CN117289167A - Battery remaining life prediction method, device and medium based on multiple neural network - Google Patents
Battery remaining life prediction method, device and medium based on multiple neural network Download PDFInfo
- Publication number
- CN117289167A CN117289167A CN202311423969.5A CN202311423969A CN117289167A CN 117289167 A CN117289167 A CN 117289167A CN 202311423969 A CN202311423969 A CN 202311423969A CN 117289167 A CN117289167 A CN 117289167A
- Authority
- CN
- China
- Prior art keywords
- battery
- neural network
- charge
- discharge
- life prediction
- 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.)
- Pending
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 64
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000012549 training Methods 0.000 claims abstract description 51
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 claims abstract description 34
- 229910052744 lithium Inorganic materials 0.000 claims abstract description 34
- 238000012795 verification Methods 0.000 claims abstract description 30
- 238000003062 neural network model Methods 0.000 claims abstract description 16
- 238000007781 pre-processing Methods 0.000 claims abstract description 16
- 230000002068 genetic effect Effects 0.000 claims abstract description 15
- 238000004590 computer program Methods 0.000 claims description 24
- 230000015654 memory Effects 0.000 claims description 22
- 230000010354 integration Effects 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 14
- 238000007600 charging Methods 0.000 claims description 13
- 238000007599 discharging Methods 0.000 claims description 13
- 238000013527 convolutional neural network Methods 0.000 claims description 9
- 230000015556 catabolic process Effects 0.000 claims description 8
- 238000006731 degradation reaction Methods 0.000 claims description 8
- 238000012216 screening Methods 0.000 claims description 7
- 210000000349 chromosome Anatomy 0.000 claims description 6
- 230000035772 mutation Effects 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 4
- 238000010200 validation analysis Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 5
- 229910001416 lithium ion Inorganic materials 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000032683 aging Effects 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- GELKBWJHTRAYNV-UHFFFAOYSA-K lithium iron phosphate Chemical compound [Li+].[Fe+2].[O-]P([O-])([O-])=O GELKBWJHTRAYNV-UHFFFAOYSA-K 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 208000032953 Device battery issue Diseases 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010280 constant potential charging Methods 0.000 description 1
- 238000010277 constant-current charging Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000010439 graphite Substances 0.000 description 1
- 229910002804 graphite Inorganic materials 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000009966 trimming Methods 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]
- G01R31/392—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]
- G01R31/367—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]
- G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The invention discloses a battery remaining life prediction method, a device and a medium based on a multiple neural network, which calculate the remaining life of each charge-discharge period according to charge-discharge parameters of a plurality of lithium batteries in different charge-discharge periods; preprocessing and normalizing the residual life of the lithium battery in each charge-discharge period to construct a training set; inputting the training set into a pre-constructed heterogeneous neural network model to perform model training; drawing a verification set from the training set, removing bad individuals in the verification set by utilizing a genetic algorithm, and reserving an optimal subset as a residual life prediction model of the retired battery; and after preprocessing and normalizing the charge and discharge data of the battery to be tested, inputting the charge and discharge data into the residual life prediction model, and outputting the residual life of the battery to be tested. The method and the device can ensure that the basic model is independent and simultaneously give consideration to prediction accuracy.
Description
Technical Field
The invention relates to the technical field of battery detection, in particular to a battery remaining life prediction method, device and medium based on a multiple neural network.
Background
Compared with the traditional battery, the lithium battery has the advantages of high output voltage, high energy density, small self-discharge, long cycle life, high reliability and the like. These advantages have led to a wider application of lithium ion batteries in more fields. Battery failure may result in reduced performance or failure of the power plant or system, thereby increasing costs. Particularly, if the lithium battery for the electric automobile is poorly managed, fire and explosion can be caused. Therefore, accurately predicting the remaining useful life of a lithium battery plays an increasingly important role in lithium battery state estimation and health management.
The life prediction method of the battery can be mainly divided into a prediction method based on an equivalent circuit model and a data-driven prediction method. Due to the highly complex chemical reactions inside lithium ion batteries, comprehensive internal state data is often difficult to detect and collect, and difficult to use for battery degradation modeling. The state of the lithium ion battery is not only influenced by environmental factors such as working temperature, circuits, loads and the like, but also has strong vulnerability, so that an accurate lithium battery degradation prediction model is difficult to establish. In recent years, a data driving method has attracted a great deal of attention in the field of battery remaining life prediction research based on a multiple neural network. The current data-driven residual life prediction method is mainly realized by applying shallow learning methods such as a support vector machine and deep learning methods such as a deep belief network, a convolutional neural network and a long-term and short-term memory network. However, due to the high sensitivity of deep learning to structural complexity, it is difficult for the existing integration framework to ensure independent basic models and simultaneously consider prediction accuracy.
Disclosure of Invention
Aiming at the defects, the invention provides a battery remaining life prediction method, a device and a medium based on a multiple neural network, which ensure that a basic model is independent and simultaneously give consideration to prediction accuracy.
The embodiment of the invention provides a battery remaining life prediction method based on a multiple neural network, which comprises the following steps:
calculating the residual life of each charge-discharge period according to the charge-discharge parameters of a plurality of lithium batteries in different charge-discharge periods;
preprocessing and normalizing the residual life of the lithium battery in each charge-discharge period to construct a training set;
inputting the training set into a pre-constructed heterogeneous neural network model to perform model training;
drawing a verification set from the training set, removing bad individuals in the verification set by utilizing a genetic algorithm, and reserving an optimal subset as a residual life prediction model of the retired battery;
and after preprocessing and normalizing the charge and discharge data of the battery to be tested, inputting the charge and discharge data into the residual life prediction model, and outputting the residual life of the battery to be tested.
Preferably, the charge and discharge parameters include a charge current, a discharge current, an external temperature, a voltage range, a charge and discharge cycle number, and an actual remaining capacity;
the residual service life RUL=L+1-i of the lithium battery in each charge-discharge period;
wherein L is the total charge-discharge cycle number of the battery which is experienced from the initial degradation to the failure, and i is the current charge-discharge cycle number of the battery.
As a preferable scheme, the heterogeneous neural network model is specifically obtained by heterogeneous integration of a long-short-time memory neural network, a convolutional neural network and a CNN-LSTM combined network.
Preferably, the drawing the verification set from the training set, and using a genetic algorithm to exclude bad individuals in the verification set, and reserving an optimal subset as a residual life prediction model of the retired battery includes:
step 1, generating an initial population by using an algorithm, and extracting an independent verification set from a training set;
step 2, calculating an integrated prediction result of the neural network subset corresponding to each chromosome by adopting an average integration algorithm on the verification set;
step 3, using the inverse of the root mean square error of the prediction result as an fitness function, and obtaining a offspring population through screening, crossing and mutation;
step 4, judging whether the iteration times reach the termination iteration times or not;
if not, returning to the step 2;
if yes, executing the step 5;
and 5, selecting an optimal individual as a residual life prediction model of the retired battery.
The embodiment of the invention provides a battery remaining life prediction device based on a multiple neural network, which comprises:
the data acquisition module is used for calculating the residual life of each charging and discharging period according to the charging and discharging parameters of the plurality of lithium batteries in different charging and discharging periods;
the pretreatment module is used for carrying out pretreatment and normalization on the residual life of the lithium battery in each charge-discharge period, and constructing a training set;
the training module is used for inputting the training set into a pre-constructed heterogeneous neural network model to perform model training;
the preferential module is used for drawing a verification set from the training set, removing bad individuals in the verification set by utilizing a genetic algorithm, and reserving an optimal subset as a residual life prediction model of the retired battery;
the prediction module is used for preprocessing and normalizing charge and discharge data of the battery to be detected, inputting the data into the residual life prediction model and outputting the residual life of the battery to be detected.
Preferably, the charge and discharge parameters include a charge current, a discharge current, an external temperature, a voltage range, a charge and discharge cycle number, and an actual remaining capacity;
the residual service life RUL=L+1-i of the lithium battery in each charge-discharge period;
wherein L is the total charge-discharge cycle number of the battery which is experienced from the initial degradation to the failure, and i is the current charge-discharge cycle number of the battery.
As a preferable scheme, the heterogeneous neural network model is specifically obtained by heterogeneous integration of a long-short-time memory neural network, a convolutional neural network and a CNN-LSTM combined network.
Preferably, the preferential module is specifically configured to perform:
step 1, generating an initial population by using an algorithm, and extracting an independent verification set from a training set;
step 2, calculating an integrated prediction result of the neural network subset corresponding to each chromosome by adopting an average integration algorithm on the verification set;
step 3, using the inverse of the root mean square error of the prediction result as an fitness function, and obtaining a offspring population through screening, crossing and mutation;
step 4, judging whether the iteration times reach the termination iteration times or not;
if not, returning to the step 2;
if yes, executing the step 5;
and 5, selecting an optimal individual as a residual life prediction model of the retired battery.
The embodiment of the invention also provides a battery remaining life prediction device based on a multiple neural network, which comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the battery remaining life prediction method based on the multiple neural network according to any one of the above embodiments.
The embodiment of the invention also provides a computer readable storage medium, which comprises a stored computer program, wherein the computer program controls a device where the computer readable storage medium is located to execute the battery remaining life prediction method based on the multiple neural network according to any one of the above embodiments when running.
According to the battery remaining life prediction method, device and medium based on the multiple neural network, the remaining life of each charging and discharging period is calculated according to the charging and discharging parameters of a plurality of lithium batteries in different charging and discharging periods; preprocessing and normalizing the residual life of the lithium battery in each charge-discharge period to construct a training set; inputting the training set into a pre-constructed heterogeneous neural network model to perform model training; drawing a verification set from the training set, removing bad individuals in the verification set by utilizing a genetic algorithm, and reserving an optimal subset as a residual life prediction model of the retired battery; and after preprocessing and normalizing the charge and discharge data of the battery to be tested, inputting the charge and discharge data into the residual life prediction model, and outputting the residual life of the battery to be tested. The method and the device can ensure that the basic model is independent and simultaneously give consideration to prediction accuracy.
Drawings
Fig. 1 is a schematic flow chart of a method for predicting remaining life of a battery based on a multiple neural network according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a heterogeneous neural network model according to an embodiment of the present invention;
FIG. 3 is a schematic flow chart of selective integration based on genetic algorithm according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a battery remaining life prediction apparatus based on a multiple neural network according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a battery remaining life prediction apparatus based on a multiple neural network according to another embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, a flow chart of a method for predicting remaining life of a battery based on a multiple neural network is provided in an embodiment of the present invention, and the method includes steps S1 to S5:
s1, calculating the residual life of each charge-discharge period according to charge-discharge parameters of a plurality of lithium batteries in different charge-discharge periods;
s2, preprocessing and normalizing the residual life of the lithium battery in each charge-discharge period to construct a training set;
s3, inputting the training set into a pre-constructed heterogeneous neural network model to perform model training;
s4, drawing a verification set from the training set, removing bad individuals in the verification set by utilizing a genetic algorithm, and reserving an optimal subset as a residual life prediction model of the retired battery;
s5, preprocessing and normalizing charge and discharge data of the battery to be tested, inputting the data into the residual life prediction model, and outputting the residual life of the battery to be tested.
In the implementation of this embodiment, data of a plurality of lithium batteries in different charge and discharge cycles are collected, and the remaining life of each charge and discharge cycle is calculated according to the data.
And preprocessing data and residual life of the lithium battery in each charge-discharge period, and converting the value range of all the extracted characteristic values into [0,1] through a normalization method. And constructing a training set, wherein the training set comprises pre-processing the normalized charge and discharge data and the corresponding residual life.
Inputting the training set into a pre-constructed heterogeneous neural network structure model, and performing model training to obtain a trained heterogeneous neural network model.
Drawing a verification set from the training set, removing bad individuals in the candidate set by utilizing a genetic algorithm according to design indexes, and further screening and reserving an optimal subset as a residual life prediction model of the retired battery;
and after preprocessing and normalizing the charge and discharge data of the battery to be tested, inputting the charge and discharge data into the residual life prediction model for life prediction, and outputting the predicted residual life of the retired battery.
The application provides a retired battery residual life prediction method based on multiple deep neural network integration. Firstly, designing a candidate set generation scheme under multi-method combined disturbance, and eliminating the internal coupling relation of a model by adopting a heterogeneous neural network structure and multi-time scale design to strengthen the diversity of candidate deep neural network sets; and then integrating and trimming a redundant model by a genetic algorithm, effectively eliminating redundant learners with poor performance, acquiring diversified optimal candidate subsets, and outputting a prediction result according to average integration. The method enhances the accuracy and the diversity of the integrated model through synchronization, improves the RUL prediction precision, and can provide powerful support for operation and maintenance decision. The prediction accuracy is considered while the independence of the basic models is ensured.
In yet another embodiment provided by the present invention, the charge-discharge parameters include a charge current, a discharge current, an external temperature, a voltage range, a number of charge-discharge cycles, and an actual remaining capacity;
the residual service life RUL=L+1-i of the lithium battery to be tested in each charge-discharge period;
wherein L is the total charge-discharge cycle number of the battery which is experienced from the initial degradation to the failure, and i is the current charge-discharge cycle number of the battery.
When the embodiment is implemented, full life test data of the lithium battery, namely remaining life data of the battery in different charge and discharge cycles, are obtained and used as a training set of a life prediction algorithm.
In the charging process, first constant current charging is performed, the current is set to a constant value, and the voltage is raised to the maximum upper limit. The voltage is kept at a constant value, which is called constant voltage charging. When the voltage is kept at a constant value, the current drops to a certain threshold value.
In the discharging process, the battery is firstly discharged at a constant current with a specific current value until the voltages of different lithium batteries are respectively reduced to specific voltage values.
The obtained charge and discharge data include charge current, discharge current, external temperature, voltage range, charge and discharge cycle number and actual remaining capacity.
The remaining life of the battery in each charge-discharge cycle is calculated as follows:
RUL=L+1-i
wherein L is the total charge-discharge cycle number of the battery which is experienced from the initial degradation to the failure, and i is the current charge-discharge cycle number of the battery.
In still another embodiment of the present invention, the heterogeneous neural network model is specifically obtained by heterogeneous integration of a long-short-term memory neural network, a convolutional neural network and a CNN-LSTM combination network.
In the implementation of this embodiment, referring to fig. 2, a schematic structural diagram of a heterogeneous neural network model provided in an embodiment of the present invention is shown; in terms of diversity of model structures, 3 different deep neural network structures are selected for heterogeneous integration, including long short time memory neural networks (LSTM), convolutional Neural Networks (CNN), and CNN-LSTM combination networks. The CNN is used for extracting deep local features, and then the deep local features are input into the LSTM for long-distance feature extraction. Finally, the features are input into the fully connected layer through transformation to obtain a prediction result.
In terms of time scale diversity, a multi-time scale scheme is employed. A set of alternative time window lengths { T1, T2, T3, & gt, tn }, are preset, and multiple sets of models are trained.
For a test sample of length ti, all subsets satisfying T ti are partitioned { T1, T2, T3,..tm }, where { T1, T2, T3,.. Tm is a subset of { T1, T2, T3,..tn }, then all models in the subset can be used as multiple time scale candidate sets for the test sample.
In yet another embodiment of the present invention, the step S4 specifically includes:
step 1, generating an initial population by using an algorithm, and extracting an independent verification set from a training set;
step 2, calculating an integrated prediction result of the neural network subset corresponding to each chromosome by adopting an average integration algorithm on the verification set;
step 3, using the inverse of the root mean square error of the prediction result as an fitness function, and obtaining a offspring population through screening, crossing and mutation;
step 4, judging whether the iteration times reach the termination iteration times or not;
if not, returning to the step 2;
if yes, executing the step 5;
and 5, selecting an optimal individual as a residual life prediction model of the retired battery.
In the implementation of this embodiment, referring to fig. 3, a schematic flow chart of selective integration based on genetic algorithm according to an embodiment of the present invention is shown; the residual life prediction model determining process specifically comprises the following steps:
an algorithm is used to generate the initial population and separate validation sets are extracted from the training set.
On the verification set, calculating an integrated prediction result of the neural network subset corresponding to each chromosome by adopting an average integration algorithm;
using the inverse of the Root Mean Square Error (RMSE) of the prediction result as an fitness function, and obtaining a offspring population through screening, crossing and mutation;
judging whether the iteration times reach the termination iteration times or not;
if not, repeating the selection process until reaching the end of the iteration algebra.
If yes, selecting the optimal individual as an integrated model of the retired battery residual life prediction model to obtain the residual life prediction model.
The application adopts a high-performance battery test system based on laboratory construction, and comprises a computer, single battery test equipment and an incubator, wherein the aging test is carried out on a lithium ion battery at the constant temperature of 25 ℃ to obtain aging data of the lithium ion battery. The experimental data is derived from the high performance battery test system built in the laboratory and the accelerated aging data of the lithium battery in the NASA database. The experimental battery was a li shen 18650 lithium iron phosphate battery with nominal capacities and voltages of 2.4Ah and 3.6V, respectively. The negative electrode and the positive electrode of the lithium battery are respectively composed of graphite and lithium iron phosphate. The error of the aging test results of the laboratory single lithium battery (short for battery 1) and the NASA single lithium battery (short for battery 2) is within 20 percent.
Compared with the traditional deep learning training model, the method and the device can effectively solve the problem of lack of diversity of the training model, eliminate the limitation of a single model structure and a fixed time scale to the diversity of the prediction model, and remarkably reduce the calculated amount of the parameter algorithm optimizing process. According to the method, the genetic algorithm is utilized to optimize the multiple depth neural network, redundant models in the model set are integrated and deleted, the optimal candidate subset is further screened out in the model set, and the accuracy and the diversity of the life prediction model are greatly improved.
The embodiment of the invention also provides a device for predicting the residual life of a battery based on a multiple neural network, referring to fig. 4, which is a schematic structural diagram of the device for predicting the residual life of a battery based on a multiple neural network, provided by the embodiment of the invention, wherein the device comprises:
the data acquisition module is used for calculating the residual life of each charging and discharging period according to the charging and discharging parameters of the plurality of lithium batteries in different charging and discharging periods;
the pretreatment module is used for carrying out pretreatment and normalization on the residual life of the lithium battery in each charge-discharge period, and constructing a training set;
the training module is used for inputting the training set into a pre-constructed heterogeneous neural network model to perform model training;
the preferential module is used for drawing a verification set from the training set, removing bad individuals in the verification set by utilizing a genetic algorithm, and reserving an optimal subset as a residual life prediction model of the retired battery;
the prediction module is used for preprocessing and normalizing charge and discharge data of the battery to be detected, inputting the data into the residual life prediction model and outputting the residual life of the battery to be detected.
It should be noted that, the battery remaining life prediction device based on a multiple neural network provided in the embodiment of the present invention can execute the battery remaining life prediction method based on a multiple neural network described in any embodiment of the foregoing embodiment, and specific functions of the battery remaining life prediction device based on a multiple neural network are not described herein.
Referring to fig. 5, a schematic structural diagram of a battery remaining life prediction apparatus based on a multiple neural network according to another embodiment of the present invention is shown. The battery remaining life prediction apparatus based on a multiple neural network of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a multi-neural network based battery remaining life prediction program. The processor, when executing the computer program, implements the steps in the embodiments of the method for predicting remaining battery life based on multiple neural networks, for example, steps S1 to S5 shown in fig. 1. Alternatively, the processor may implement the functions of the modules in the above-described device embodiments when executing the computer program.
The computer program may be divided into one or more modules/units, which are stored in the memory and executed by the processor to accomplish the present invention, for example. The one or more modules/units may be a series of computer program instruction segments capable of performing specific functions for describing the execution of the computer program in the multi-neural network-based battery remaining life prediction apparatus. For example, the computer program may be divided into modules, and specific functions of each module are not described herein.
The battery remaining life prediction device based on the multiple neural network can be a desktop computer, a notebook computer, a palm computer, a cloud server and other computing equipment. The multiple neural network-based battery remaining life prediction device may include, but is not limited to, a processor, a memory. It will be appreciated by those skilled in the art that the schematic diagram is merely an example of a multiple neural network-based battery remaining life prediction apparatus, and is not limiting of the multiple neural network-based battery remaining life prediction apparatus, and may include more or less components than those illustrated, or may combine certain components, or different components, e.g., the multiple neural network-based battery remaining life prediction apparatus may further include an input-output device, a network access device, a bus, etc.
The processor may be a central processing unit (Central Processing Unit, CPU), other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. The general-purpose processor may be a microprocessor or the processor may be any conventional processor or the like, which is a control center of the multiple neural network-based battery remaining life prediction apparatus, and connects various parts of the entire multiple neural network-based battery remaining life prediction apparatus using various interfaces and lines.
The memory may be used to store the computer program and/or module, and the processor may implement various functions of the multiple neural network-based battery remaining life prediction apparatus by running or executing the computer program and/or module stored in the memory and invoking data stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program (such as a sound playing function, an image playing function, etc.) required for at least one function, and the like; the storage data area may store data (such as audio data, phonebook, etc.) created according to the use of the handset, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, memory, plug-in hard disk, smart Media Card (SMC), secure Digital (SD) Card, flash Card (Flash Card), at least one disk storage device, flash memory device, or other volatile solid-state storage device.
Wherein the module/unit integrated with the multi-neural network-based battery remaining life prediction apparatus may be stored in a computer-readable storage medium if implemented in the form of a software functional unit and sold or used as a separate product. Based on such understanding, the present invention may implement all or part of the flow of the method of the above embodiment, or may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when the computer program is executed by a processor, the computer program may implement the steps of each of the method embodiments described above. Wherein the computer program comprises computer program code, which may be in the form of code, object code, executable files, or in some intermediate form, etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), an electrical carrier signal, a telecommunications signal, a software distribution medium, and so forth.
While the foregoing is directed to the preferred embodiments of the present invention, it will be appreciated by those skilled in the art that changes and modifications may be made without departing from the principles of the invention, such changes and modifications are also intended to be within the scope of the invention.
Claims (10)
1. A battery remaining life prediction method based on a multiple neural network, the method comprising:
calculating the residual life of each charge-discharge period according to the charge-discharge parameters of a plurality of lithium batteries in different charge-discharge periods;
preprocessing and normalizing the residual life of the lithium battery in each charge-discharge period to construct a training set;
inputting the training set into a pre-constructed heterogeneous neural network model to perform model training;
drawing a verification set from the training set, removing bad individuals in the verification set by utilizing a genetic algorithm, and reserving an optimal subset as a residual life prediction model of the retired battery;
and after preprocessing and normalizing the charge and discharge data of the battery to be tested, inputting the charge and discharge data into the residual life prediction model, and outputting the residual life of the battery to be tested.
2. The method for predicting the remaining life of a battery based on a multiple neural network according to claim 1, wherein the charge and discharge parameters include a charge current, a discharge current, an external temperature, a voltage range, the number of charge and discharge cycles, and an actual remaining capacity;
the residual service life RUL=L+1-i of the lithium battery in each charge-discharge period;
wherein L is the total charge-discharge cycle number of the battery which is experienced from the initial degradation to the failure, and i is the current charge-discharge cycle number of the battery.
3. The method for predicting the remaining life of a battery based on a multiple neural network according to claim 1, wherein the heterogeneous neural network model is obtained by heterogeneous integration of a long-short-term memory neural network, a convolutional neural network and a CNN-LSTM combined network.
4. The method for predicting remaining life of a battery based on a multiple neural network according to claim 1, wherein the steps of decimating a validation set from the training set, and removing bad individuals in the validation set using a genetic algorithm, and reserving an optimal subset as a remaining life prediction model of retired batteries, comprise:
step 1, generating an initial population by using an algorithm, and extracting an independent verification set from a training set;
step 2, calculating an integrated prediction result of the neural network subset corresponding to each chromosome by adopting an average integration algorithm on the verification set;
step 3, using the inverse of the root mean square error of the prediction result as an fitness function, and obtaining a offspring population through screening, crossing and mutation;
step 4, judging whether the iteration times reach the termination iteration times or not;
if not, returning to the step 2;
if yes, executing the step 5;
and 5, selecting an optimal individual as a residual life prediction model of the retired battery.
5. A battery remaining life prediction apparatus based on a multiple neural network, the apparatus comprising:
the data acquisition module is used for calculating the residual life of each charging and discharging period according to the charging and discharging parameters of the plurality of lithium batteries in different charging and discharging periods;
the pretreatment module is used for carrying out pretreatment and normalization on the residual life of the lithium battery in each charge-discharge period, and constructing a training set;
the training module is used for inputting the training set into a pre-constructed heterogeneous neural network model to perform model training;
the preferential module is used for drawing a verification set from the training set, removing bad individuals in the verification set by utilizing a genetic algorithm, and reserving an optimal subset as a residual life prediction model of the retired battery;
the prediction module is used for preprocessing and normalizing charge and discharge data of the battery to be detected, inputting the data into the residual life prediction model and outputting the residual life of the battery to be detected.
6. The multi-neural network-based battery remaining life prediction apparatus of claim 5, wherein the charge and discharge parameters include a charge current, a discharge current, an external temperature, a voltage range, a number of charge and discharge cycles, and an actual remaining capacity;
the residual service life RUL=L+1-i of the lithium battery in each charge-discharge period;
wherein L is the total charge-discharge cycle number of the battery which is experienced from the initial degradation to the failure, and i is the current charge-discharge cycle number of the battery.
7. The battery remaining life prediction device based on the multiple neural network according to claim 5, wherein the heterogeneous neural network model is obtained by heterogeneous integration of a long-short-term memory neural network, a convolutional neural network and a CNN-LSTM combined network.
8. The multi-neural network-based battery remaining life prediction apparatus of claim 5, wherein the preferential module is specifically configured to perform:
step 1, generating an initial population by using an algorithm, and extracting an independent verification set from a training set;
step 2, calculating an integrated prediction result of the neural network subset corresponding to each chromosome by adopting an average integration algorithm on the verification set;
step 3, using the inverse of the root mean square error of the prediction result as an fitness function, and obtaining a offspring population through screening, crossing and mutation;
step 4, judging whether the iteration times reach the termination iteration times or not;
if not, returning to the step 2;
if yes, executing the step 5;
and 5, selecting an optimal individual as a residual life prediction model of the retired battery.
9. A multi-neural network-based battery remaining life prediction apparatus comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the multi-neural network-based battery remaining life prediction method of any one of claims 1 to 4 when the computer program is executed.
10. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored computer program, wherein the computer program when run controls a device in which the computer readable storage medium is located to perform the multi-neural network based battery remaining life prediction method according to any one of claims 1 to 4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311423969.5A CN117289167A (en) | 2023-10-27 | 2023-10-27 | Battery remaining life prediction method, device and medium based on multiple neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311423969.5A CN117289167A (en) | 2023-10-27 | 2023-10-27 | Battery remaining life prediction method, device and medium based on multiple neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117289167A true CN117289167A (en) | 2023-12-26 |
Family
ID=89244518
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311423969.5A Pending CN117289167A (en) | 2023-10-27 | 2023-10-27 | Battery remaining life prediction method, device and medium based on multiple neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117289167A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117709205A (en) * | 2024-02-05 | 2024-03-15 | 华南师范大学 | Method, device, equipment and medium for predicting residual service life of aero-engine |
-
2023
- 2023-10-27 CN CN202311423969.5A patent/CN117289167A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117709205A (en) * | 2024-02-05 | 2024-03-15 | 华南师范大学 | Method, device, equipment and medium for predicting residual service life of aero-engine |
CN117709205B (en) * | 2024-02-05 | 2024-05-07 | 华南师范大学 | Method, device, equipment and medium for predicting residual service life of aero-engine |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Li et al. | Lithium-ion battery capacity estimation—A pruned convolutional neural network approach assisted with transfer learning | |
CN110224192B (en) | Method for predicting service life of power battery by gradient utilization | |
CN107957562B (en) | Online prediction method for residual life of lithium ion battery | |
CN109324291B (en) | Prediction method for predicting service life of proton exchange membrane fuel cell | |
Liu et al. | Future ageing trajectory prediction for lithium-ion battery considering the knee point effect | |
CN109001640B (en) | Data processing method and device for power battery | |
CN107132490B (en) | Method for estimating state of charge of lithium battery pack | |
CN110082682B (en) | Lithium battery state of charge estimation method | |
CN108009585B (en) | Lead-acid battery health state prediction method based on local information fusion | |
CN117289167A (en) | Battery remaining life prediction method, device and medium based on multiple neural network | |
CN113125960A (en) | Vehicle-mounted lithium ion battery charge state prediction method based on random forest model | |
CN109800446B (en) | Method and device for estimating voltage inconsistency in discharging process of lithium ion battery | |
CN112686380A (en) | Neural network-based echelon power cell consistency evaluation method and system | |
CN111983474A (en) | Lithium ion battery life prediction method and system based on capacity decline model | |
CN115508711B (en) | Electric vehicle power lithium battery residual life prediction method based on deep network model | |
CN112666479A (en) | Battery service life prediction method based on charging cycle fusion | |
CN113484774A (en) | Lithium battery pack capacity estimation method based on temperature calibration and neural network | |
CN114578234A (en) | Lithium ion battery degradation and capacity prediction model considering causality characteristics | |
Sheikh et al. | A data-driven comparative analysis of lithium-ion battery state of health and capacity estimation | |
CN113657030A (en) | Method for predicting remaining service life of lithium battery based on Gaussian process regression | |
CN112084459A (en) | Method and device for predicting battery charge-discharge cycle life, electronic terminal and storage medium | |
Song et al. | Capacity estimation method of lithium-ion batteries based on deep convolution neural network | |
Lu et al. | Deep learning to predict battery voltage behavior after uncertain cycling-induced degradation | |
CN115542168A (en) | Lithium battery residual service life prediction method based on fusion data driving model | |
CN115291113A (en) | Unified joint estimation method and system for SOC, SOH and RUL of lithium ion battery |
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 |