CN117686935B - Battery RUL prediction method based on voltage probability density - Google Patents
Battery RUL prediction method based on voltage probability density Download PDFInfo
- Publication number
- CN117686935B CN117686935B CN202410123299.3A CN202410123299A CN117686935B CN 117686935 B CN117686935 B CN 117686935B CN 202410123299 A CN202410123299 A CN 202410123299A CN 117686935 B CN117686935 B CN 117686935B
- Authority
- CN
- China
- Prior art keywords
- voltage
- probability density
- battery
- capacity
- characteristic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 59
- 230000036541 health Effects 0.000 claims abstract description 24
- 230000008569 process Effects 0.000 claims abstract description 16
- 238000000354 decomposition reaction Methods 0.000 claims description 19
- 238000012360 testing method Methods 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 11
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 8
- 238000012545 processing Methods 0.000 abstract description 6
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 16
- 229910052744 lithium Inorganic materials 0.000 description 16
- 230000015556 catabolic process Effects 0.000 description 11
- 238000006731 degradation reaction Methods 0.000 description 11
- 230000000875 corresponding effect Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 4
- 238000005259 measurement Methods 0.000 description 3
- 238000000691 measurement method Methods 0.000 description 3
- 241000714933 Chryseobacterium nakagawai Species 0.000 description 2
- 230000032683 aging Effects 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003487 electrochemical reaction Methods 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
Landscapes
- Secondary Cells (AREA)
Abstract
The invention provides a battery RUL prediction method based on voltage probability density, which comprises the following steps: taking the positive and negative 0.05V range of the voltage corresponding to the peak value of the voltage probability density curves a as a characteristic voltage interval, calculating the voltage probability density distribution in the characteristic voltage interval under different cycle periods, and drawing a voltage probability density curve b; extracting peaks of a voltage probability density curve b under different cycle periods, drawing a voltage probability density peak curve, and taking the voltage probability density peak curve as a health feature HF; taking health characteristic HF as characteristic data sequence of batteryAnd using characteristic data sequencesAnd constructing a ridge regression capacity prediction model by using the battery capacity data to obtain a capacity prediction value C. According to the scheme, the voltage probability density peak curve of the battery is drawn to serve as the health characteristic HF, so that the prediction process only needs to collect data in a characteristic voltage interval, the data calculation amount is greatly reduced, and the processing efficiency is improved.
Description
Technical Field
The invention relates to the technical field of battery health management, in particular to a battery RUL prediction method based on voltage probability density.
Background
Lithium batteries are widely used in various fields such as electronic products, new energy automobiles, energy storage systems and the like because of their advantages of high energy density, long cycle life, low self-discharge rate and the like. However, with the increase of the charge-discharge cycle period, the battery can have performance degradation phenomena such as available capacity reduction, and if the battery is not replaced in time, huge safety risks can be brought to the equipment, and the stability and reliability of the system are seriously affected. The RUL of a lithium battery refers to the number of charge-discharge cycles that the battery capacity decays to a failure threshold in the current state, and is an important indicator for characterizing the aging and health condition of the battery. The RUL of the lithium battery can be accurately predicted to obtain key information of the battery state in advance, and the aged or failed battery can be timely replaced, so that the RUL is important for improving the stability and reliability of a power supply system.
Because the direct measurement of the capacity of the lithium battery needs to go deep into the battery, and damage to the internal structure of the battery and damage the tightness and integrity of the battery, in practical application, the direct measurement method is not commonly used, but an indirect measurement method is adopted, and the indirect measurement method is used for deducing the internal state and performance of the battery by monitoring the performance parameters of the lithium battery, such as voltage, current, internal resistance, temperature and other data as health characteristics and capacity degradation trend to perform modeling and analysis, so as to obtain the RUL of the battery. The RUL prediction method based on data driving can directly analyze and mine factors influencing the degradation trend of the battery performance from monitoring measurement indexes of the battery, does not need to relate to complex and professional electrochemical reaction problems, has remarkable universality, and still has the following defects: (1) Finding a health feature highly correlated with the battery capacity degradation trend is not easy, which directly affects whether the actual change in battery capacity can be accurately described; (2) The current signal decomposition method often has a modal aliasing phenomenon, which makes it difficult to accurately capture the characteristic information of each component, resulting in lower accuracy of the prediction result.
Therefore, in the field of lithium battery RUL prediction, the existing method needs to be improved and optimized in the aspects of health feature data collection, signal decomposition technology, modeling and prediction algorithm and the like, so that a more accurate lithium battery RUL prediction result can be realized, and powerful support is provided for battery state estimation and fault early warning technology.
Disclosure of Invention
Based on the problems in the prior art, the invention aims to provide a battery RUL prediction method based on voltage probability density so as to improve and optimize the existing prediction method.
The invention provides a battery RUL prediction method based on voltage probability density, which comprises the following steps:
s1: collecting multiple groups of charging voltage data of the battery in a plurality of charge-discharge cycle tests, and drawing multiple voltage probability density curves a;
S2: taking the positive and negative 0.05V range of the voltage corresponding to the peak value of all the voltage probability density curves a as a characteristic voltage interval, calculating the voltage probability density distribution in the characteristic voltage interval under different cycle periods, and drawing a voltage probability density curve b;
S3: extracting peaks of a voltage probability density curve b under different cycle periods, drawing a voltage probability density peak curve, and taking the voltage probability density peak curve as a health feature HF;
S4: taking health characteristic HF as characteristic data sequence of battery And utilize the characteristic data sequence/>And battery capacity data to construct a ridge regression capacity prediction model;
S5: sequence of characteristic data Inputting a ridge regression capacity prediction model to obtain a capacity prediction value C;
S6: judging whether the capacity predicted value C is smaller than the failure threshold value T or not, if the capacity predicted value C is smaller than the failure threshold value T, stopping cycle prediction, and outputting the current cycle number as RUL of the battery of the model to be detected; otherwise, go to step S5.
According to an embodiment of the present invention, step S4 further comprises the step of sequencing the feature dataThe method comprises the steps of dividing a training set and a testing set, constructing a ridge regression capacity prediction model by using the training set and battery capacity data, and then verifying the accuracy and reliability of the ridge regression capacity prediction model by using the testing set.
According to an embodiment of the present invention, the characteristic data is sequenced in step S5The input ridge regression capacity prediction model specifically comprises:
s51: multi-scale decomposition of training set into high frequency components using EEMD algorithm And low frequency component/>；
S52: will high frequency componentAnd low frequency component/>As the input of the multi-scale prediction model, predicting through the GRU model and the MLR model respectively;
s53: overlapping and reconstructing the prediction result to obtain a new characteristic data sequence ；
S54: sequence of new characteristic dataAnd inputting the trained ridge regression capacity prediction model to perform capacity prediction.
In step S51, the training set is multi-scale decomposed into high frequency components using the EEMD algorithm according to an embodiment of the present inventionAnd low frequency component/>The method specifically comprises the following steps:
sequence of orientation feature data Adding white noise signal/>Acquisition of New sequence/>：
，
Wherein i is the number of times white noise is added;
Will be EMD decomposition is carried out to obtain the form of each IMF component sum and the residual component/>, after the decomposition：
，
In the method, in the process of the invention,The jth IMF component obtained by decomposing after adding white noise for the ith time, wherein the value range of j is 1-n, n represents that n IMFs can be obtained by decomposing togetherThe remainder of each decomposition, i.e., the original signal minus the sum of each set of IMFs;
repeating the above two steps for M times, adding the IMF components obtained each time, and then averaging to obtain a final result:
，
In the method, in the process of the invention, The mean value of the j-th IMF component obtained after EEMD decomposition is obtained for the characteristic data sequence.
According to an embodiment of the present invention, the drawing of the voltage probability density curve a in step S1 includes: and drawing a plurality of charging voltage-time curves by using the collected plurality of groups of battery charging voltage data, counting the occurrence times of each voltage point in the charging voltage-time curves by using a point counting method, and dividing the occurrence times of each voltage value by the total data quantity to obtain corresponding probability distribution.
The invention has the following beneficial effects:
According to the battery RUL prediction method based on the voltage probability density, provided by the invention, the positive and negative 0.05V range of the voltages corresponding to the peak values of the voltage probability density curves a is used as the characteristic voltage interval, the voltage probability density distribution in the characteristic voltage interval under different cycle periods is calculated, the voltage probability density curve b is drawn, and the voltage probability density peak value curve b is used as the health characteristic HF, so that the data in the characteristic voltage interval is only required to be collected in the prediction process, and compared with the existing prediction method, the data calculation amount is greatly reduced, and the processing efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments or the technical solutions in the prior art, the following description will briefly introduce the drawings that are needed in the embodiments or the description of the prior art, it is obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a block diagram of steps of a battery RUL prediction method based on voltage probability density;
FIG. 2 is a block flow diagram of a method for predicting battery RUL based on voltage probability density according to an embodiment of the present invention;
FIG. 3 is a graph of charge voltage versus time for a first cycle of a lithium battery in an embodiment of the invention;
FIG. 4 is a graph of voltage probability density for a first cycle period of a lithium battery in an embodiment of the invention;
FIG. 5 is a graph of voltage probability density for different cycle periods of a lithium battery in an embodiment of the invention;
FIG. 6 is a graph of the peak voltage probability density of a lithium battery in an embodiment of the invention;
fig. 7 is a graph showing degradation of Cell No. 1 battery capacity in the embodiment of the present invention;
FIG. 8 is a network architecture diagram of a GRU model in an embodiment of the invention;
FIG. 9 is a new health profile of a lithium battery in an embodiment of the invention A graph;
fig. 10 is a result of predicting RUL of a lithium battery according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments refers to the accompanying drawings, which illustrate specific embodiments in which the invention may be practiced. The directional terms mentioned in the present invention, such as [ upper ], [ lower ], [ front ], [ rear ], [ left ], [ right ], [ inner ], [ outer ], [ side ], etc., are only referring to the directions of the attached drawings. Accordingly, directional terminology is used to describe and understand the invention and is not limiting of the invention. In the drawings, like elements are designated by like reference numerals.
The invention provides a battery RUL prediction method based on voltage probability density, which comprises the following steps:
s1: collecting multiple groups of charging voltage data of the battery in a plurality of charge-discharge cycle tests, and drawing multiple voltage probability density curves a;
S2: taking the positive and negative 0.05V range of the voltage corresponding to the peak value of all the voltage probability density curves a as a characteristic voltage interval, calculating the voltage probability density distribution in the characteristic voltage interval under different cycle periods, and drawing a voltage probability density curve b;
S3: extracting peaks of a voltage probability density curve b under different cycle periods, drawing a voltage probability density peak curve, and taking the voltage probability density peak curve as a health feature HF;
S4: taking health characteristic HF as characteristic data sequence of battery And utilize the characteristic data sequence/>And battery capacity data to construct a ridge regression capacity prediction model;
S5: sequence of characteristic data Inputting a ridge regression capacity prediction model to obtain a capacity prediction value C;
S6: judging whether the capacity predicted value C is smaller than the failure threshold value T or not, if the capacity predicted value C is smaller than the failure threshold value T, stopping cycle prediction, and outputting the current cycle number as RUL of the battery of the model to be detected; otherwise, go to step S5, i.e. feature sequence The next feature value of (3) is input into a ridge regression capacity prediction model to obtain a new capacity prediction value C.
Compared with the prior art, in the battery RUL prediction method based on the voltage probability density, the positive and negative 0.05V range of the voltage corresponding to the peak value of the voltage probability density curves a is used as the characteristic voltage interval, the voltage probability density distribution in the characteristic voltage interval under different cycle periods is calculated, the voltage probability density curve b is drawn, the voltage probability density peak value curve b is used as the health characteristic HF, the fact that the prediction process only needs to collect data in the characteristic voltage interval is achieved, and compared with the existing prediction method which utilizes complete charging voltage data prediction, the data calculation amount is greatly reduced, and therefore the processing efficiency is improved.
Specifically, in the experiment, a Cell No. 1 battery in a battery pack is selected as a test object, the rated capacity of the battery pack is 740mAh, a charge-discharge aging experiment of the battery is performed at a constant environmental temperature of 40 ℃, a constant-current constant-voltage mode is adopted to charge the battery, and the battery data is measured after 100 times of cycles by simulating the discharge process of the running working condition of an Artemis urban area, so that the voltage data is measured 78 groups. Fig. 2 is a flow chart of predicting the RUL of the lithium battery by adopting the method for predicting the RUL of the battery in the embodiment, which specifically includes the following steps:
(1) Collecting multiple groups of charging voltage data of the lithium battery in a plurality of charge-discharge cycle tests, carrying out interpolation processing on abnormal voltage data during the collection, and correspondingly drawing multiple charging voltage-time curves, as shown in fig. 3; specifically, in order to reduce the number of data measurement, when data is acquired, the data of the battery is measured after every 100 cycles, and 78 groups of voltage data are measured in total; counting the occurrence times of each voltage point in each charging voltage-time curve by using a point method with 0.01V as a unit, dividing the occurrence times of each voltage value by the total data quantity to obtain corresponding probability density, and drawing a voltage probability density curve a as shown in fig. 4;
(2) According to the range of the peak value of the coverage voltage probability density curve a, taking the voltage (i.e., [3.74V,3.84V ]) within the range of 3.79 V+/-0.05V of the abscissa voltage value corresponding to the peak point as a characteristic voltage interval, and collecting only the voltage data of the interval in the subsequent charging cycle period, so that the data volume is greatly reduced;
(3) Calculating voltage probability density distribution in a characteristic voltage interval under different cycle periods, and drawing a voltage probability density curve b, as shown in fig. 5;
(4) Extracting peaks in the voltage probability density curve b under different cycle periods, and drawing a voltage probability density peak curve graph, as shown in fig. 6;
As shown in fig. 7, the Cell No. 1 battery capacity degradation curve is obtained by analyzing the correlation between the voltage probability density peak curve and the battery capacity degradation curve using pearson correlation coefficient (Pearson Correlation Coefficient), and the calculation formula is as follows:
，
In the method, in the process of the invention, 、/>Corresponding voltage probability density peak and average value thereof,/>、/>For battery capacity and its average, n is the data length. The voltage probability density peak curve and the battery capacity degradation are obviously positively correlated by the calculation of the formula, and have extremely strong correlation, so that the voltage probability density peak curve can be used as Health Feature (HF) to indirectly predict the capacity of the battery.
(5) The health characteristic HF is taken as characteristic data sequence of the battery and is recorded asThe first 50 data are divided into training sets and the last 28 data are divided into test sets. Constructing a ridge regression capacity prediction model by using the training set and the battery capacity data, and then verifying the accuracy and reliability of the model on capacity prediction by using a test set;
(6) The training set is subjected to multi-scale decomposition by using EEMD (ensemble empirical mode decomposition) algorithm, and then a plurality of components obtained after decomposition are subjected to classification processing, and are divided into high-frequency components according to the size of frequency And low frequency component/>Wherein the high frequency component/>Reflecting the local degradation trend of the battery, while the low frequency component/>The main trend of battery degradation can be reflected, and the specific process is as follows:
(6.1) sequence of orientation feature data Adding white noise signal/>Acquisition of New sequence/>：
，
Wherein i is the number of times white noise is added;
(6.2) will EMD (empirical mode decomposition) decomposition is performed to obtain the form of each IMF component sum and the decomposed residual component/>：
，
In the method, in the process of the invention,The jth IMF (INTRINSIC MODE FUNCTIONS, natural mode component) component obtained by decomposing after adding white noise for the ith time, wherein the value range of j is 1-n, n represents that n IMFs can be obtained by decomposing togetherThe remaining part obtained by each decomposition, namely the sum of the original signal minus each group of IMFs, represents the part of the original signal which cannot be decomposed into IMFs, including noise, aperiodic components and other irregular signal components;
(6.3) repeating the above two steps M times, and adding the IMF components obtained each time and averaging again as a final result:
，
In the method, in the process of the invention, The mean value of the j-th IMF component obtained after EEMD decomposition is obtained for the characteristic data sequence.
(7) For each component characteristic, high frequency components are addedAnd low frequency component/>As the input of the multi-scale prediction model, inputting GRU (Gate Recurrent Unit) model and MLR (mixed logistic regression) model respectively to predict, and then overlapping and reconstructing the prediction result to obtain a new characteristic data sequence/>. As shown in fig. 8, which is a network structure diagram of the GRU model, the calculation formula of the state and output of the GRU model is:
，
，
，
，
In the method, in the process of the invention, To reset the gate,/>Is the hidden state of the previous moment,/>For input at the present moment,/>To update the door,/>Hidden state calculated for reset gate,/>To update the hidden state of the update gate,/>、/>And/>Is a weight matrix,/>And/>To activate the function.
The matrix expression and expansion of the MLR model are:
，
，
In the method, in the process of the invention, Representing data to be predicted,/>Representing historical data,/>Is a regression coefficient,/>Is a random disturbance. And carrying out parameter solving by using a least square method to obtain a regression function, wherein the regression function is specifically shown as follows:
。
(8) After the multi-scale prediction model, new health characteristics shown in figure 9 are obtained It can be observed that the predictive effect is very accurate and very close to the actual trend of the fluctuation of the health features.
(9) Will be new health featuresAs a characteristic data sequence of the battery and noted/>And inputting the data into a trained ridge regression capacity prediction model to perform capacity prediction. After the prediction is finished, if the cycle number is N (N is a positive integer), judging whether the capacity predicted value C at the moment is smaller than a failure threshold value T or not; if the predicted capacity C is smaller than the failure threshold T, the cycle prediction is stopped, and the number of cycle periods that have been currently experienced is output as the RUL of the battery of the model to be measured, and the prediction result is shown in fig. 10, in this embodiment, 80% of the rated capacity (i.e., 590 mAh) is used as the failure threshold of the lithium battery.
In order to evaluate the performance of the model more comprehensively, the following 3 indexes are adopted for evaluation, namely absolute error, average absolute error and root mean square error, and the calculation formula is as follows:
，
，
，
In the method, in the process of the invention, Is the number of cycles at the end of battery life in a real situation,/>For the RUL predictor, n is the number of predicted cycle cycles,/>As the actual value of the capacity,/>Is a capacity predictor. According to the result of the RUL prediction in fig. 10, the prediction curve is observed to be closer to the true value, and the actual capacity degradation trend can be reflected, which indicates that the scheme provided by the invention can give a more accurate RUL result, wherein the error evaluation index AE value is 0, and the mae and RMSE are respectively 0.0042 and 0.0052, which shows higher prediction accuracy. Finally, the predicted cycle number 63 when the failure threshold is reached is outputted as the RUL value of the battery of the model to be tested.
In summary, in the battery RUL prediction method based on the voltage probability density provided by the embodiment of the invention, by drawing the voltage probability density peak curve of the battery as the health feature HF, the prediction process is realized by only collecting the data in the feature voltage interval, and compared with the existing prediction method by using complete charge voltage data prediction, the data calculation amount is greatly reduced, so that the processing efficiency is improved; meanwhile, by designing a multi-scale prediction model, the health feature sequence is decomposed into a high-frequency component and a low-frequency component, noise interference in data can be reduced, a simpler fluctuation rule and obvious frequency features are provided, key factors causing capacity attenuation can be comprehensively captured, capacity prediction is performed through a ridge regression model, and therefore the accuracy of RUL prediction results is improved.
It should be noted that, although the present invention has been described in terms of the above embodiments, the above embodiments are not intended to limit the invention, and those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention, so that the scope of the invention is defined by the appended claims.
Claims (5)
1. The battery RUL prediction method based on the voltage probability density is characterized by comprising the following steps of:
s1: collecting multiple groups of charging voltage data of the battery in a plurality of charge-discharge cycle tests, and drawing multiple voltage probability density curves a;
S2: taking the positive and negative 0.05V range of the voltage corresponding to the peak value of all the voltage probability density curves a as a characteristic voltage interval, calculating the voltage probability density distribution in the characteristic voltage interval under different cycle periods, and drawing a voltage probability density curve b;
S3: extracting peaks of a voltage probability density curve b under different cycle periods, drawing a voltage probability density peak curve, and taking the voltage probability density peak curve as a health feature HF;
S4: taking health characteristic HF as characteristic data sequence of battery And utilize the characteristic data sequence/>And battery capacity data to construct a ridge regression capacity prediction model;
S5: sequence of characteristic data Inputting a ridge regression capacity prediction model to obtain a capacity prediction value C;
S6: judging whether the capacity predicted value C is smaller than the failure threshold value T or not, if the capacity predicted value C is smaller than the failure threshold value T, stopping cycle prediction, and outputting the current cycle number as RUL of the battery of the model to be detected; otherwise, go to step S5.
2. The method of claim 1, wherein step S4 further comprises the step of sequencing the feature dataThe method comprises the steps of dividing a training set and a testing set, constructing a ridge regression capacity prediction model by using the training set and battery capacity data, and then verifying the accuracy and reliability of the ridge regression capacity prediction model by using the testing set.
3. The battery RUL prediction method based on voltage probability density according to claim 2, wherein the characteristic data sequence is set in step S5The input ridge regression capacity prediction model specifically comprises:
s51: multi-scale decomposition of training set into high frequency components using EEMD algorithm And low frequency component/>；
S52: will high frequency componentAnd low frequency component/>As input to the multiscale prediction model, the high frequency component/>Prediction by GRU model and low frequency component/>Predicting through an MLR model;
s53: overlapping and reconstructing the prediction result to obtain a new characteristic data sequence ；
S54: sequence of new characteristic dataAnd inputting the trained ridge regression capacity prediction model to perform capacity prediction.
4. The battery RUL prediction method based on voltage probability density according to claim 3, wherein the training set is multi-scale decomposed into high frequency components using EEMD algorithm in step S51And low frequency component/>The method specifically comprises the following steps:
s511: sequence of orientation feature data Adding white noise signal/>Acquisition of New sequence/>：
，
Wherein i is the number of times white noise is added;
S512: will be EMD decomposition is carried out to obtain the form of each IMF component sum and the residual component/>, after the decomposition：
，
In the method, in the process of the invention,The jth IMF component obtained by decomposing after adding white noise for the ith time, wherein the value range of j is 1-n, n represents that n IMFs can be obtained by decomposing togetherThe remainder of each decomposition, i.e., the original signal minus the sum of each set of IMFs;
repeating the steps S511 and S512 for M times, adding the IMF components obtained each time and averaging to obtain a final result:
，
In the method, in the process of the invention, The mean value of the j-th IMF component obtained after EEMD decomposition is obtained for the characteristic data sequence.
5. The method for predicting the RUL of the battery based on the voltage probability density according to claim 1, wherein the drawing of the voltage probability density curve a in the step S1 includes: and drawing a plurality of charging voltage-time curves by using the collected plurality of groups of battery charging voltage data, counting the occurrence times of each voltage point in the charging voltage-time curves by using a point counting method, and dividing the occurrence times of each voltage value by the total data quantity to obtain corresponding probability distribution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410123299.3A CN117686935B (en) | 2024-01-30 | 2024-01-30 | Battery RUL prediction method based on voltage probability density |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410123299.3A CN117686935B (en) | 2024-01-30 | 2024-01-30 | Battery RUL prediction method based on voltage probability density |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117686935A CN117686935A (en) | 2024-03-12 |
| CN117686935B true CN117686935B (en) | 2024-05-14 |
Family
ID=90137371
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410123299.3A Active CN117686935B (en) | 2024-01-30 | 2024-01-30 | Battery RUL prediction method based on voltage probability density |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117686935B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104090238A (en) * | 2014-06-19 | 2014-10-08 | 哈尔滨工业大学 | Lithium ion battery remaining service life prediction method based on AR ensemble learning model |
| CN110442941A (en) * | 2019-07-25 | 2019-11-12 | 桂林电子科技大学 | It is a kind of to be tracked and RUL prediction technique based on the battery status for improving particle filter and process noise features fusion algorithm |
| CN111458649A (en) * | 2020-04-23 | 2020-07-28 | 国网陕西省电力公司汉中供电公司 | Rapid detection method for health degree of battery module |
| CN112485693A (en) * | 2020-11-19 | 2021-03-12 | 上海电力大学 | Battery health state rapid evaluation method based on temperature probability density function |
| CN113239556A (en) * | 2021-05-21 | 2021-08-10 | 中国工程物理研究院总体工程研究所 | Fatigue damage rate multiple estimation method of random acceleration power spectral density |
| CN113884889A (en) * | 2021-10-29 | 2022-01-04 | 蜂巢能源(上海)有限公司 | Battery safety early warning method and device, storage medium and electronic equipment |
| CN115389946A (en) * | 2022-09-20 | 2022-11-25 | 金陵科技学院 | Lithium battery health state estimation method based on isobaric rise energy and improved GRU |
| CN115659612A (en) * | 2022-10-12 | 2023-01-31 | 上海电力大学 | Method for predicting residual life of offshore wind turbine generator with typhoon influence taken into consideration |
| CN115704867A (en) * | 2021-08-17 | 2023-02-17 | 上海理工大学 | Method for estimating battery capacity and predicting remaining service life of lithium ion battery |
| WO2024013281A1 (en) * | 2022-07-14 | 2024-01-18 | Hitachi Energy Ltd | Method for lifetime prediction and monitoring |
-
2024
- 2024-01-30 CN CN202410123299.3A patent/CN117686935B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104090238A (en) * | 2014-06-19 | 2014-10-08 | 哈尔滨工业大学 | Lithium ion battery remaining service life prediction method based on AR ensemble learning model |
| CN110442941A (en) * | 2019-07-25 | 2019-11-12 | 桂林电子科技大学 | It is a kind of to be tracked and RUL prediction technique based on the battery status for improving particle filter and process noise features fusion algorithm |
| CN111458649A (en) * | 2020-04-23 | 2020-07-28 | 国网陕西省电力公司汉中供电公司 | Rapid detection method for health degree of battery module |
| CN112485693A (en) * | 2020-11-19 | 2021-03-12 | 上海电力大学 | Battery health state rapid evaluation method based on temperature probability density function |
| CN113239556A (en) * | 2021-05-21 | 2021-08-10 | 中国工程物理研究院总体工程研究所 | Fatigue damage rate multiple estimation method of random acceleration power spectral density |
| CN115704867A (en) * | 2021-08-17 | 2023-02-17 | 上海理工大学 | Method for estimating battery capacity and predicting remaining service life of lithium ion battery |
| CN113884889A (en) * | 2021-10-29 | 2022-01-04 | 蜂巢能源(上海)有限公司 | Battery safety early warning method and device, storage medium and electronic equipment |
| WO2024013281A1 (en) * | 2022-07-14 | 2024-01-18 | Hitachi Energy Ltd | Method for lifetime prediction and monitoring |
| CN115389946A (en) * | 2022-09-20 | 2022-11-25 | 金陵科技学院 | Lithium battery health state estimation method based on isobaric rise energy and improved GRU |
| CN115659612A (en) * | 2022-10-12 | 2023-01-31 | 上海电力大学 | Method for predicting residual life of offshore wind turbine generator with typhoon influence taken into consideration |
Non-Patent Citations (4)
| Title |
|---|
| 基于CNN-BiLSTM的锂电池剩余使用寿命概率密度预测;刘泽 等;《电源技术》;20230131;第第47卷卷(第第1期期);57至61段 * |
| 多尺度分解下GRU-MLR组合的锂电池剩余使用寿命预测方法;武明虎 等;《储能科学与技术》;20230731;第第12卷卷(第第7期期);2220至2227页 * |
| 张金龙 ; 佟微 ; 孙叶宁 ; 李端凯 ; 漆汉宏 ; 魏艳君 ; .锂电池健康状态估算方法综述.电源学报.2017,(第02期),全文. * |
| 锂电池健康状态估算方法综述;张金龙;佟微;孙叶宁;李端凯;漆汉宏;魏艳君;;电源学报;20170315(第02期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117686935A (en) | 2024-03-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112731159B (en) | Method for pre-judging and positioning battery faults of battery compartment of energy storage power station | |
| CN111398833B (en) | Battery health state assessment method | |
| CN111146514B (en) | Lithium ion battery module operation safety evaluation prediction method and system and electronic equipment | |
| CN113805064B (en) | Lithium ion battery pack health state prediction method based on deep learning | |
| CN112092675A (en) | Battery thermal runaway early warning method, system and server | |
| CN112327167B (en) | Battery capacity diving risk assessment method and system | |
| CN113447828B (en) | Lithium battery temperature estimation method and system based on Bayesian neural network | |
| CN115366683A (en) | Fault diagnosis strategy for new energy automobile power battery multi-dimensional model fusion | |
| CN114818831B (en) | Bidirectional lithium ion battery fault detection method and system based on multi-source perception | |
| CN116401585B (en) | Energy storage battery failure risk assessment method based on big data | |
| CN111983474A (en) | Lithium ion battery life prediction method and system based on capacity decline model | |
| CN110568360A (en) | lithium battery aging diagnosis method based on fuzzy logic algorithm | |
| CN115480180A (en) | New energy battery health diagnosis and analysis method | |
| Kamali et al. | Novel SOH estimation of lithium-ion batteries for real-time embedded applications | |
| CN114646888A (en) | Assessment method and system for capacity attenuation of power battery | |
| CN114280490A (en) | Lithium ion battery state of charge estimation method and system | |
| Yu et al. | SOH estimation method for lithium-ion battery based on discharge characteristics | |
| CN117686935B (en) | Battery RUL prediction method based on voltage probability density | |
| CN114355218A (en) | Lithium ion battery charge state prediction method based on multi-feature quantity screening | |
| CN113255215B (en) | Lithium battery health state estimation method based on voltage segments | |
| Liu et al. | Prediction method for battery self-discharge voltage drop based on pre-classifier | |
| Xu et al. | Prediction of Lithium-ion Battery Remaining Useful Life Based on Long Short Term Memory | |
| CN115825755B (en) | Method for evaluating consistency of voltages of battery cells of energy storage battery | |
| Zhang et al. | State of health estimation for lithium batteries using improved weighted grey relational method | |
| Li et al. | Estimation of State-of-Health for Lithium-Ion Battery Based on Increment Capacity Analysis Method and Long Short-Term Memory Neural Network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |