteries based on entropy for electric vehicles | |
| Sun et al. | An online data-driven fault diagnosis and thermal runaway early warning for electric vehicle batteries | |
| Li et al. | Fault diagnosis for lithium-ion batteries in electric vehicles based on signal decomposition and two-dimensional feature clustering | |
| CN113447828B (en) | Lithium battery temperature estimation method and system based on Bayesian neural network | |
| CN112946522A (en) | On-line monitoring method for short-circuit fault in battery energy storage system caused by low-temperature working condition | |
| CN112014745A (en) | Three-level fault diagnosis method for power battery | |
| CN114559819A (en) | Electric vehicle battery safety early warning method based on signal processing | |
| CN107807860B (en) | Power failure analysis method and system based on matrix decomposition | |
| CN111257753B (en) | Battery system fault diagnosis method | |
| Wang et al. | Voltage correlation-based principal component analysis method for short circuit fault diagnosis of series battery pack | |
| Vianna et al. | Proton exchange membrane fuel cells (PEMFC) impedance estimation using regression analysis | |
| Kim et al. | Outlier mining-based fault diagnosis for multiceli lithium-ion batteries using a low-priced microcontroller | |
| CN114239734A (en) | Distributed vehicle-mounted health management system | |
| Ding et al. | Diagnosis of connection fault for parallel-connected lithium-ion batteries based on long short-term memory networks | |
| Liao et al. | Fault diagnosis of lithium-ion batteries based on wavelet packet decomposition and Manhattan average distance | |
| Gu et al. | An Early minor-fault diagnosis method for lithium-ion battery packs based on unsupervised learning | |
| CN113935423A (en) | Power battery fault early warning method and system for coupling fuzzy control rule with Elman neural network | |
| CN117169761A (en) | Battery state evaluation method, apparatus, device, storage medium, and program product | |
| Wu et al. | Research on short-circuit fault-diagnosis strategy of lithium-ion battery in an energy-storage system based on voltage cosine similarity | |
| Wang et al. | An inconsistency fault diagnosis method for lithium-ion cells in the battery pack based on piecewise dimensionality reduction and outlier identification | |
| Wang et al. | A double-layer fault diagnosis strategy for electric vehicle batteries based on Gaussian mixture model | |
| CN111352365A (en) | Dustproof ventilation type electric power and electrical equipment cabinet and control method | |
| CN116381493A (en) | Battery pack fault detection method, system, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20231114 |