CN115508732A - Battery pack service life prediction method and device - Google Patents
Battery pack service life prediction method and device Download PDFInfo
- Publication number
- CN115508732A CN115508732A CN202211274814.5A CN202211274814A CN115508732A CN 115508732 A CN115508732 A CN 115508732A CN 202211274814 A CN202211274814 A CN 202211274814A CN 115508732 A CN115508732 A CN 115508732A
- Authority
- CN
- China
- Prior art keywords
- service life
- battery pack
- data
- life prediction
- 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
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000012549 training Methods 0.000 claims description 52
- 238000012795 verification Methods 0.000 claims description 43
- 238000010801 machine learning Methods 0.000 claims description 31
- 238000012545 processing Methods 0.000 claims description 25
- 238000012360 testing method Methods 0.000 claims description 22
- 238000010606 normalization Methods 0.000 claims description 18
- 238000000605 extraction Methods 0.000 claims description 17
- 238000004458 analytical method Methods 0.000 claims description 16
- 238000001914 filtration Methods 0.000 claims description 16
- 238000004140 cleaning Methods 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 12
- 238000010219 correlation analysis Methods 0.000 claims description 12
- 230000002159 abnormal effect Effects 0.000 claims description 10
- 238000013101 initial test Methods 0.000 claims description 10
- 238000012544 monitoring process Methods 0.000 claims description 9
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 230000000306 recurrent effect Effects 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 2
- 238000012423 maintenance Methods 0.000 abstract description 13
- 238000004364 calculation method Methods 0.000 abstract description 10
- 238000007726 management method Methods 0.000 description 15
- 230000003993 interaction Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 238000005259 measurement Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000007599 discharging Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012821 model calculation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000000275 quality assurance Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000022131 cell cycle Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013524 data verification Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The application provides a method and a device for predicting the service life of a battery pack, wherein the method comprises the following steps: a battery pack service life prediction model is constructed in advance; acquiring a service life prediction instruction and a service life prediction characteristic value of a battery pack input by a user; calling a battery pack service life prediction model according to the battery pack service life prediction instruction and the service life prediction characteristic value to predict the service life of the battery pack to obtain a prediction result; and outputting the prediction result, and optimizing the battery management strategy based on the prediction result. Therefore, the method can accurately and quickly realize the service life prediction of the battery pack, has low cost, high calculation speed and high prediction accuracy, can optimize the battery management strategy in time, can intervene in advance to resolve risks for the battery pack with the risks, and has positive significance for the service life maintenance of the battery and the public praise maintenance of products.
Description
Technical Field
The application relates to the technical field of data processing, in particular to a method and a device for predicting service life of a battery pack.
Background
In recent years, with the increasing energy crisis, new energy automobiles have become the development focus of the automobile industry in the future due to the excellent energy-saving and environment-friendly characteristics of the new energy automobiles. The battery life of the battery pack carried on the new energy automobile directly influences the performance and the running condition of the new energy automobile, so that the prediction of the battery life of the battery pack becomes a crucial link in the research of the new energy automobile. The existing battery pack service life prediction method is generally based on a battery cell cycle service life test and an empirical degradation model, the cycle number and the discharge capacity corresponding to each week are counted, the discharge capacity data from the m time to the n time in the cycle service life test are selected to carry out fitting analysis, the value of a coefficient to be determined in the empirical model is solved to obtain a cycle service life prediction model, the discharge capacity at the end of the service life of the battery is substituted into the model according to the specified cycle service life termination condition of the battery to obtain the cycle number, and therefore the cycle service life of the battery is predicted. However, in practice, it is found that the existing battery cell test has high cost and long period, the difference between the battery cell test working condition and the market vehicle working condition is large, and a large error exists in the service life of the whole battery cell extrapolated according to the service life of the battery cell. Therefore, the existing method has the disadvantages of high cost, long period and large error.
Disclosure of Invention
The embodiment of the application aims to provide a battery pack service life prediction method and device, which can accurately and quickly realize battery pack service life prediction, are low in cost, high in calculation speed and high in prediction accuracy, can intervene in the risk resolution of a battery pack with a risk in advance, and have positive significance on battery service life maintenance and product public praise maintenance.
A first aspect of an embodiment of the present application provides a method for predicting a life of a battery pack, including:
a battery pack service life prediction model is constructed in advance;
acquiring a service life prediction instruction and a service life prediction characteristic value of a battery pack input by a user;
calling the battery pack service life prediction model according to the battery pack service life prediction instruction and the service life prediction characteristic value to predict the service life of the battery pack to obtain a prediction result;
and outputting the prediction result, and optimizing the battery management strategy based on the prediction result.
In the implementation process, the method can be used for constructing a battery pack service life prediction model in advance; then acquiring a service life prediction instruction and a service life prediction characteristic value of the battery pack input by a user; then, calling a battery pack service life prediction model according to the battery pack service life prediction instruction and the service life prediction characteristic value to predict the service life of the battery pack to obtain a prediction result; and finally, outputting the prediction result, and optimizing the battery management strategy based on the prediction result. Therefore, the implementation of the implementation mode can accurately and quickly realize the service life prediction of the battery pack, has low cost, high calculation speed and high prediction accuracy, can intervene in the risk resolution of the battery pack with the risk in advance, and has positive significance on the service life maintenance of the battery and the public praise maintenance of products.
Further, the pre-building a battery pack life prediction model includes:
capturing specified sample data from a new energy remote monitoring big data platform;
performing data processing on the sample data to obtain processed target sample data;
and training a pre-deployed machine learning model through the target sample data to obtain a battery pack service life prediction model.
Further, the performing data processing on the sample data to obtain processed target sample data includes:
performing data cleaning on the sample data to remove abnormal data and low-value data to obtain cleaned data;
performing feature extraction on the cleaned data to obtain feature data;
performing characteristic correlation analysis on the characteristic data to obtain an analysis result;
and performing characteristic filtering on the characteristic data according to the analysis result to obtain target sample data.
Further, the training of the pre-deployed machine learning model through the target sample data to obtain a battery pack life prediction model includes:
a machine learning model is deployed in advance by adopting a recurrent neural network; wherein the model unit of the machine learning model is a door control cycle unit;
dividing the target sample data into an initial training set, an initial verification set and an initial test set according to a preset division ratio;
respectively carrying out normalization processing on the initial training set, the initial verification set and the initial test set to obtain a training set, a verification set and a test set;
training the machine learning model according to a pre-constructed loss function and the training set to obtain a trained model;
performing model precision verification on the trained model through the test set and the verification set to obtain a verification result;
and when the verification result meets a preset precision threshold, determining the trained model as a battery pack service life prediction model.
Further, the calling the battery pack life prediction model according to the battery pack life prediction instruction and the life prediction characteristic value to predict the life of the battery pack to obtain a prediction result includes:
normalizing the service life prediction characteristic value according to the service life prediction instruction of the battery pack to obtain a normalized characteristic value;
calling the battery pack service life prediction model according to the normalized characteristic value to predict the service life of the battery pack to obtain a preliminary prediction result;
and performing inverse normalization processing on the preliminary prediction result to obtain a final prediction result.
A second aspect of the embodiments of the present application provides a battery pack life prediction apparatus, including:
the model building unit is used for building a battery pack service life prediction model in advance;
the system comprises an acquisition unit, a service life prediction unit and a service life prediction characteristic value, wherein the acquisition unit is used for acquiring a service life prediction instruction and a service life prediction characteristic value of a battery pack input by a user;
the prediction unit is used for calling the battery pack service life prediction model to predict the service life of the battery pack according to the battery pack service life prediction instruction and the service life prediction characteristic value to obtain a prediction result;
and the output unit is used for outputting the prediction result and optimizing the battery management strategy based on the prediction result.
In the implementation process, the device can pre-construct a battery pack service life prediction model through a model construction unit; acquiring a service life prediction instruction and a service life prediction characteristic value of a battery pack input by a user through an acquisition unit; calling a battery pack service life prediction model to predict the service life of the battery pack according to the battery pack service life prediction instruction and the service life prediction characteristic value through a prediction unit to obtain a prediction result; and outputting the prediction result through an output unit, and optimizing the battery management strategy based on the prediction result. Therefore, the implementation of the implementation mode can accurately and quickly realize the service life prediction of the battery pack, has low cost, high calculation speed and high prediction accuracy, can intervene in the risk resolution of the battery pack with the risk in advance, and has positive significance on the service life maintenance of the battery and the public praise maintenance of products.
Further, the model building unit includes:
the capturing subunit is used for capturing specified sample data from the new energy remote monitoring big data platform;
the processing subunit is used for carrying out data processing on the sample data to obtain processed target sample data;
and the training subunit is used for training a pre-deployed machine learning model through the target sample data to obtain a battery pack service life prediction model.
Further, the processing subunit includes:
the cleaning module is used for cleaning the data of the sample data to remove abnormal data and low-value data and obtain cleaned data;
the extraction module is used for extracting the characteristics of the cleaned data to obtain characteristic data;
the analysis module is used for carrying out characteristic correlation analysis on the characteristic data to obtain an analysis result;
and the filtering module is used for carrying out characteristic filtering on the characteristic data according to the analysis result to obtain target sample data.
Further, the training subunit includes:
the deployment module is used for adopting a recurrent neural network to deploy a machine learning model in advance; wherein the model unit of the machine learning model is a door control cycle unit;
the dividing module is used for dividing the target sample data into an initial training set, an initial verification set and an initial test set according to a preset dividing proportion;
the normalization module is used for respectively carrying out normalization processing on the initial training set, the initial verification set and the initial test set to obtain a training set, a verification set and a test set;
the training module is used for training the machine learning model according to a pre-constructed loss function and the training set to obtain a trained model;
the verification module is used for performing model precision verification on the trained model through the test set and the verification set to obtain a verification result;
and the determining module is used for determining the trained model as a battery pack service life prediction model when the verification result meets a preset precision threshold.
Further, the prediction unit includes:
the normalizing subunit is used for normalizing the service life prediction characteristic value according to the service life prediction instruction of the battery pack to obtain a normalized characteristic value;
the predicting subunit is used for calling the battery pack service life predicting model according to the normalized characteristic value to predict the service life of the battery pack to obtain a preliminary predicting result;
and the inverse normalization subunit is used for performing inverse normalization processing on the preliminary prediction result to obtain a final prediction result.
A third aspect of the embodiments of the present application provides an electronic device, including a memory and a processor, where the memory is used to store a computer program, and the processor runs the computer program to make the electronic device execute the method for predicting the life of a battery pack according to any one of the first aspect of the embodiments of the present application.
A fourth aspect of the present embodiment provides a computer-readable storage medium, which stores computer program instructions, where the computer program instructions, when read and executed by a processor, perform the method for predicting the life of a battery pack according to any one of the first aspect of the present embodiment.
Drawings
To more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
Fig. 1 is a schematic flowchart of a method for predicting a life of a battery pack according to an embodiment of the present disclosure;
fig. 2 is a schematic structural diagram of another device for predicting life of a battery pack according to an embodiment of the present disclosure;
FIG. 3 is a schematic diagram of an error curve of a training process according to an embodiment of the present application;
fig. 4 is a schematic diagram illustrating an error comparison between a predicted capacity and an actual capacity according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures. Meanwhile, in the description of the present application, the terms "first", "second", and the like are used only for distinguishing the description, and are not construed as indicating or implying relative importance.
Example 1
Referring to fig. 1, fig. 1 is a schematic flowchart illustrating a method for predicting a lifetime of a battery pack according to an embodiment of the present disclosure. The battery pack life prediction method comprises the following steps:
s101, capturing specified sample data from a new energy remote monitoring big data platform.
And S102, data cleaning is carried out on the sample data to remove abnormal data and low-value data, and cleaned data are obtained.
And S103, performing feature extraction on the cleaned data to obtain feature data.
And S104, performing characteristic correlation analysis on the characteristic data to obtain an analysis result.
And S105, performing feature filtering on the feature data according to the analysis result to obtain target sample data.
S106, adopting a recurrent neural network to pre-deploy a machine learning model; wherein the model unit of the machine learning model is a gate control cycle unit.
And S107, dividing the target sample data into an initial training set, an initial verification set and an initial test set according to a preset dividing ratio.
And S108, respectively carrying out normalization processing on the initial training set, the initial verification set and the initial test set to obtain a training set, a verification set and a test set.
And S109, training the machine learning model according to the pre-constructed loss function and the training set to obtain the trained model.
And S110, performing model precision verification on the trained model through the test set and the verification set to obtain a verification result.
And S111, when the verification result meets a preset precision threshold, determining the trained model as a battery pack service life prediction model.
And S112, acquiring a service life prediction instruction and a service life prediction characteristic value of the battery pack input by the user.
In this embodiment, when it is necessary to evaluate the performance of the battery pack after a certain policy parameter is changed, such as reducing the number of deep charging times, reducing the number of fast charging times, reducing the charging rate, and other policy changes, by defining each characteristic value after the policy change, each characteristic value may be a battery pack life prediction instruction and a life prediction characteristic value.
And S113, normalizing the service life prediction characteristic value according to the service life prediction instruction of the battery pack to obtain a normalized characteristic value.
And S114, calling a battery pack service life prediction model according to the normalized characteristic value to predict the service life of the battery pack, and obtaining a preliminary prediction result.
And S115, performing inverse normalization processing on the preliminary prediction result to obtain a final prediction result.
And S116, outputting a prediction result, and optimizing a battery management strategy based on the prediction result.
In this embodiment, each feature value is input into the model to be calculated, so that a prediction result can be obtained, where the prediction result includes the battery pack life after the policy is changed. Based on the prediction results under different strategy parameters, different battery life performances can be obtained, and then the battery management strategy can be continuously optimized according to the prediction results, so that the method has an important effect on the optimization of the battery pack full-life-cycle management strategy.
In this embodiment, the method may be applied to a battery pack life prediction maintenance system based on big data and machine learning, and the system may include a data acquisition module, a feature engineering module, a machine learning module, a prediction update module, and a human-computer interaction module. The data acquisition module is connected with the new energy remote monitoring big data platform and is responsible for capturing specified sample data from the big data platform; the characteristic engineering module is connected with the data acquisition module, and receives the sample data and performs data cleaning, characteristic extraction, characteristic correlation analysis and characteristic filtering; the machine learning module is connected with the feature engineering module, receives the processed sample data and is responsible for deployment training and verification of the machine learning model; the prediction updating module is connected with the machine learning module and the human-computer interaction module, a user inputs characteristic parameters through the human-computer interaction module, the prediction updating module directly calls a trained model in the machine learning module to perform prediction calculation, and a visual result is output through the human-computer interaction module.
In this embodiment, the data acquisition module can access a new energy remote monitoring big data platform, and acquire required original feature data by means of computing power of the server; then the data flow is to the characteristic engineering module, carry on data cleaning, characteristic extraction, characteristic correlation analysis and characteristic filtration; then the data flow to a machine learning module, the work of model building, model training, model test evaluation, actual measurement data verification and the like is completed, a GRU neural network model is built at first, sample data is divided into a training set, a verification set and a test set according to a certain proportion, model training and test evaluation are carried out, the hyper-parameters of the model are dynamically adjusted according to error results, the optimal model is finally determined, then the model prediction results are verified based on multiple groups of effective actual measurement data, the average error is 1.05%, and the high-precision advantage of the model for predicting the service life of the battery is proved; and (3) inputting a corresponding characteristic value through a human-computer interaction interface by a user, starting a prediction updating module, calling the trained machine learning model for calculation, obtaining the predicted service life of the battery pack, and displaying the result on the human-computer interaction interface. By defining the characteristic value, the service life predicted value of the battery pack in the characteristic state can be obtained, and the battery pack with abnormal service life attenuation can be found in advance; based on the battery life performance under different strategy parameters, the battery management strategy can be continuously optimized.
In this embodiment, the data acquisition module is responsible for acquiring original sample data, and can access data resources in the new energy remote monitoring big data platform, and acquire the required sample data through a customized feature extraction algorithm and computing resources of the server. The method comprises the steps of taking all vehicles on line of a certain vehicle type carrying a certain battery pack as research objects, extracting charging and running data of the vehicle type from the time of putting operation to a certain time node, taking continuous charging and running data of each time as a sub-segment, and extracting the characteristics of each charging and discharging.
In this embodiment, the feature engineering module is responsible for data cleaning, feature extraction, feature correlation analysis, and feature filtering of sample data.
Specifically, the functions of the feature engineering module include:
(1) data cleaning:
the data cleaning comprises two aspects, namely removing samples containing abnormal values, such as abnormal mileage and data missing, and removing low-value samples, such as vehicles with few operation days, vehicles with low total mileage, samples with low charging depth and the like.
(2) Feature extraction:
the characteristic extraction is carried out in two steps, firstly, the characteristic extraction is carried out according to the day, the charging and discharging segments of each vehicle for many times each day are subjected to characteristic statistics, and the data of each vehicle for each day are summarized into a sample; then extracting according to stages, taking each vehicle as an example, counting the effective operation days, namely the days of charging and driving behaviors, sorting according to dates, averagely dividing the sample data of each vehicle into n sections (n is suggested to be 2-6), wherein the initial date of each stage is the first day, so that the influence of the historical charging and discharging behaviors on the service life of the battery pack is kept in each stage of data, for example, the effective days of a certain vehicle is 300 days, and the data are divided into 6 stages, the 1 st section is the 1 st day to the 50 th day, the 2 nd section is the 1 st day to the 100 th day, the 3 rd section is the 1 st day to the 150 th day, and the like.
The extraction of the target characteristic battery capacity is also divided into two steps of extraction by days and extraction by stages, firstly, the ampere-hour integration is carried out on each charging segment to obtain the charging capacity CHAh, and further the current capacity Q = (CHAh 100)/(CHsoc factor) is obtained, wherein CHsoc is the charging depth, and factor is a conversion coefficient and is related to mileage. When extracting according to the day, the capacity abnormal value is removed, and then the average value of the capacity is calculated to be used as the capacity result of a single day. In the case of extraction by stage, the average capacity (for example, m = 15) of m days at the end of each stage is calculated as the capacity result of each stage.
(3) And (3) characteristic correlation analysis:
the characteristic correlation analysis is to calculate the Pearson correlation coefficient between every two characteristics, and the correlation analysis result is used as the important basis of characteristic filtering.
(4) Characteristic filtering:
the feature filtering is to remove some low-value features, which can cause model overfitting and reduce accuracy on one hand and increase model training time on the other hand. Feature filtering follows several principles: the method comprises the steps of eliminating the features with the variance of 0, eliminating the features with the bias or the peak absolute value larger, eliminating the features with the minimum correlation with a target, and reserving one feature when the correlation coefficient of the two features is higher than a threshold value.
In this embodiment, the machine learning module is responsible for model building, model training, model testing, and actual measurement verification.
Specifically, the functions of the feature engineering module include:
(1) building a model:
for modeling of a time series type, a Recurrent Neural Network (RNN) is mostly adopted, and the RNN has certain memory capacity and can sequentially process information with any length according to a time sequence. Compared with the LSTM, the GRU has one less 'gating' parameter, can achieve the effect equivalent to the LSTM, and can effectively reduce the calculation cost and the time cost, so the GRU is selected as the model unit.
(2) Model training:
before training, data division and data normalization are required to be carried out on sample data, in the embodiment, a training set, a verification set and a test set are divided according to the proportion of 7.5. The loss function used for training is the mean square error MSE, which is as follows:
wherein MSE represents the loss function,
indicates the prediction result, Y i Denotes the actual value, i denotes the sequence number, and n denotes the maximum value of the sequence number. The error curve of the training process is shown in fig. 3, and the two curves represent the error results of the training set and the validation set respectively.
(3) And (3) testing a model:
after the model training is finished, the test set is used for verifying the model precision, and the evaluation is carried out through qualitative dimension and quantitative dimension, specifically, the qualitative dimension is the goodness of fit of the observation test set real value and the model predicted value curve, the quantitative dimension is two parameters of mean square error and mean error of the observation real value and the predicted value, and the curve fit and the error are small, so that the high model precision is indicated.
(4) And (3) actual measurement verification:
in order to verify the accuracy of the model more fully, 20 groups of capacity data actually measured by different vehicles and different mileage are adopted for verification, for each actually measured sample, each characteristic value in a corresponding state is firstly extracted and input into the model to generate predicted capacity, and the error between the predicted capacity and the actually measured capacity is compared, specifically, as shown in fig. 4, the average error of all the actually measured samples is 1.05%, which indicates that the model has high prediction accuracy.
In this embodiment, the prediction update module is responsible for calling the trained model to perform calculation and outputting the prediction result. And the prediction updating module receives each characteristic value transmitted by the human-computer interaction module, calls model calculation after normalization, performs inverse normalization on the calculation result to obtain a final prediction result, and outputs the final prediction result to the human-computer interaction module.
In this embodiment, the human-computer interaction module is responsible for receiving information input by a user, displaying a model calculation result to the user, and providing results comparing, recording, storing and drawing functions. For example, the battery pack capacity of a certain type of vehicle after 5 years needs to be evaluated, the charging habits and driving habits of the vehicle for several months are analyzed, the characteristic average value of each month of the vehicle is extracted, corresponding characteristic values after 5 years can be obtained through equal proportion conversion, and the characteristic values are input through a human-computer interaction module, so that the predicted capacity can be obtained in real time.
For example, the life performance of the battery pack after a certain strategy parameter is changed needs to be evaluated, such as strategy changes of reducing deep charge times, reducing fast charge times, reducing charge multiplying power and the like, each characteristic value after the strategy changes is defined and is input into the model for calculation, the life of the battery pack after the strategy changes can be obtained, the battery management strategy can be continuously optimized based on the life performance of the battery under different strategy parameters, and the method plays an important role in optimizing the full life cycle management strategy of the battery pack.
In this embodiment, the execution subject of the method may be a computing device such as a computer and a server, and is not limited in this embodiment.
In this embodiment, an execution subject of the method may also be an intelligent device such as a smart phone and a tablet computer, which is not limited in this embodiment.
Therefore, the battery pack service life prediction method described in the embodiment can be used for evaluating whether the performance of the service life of a certain battery pack meets the quality assurance requirement, so that the method has positive significance on battery service life maintenance and product public praise maintenance; the influence of each influence factor on the service life of the battery can be quantitatively analyzed, so that a user can obtain the service life result of the battery pack only by inputting parameter values related to the strategy through the human-computer interaction module, the battery service life performance based on different strategy parameters is facilitated, the battery management strategy is continuously optimized, and the optimization of the full life cycle management strategy of the battery pack is facilitated.
Example 2
Referring to fig. 2, fig. 2 is a schematic structural diagram of a device for predicting a lifetime of a battery pack according to an embodiment of the present disclosure. As shown in fig. 2, the battery pack life prediction apparatus includes:
a model building unit 210, configured to pre-build a battery pack life prediction model;
an obtaining unit 220, configured to obtain a battery pack life prediction instruction and a life prediction characteristic value input by a user;
the prediction unit 230 is configured to call a battery pack life prediction model according to the battery pack life prediction instruction and the life prediction characteristic value to predict the life of the battery pack, so as to obtain a prediction result;
and an output unit 240 for outputting the prediction result and optimizing the battery management strategy based on the prediction result.
As an alternative embodiment, the model building unit 210 includes:
the grasping subunit 211 is configured to grasp specified sample data from the new energy remote monitoring big data platform;
a processing subunit 212, configured to perform data processing on the sample data to obtain processed target sample data;
and the training subunit 213 is configured to train the pre-deployed machine learning model through the target sample data, so as to obtain a battery pack life prediction model.
As an alternative embodiment, the processing subunit 212 includes:
the cleaning module is used for cleaning the data of the sample data so as to remove abnormal data and low-value data and obtain the cleaned data;
the extraction module is used for extracting the characteristics of the cleaned data to obtain characteristic data;
the analysis module is used for carrying out characteristic correlation analysis on the characteristic data to obtain an analysis result;
and the filtering module is used for performing characteristic filtering on the characteristic data according to the analysis result to obtain target sample data.
As an alternative embodiment, the training subunit 213 includes:
the deployment module is used for deploying the machine learning model in advance by adopting a recurrent neural network; wherein the model unit of the machine learning model is a door control cycle unit;
the dividing module is used for dividing target sample data into an initial training set, an initial verification set and an initial test set according to a preset dividing ratio;
the normalization module is used for respectively carrying out normalization processing on the initial training set, the initial verification set and the initial test set to obtain a training set, a verification set and a test set;
the training module is used for training the machine learning model according to a pre-constructed loss function and a training set to obtain a trained model;
the verification module is used for performing model precision verification on the trained model through the test set and the verification set to obtain a verification result;
and the determining module is used for determining the trained model as the battery pack service life prediction model when the verification result meets a preset precision threshold.
As an alternative embodiment, the prediction unit 230 includes:
the normalizing subunit 231 is configured to perform normalization processing on the life prediction feature value according to the battery pack life prediction instruction to obtain a normalized feature value;
the predicting subunit 232 is configured to invoke a battery pack life prediction model according to the normalized feature value to perform battery pack life prediction, so as to obtain a preliminary prediction result;
and an inverse normalization subunit 233, configured to perform inverse normalization processing on the preliminary prediction result to obtain a final prediction result.
In this embodiment, for the explanation of the device for predicting the lifetime of a battery pack, reference may be made to the description in embodiment 1, and details are not repeated in this embodiment.
Therefore, the battery pack service life prediction device described in the embodiment can evaluate whether the performance of the service life of a certain battery pack meets the quality assurance requirement, so that the battery pack service life prediction device has positive significance for battery service life maintenance and product public praise maintenance; the influence of each influence factor on the service life of the battery can be quantitatively analyzed, so that a user can obtain the service life result of the battery pack only by inputting parameter values related to the strategy through the human-computer interaction module, thereby being beneficial to the performance of the service life of the battery based on different strategy parameters, continuously optimizing the battery management strategy and further being beneficial to the optimization of the full life cycle management strategy of the battery pack.
An embodiment of the present application provides an electronic device, which includes a memory and a processor, where the memory is used to store a computer program, and the processor runs the computer program to make the electronic device execute the method for predicting the life of a battery pack in embodiment 1 of the present application.
An embodiment of the present application provides a computer-readable storage medium, which stores computer program instructions, where the computer program instructions, when read and executed by a processor, perform the method for predicting the life of a battery pack in embodiment 1 of the present application.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solutions of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application. It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily think of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises the element.
Claims (10)
1. A method for predicting a life of a battery pack, comprising:
a battery pack service life prediction model is constructed in advance;
acquiring a service life prediction instruction and a service life prediction characteristic value of a battery pack input by a user;
calling the battery pack service life prediction model according to the battery pack service life prediction instruction and the service life prediction characteristic value to predict the service life of the battery pack to obtain a prediction result;
and outputting the prediction result, and optimizing the battery management strategy based on the prediction result.
2. The method for predicting the service life of the battery pack according to claim 1, wherein the pre-constructing a service life prediction model of the battery pack comprises:
capturing specified sample data from a new energy remote monitoring big data platform;
performing data processing on the sample data to obtain processed target sample data;
and training a pre-deployed machine learning model through the target sample data to obtain a battery pack service life prediction model.
3. The method according to claim 2, wherein the performing data processing on the sample data to obtain processed target sample data includes:
data cleaning is carried out on the sample data to remove abnormal data and low-value data, and cleaned data are obtained;
performing feature extraction on the cleaned data to obtain feature data;
performing characteristic correlation analysis on the characteristic data to obtain an analysis result;
and performing feature filtering on the feature data according to the analysis result to obtain target sample data.
4. The method according to claim 2, wherein training a pre-deployed machine learning model with the target sample data to obtain a battery pack life prediction model comprises:
a machine learning model is deployed in advance by adopting a recurrent neural network; wherein the model unit of the machine learning model is a door control cycle unit;
dividing the target sample data into an initial training set, an initial verification set and an initial test set according to a preset division ratio;
respectively carrying out normalization processing on the initial training set, the initial verification set and the initial test set to obtain a training set, a verification set and a test set;
training the machine learning model according to a pre-constructed loss function and the training set to obtain a trained model;
performing model precision verification on the trained model through the test set and the verification set to obtain a verification result;
and when the verification result meets a preset precision threshold, determining the trained model as a battery pack service life prediction model.
5. The method for predicting the service life of a battery pack according to claim 1, wherein the step of calling the battery pack service life prediction model according to the battery pack service life prediction instruction and the service life prediction characteristic value to predict the service life of the battery pack to obtain a prediction result comprises the following steps:
normalizing the service life prediction characteristic value according to the service life prediction instruction of the battery pack to obtain a normalized characteristic value;
calling the battery pack service life prediction model according to the normalized characteristic value to predict the service life of the battery pack to obtain a preliminary prediction result;
and performing inverse normalization processing on the preliminary prediction result to obtain a final prediction result.
6. A battery pack life prediction apparatus, comprising:
the model building unit is used for building a battery pack service life prediction model in advance;
the device comprises an acquisition unit, a service life prediction unit and a service life prediction characteristic value, wherein the acquisition unit is used for acquiring a service life prediction instruction and a service life prediction characteristic value of a battery pack input by a user;
the prediction unit is used for calling the battery pack service life prediction model to predict the service life of the battery pack according to the battery pack service life prediction instruction and the service life prediction characteristic value to obtain a prediction result;
and the output unit is used for outputting the prediction result and optimizing the battery management strategy based on the prediction result.
7. The battery pack life prediction apparatus according to claim 6, wherein the model construction unit includes:
the capturing subunit is used for capturing specified sample data from the new energy remote monitoring big data platform;
the processing subunit is used for carrying out data processing on the sample data to obtain processed target sample data;
and the training subunit is used for training a pre-deployed machine learning model through the target sample data to obtain a battery pack service life prediction model.
8. The battery pack life prediction device of claim 7, wherein the processing subunit comprises:
the cleaning module is used for cleaning the data of the sample data to remove abnormal data and low-value data and obtain cleaned data;
the extraction module is used for extracting the characteristics of the cleaned data to obtain characteristic data;
the analysis module is used for carrying out characteristic correlation analysis on the characteristic data to obtain an analysis result;
and the filtering module is used for carrying out feature filtering on the feature data according to the analysis result to obtain target sample data.
9. An electronic device, comprising a memory for storing a computer program and a processor that executes the computer program to cause the electronic device to perform the method of battery pack life prediction of any of claims 1-5.
10. A readable storage medium having stored thereon computer program instructions which, when read and executed by a processor, perform the method of battery pack life prediction according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211274814.5A CN115508732A (en) | 2022-10-18 | 2022-10-18 | Battery pack service life prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211274814.5A CN115508732A (en) | 2022-10-18 | 2022-10-18 | Battery pack service life prediction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115508732A true CN115508732A (en) | 2022-12-23 |
Family
ID=84510232
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211274814.5A Pending CN115508732A (en) | 2022-10-18 | 2022-10-18 | Battery pack service life prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115508732A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116011993A (en) * | 2023-01-10 | 2023-04-25 | 九源云(广州)智能科技有限公司 | Storage battery health management system based on CPS architecture |
-
2022
- 2022-10-18 CN CN202211274814.5A patent/CN115508732A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116011993A (en) * | 2023-01-10 | 2023-04-25 | 九源云(广州)智能科技有限公司 | Storage battery health management system based on CPS architecture |
| CN116011993B (en) * | 2023-01-10 | 2024-01-30 | 九源云(广州)智能科技有限公司 | Storage battery health management system based on CPS architecture |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112782591B (en) | Lithium battery SOH long-term prediction method based on multi-battery data fusion | |
| CN113158947B (en) | Power battery health scoring method, system and storage medium | |
| CN106556802A (en) | A kind of accumulator battery exception cell recognition methodss and device | |
| CN105373894A (en) | Inspection data-based power marketing service diagnosis model establishing method and system | |
| CN110852476A (en) | Passenger flow prediction method and device, computer equipment and storage medium | |
| CN115343621A (en) | Power battery health state prediction method and device based on data driving | |
| CN115508711B (en) | Electric vehicle power lithium battery residual life prediction method based on deep network model | |
| CN112651785A (en) | Real-time monitoring method and system for transaction amount | |
| CN115508732A (en) | Battery pack service life prediction method and device | |
| CN114202243A (en) | Engineering project management risk early warning method and system based on random forest | |
| CN111626514A (en) | Electric vehicle charging load prediction method and device | |
| CN116703464A (en) | Electric automobile charging demand modeling method and device, electronic equipment and storage medium | |
| CN117471346A (en) | Method and system for determining remaining life and health status of retired battery module | |
| CN114924203A (en) | Battery SOH prediction analysis method and electric automobile | |
| CN115700717A (en) | Power distribution analysis method based on electric automobile power consumption demand | |
| CN118376936A (en) | Intelligent diagnosis method and system for lithium battery state | |
| CN117330987B (en) | Method, system, medium and apparatus for time-based battery state of health assessment | |
| CN118050653A (en) | Method and device for predicting the battery of a diagnostic device using a multi-element converter model | |
| CN112287979A (en) | Mutual information-based energy storage battery state judgment method | |
| CN116794544A (en) | Method for detecting life of power battery, computer-readable storage medium, and vehicle | |
| CN115865716B (en) | Network state analysis method, system and computer readable medium | |
| CN117048524A (en) | Method and device for detecting vehicle faults, vehicle and storage medium | |
| CN113538735B (en) | Vehicle use identification method and device | |
| CN115828161A (en) | Automobile fault type prediction method and device based on recurrent neural network | |
| CN114545276A (en) | Power battery service life prediction method based on capacity test and Internet of vehicles big data |
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 |