CN112731159B - Method for pre-judging and positioning battery faults of battery compartment of energy storage power station - Google Patents
Method for pre-judging and positioning battery faults of battery compartment of energy storage power station Download PDFInfo
- Publication number
- CN112731159B CN112731159B CN202011543908.9A CN202011543908A CN112731159B CN 112731159 B CN112731159 B CN 112731159B CN 202011543908 A CN202011543908 A CN 202011543908A CN 112731159 B CN112731159 B CN 112731159B
- Authority
- CN
- China
- Prior art keywords
- battery
- fault
- data
- model
- energy storage
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/3644—Constructional arrangements
- G01R31/3648—Constructional arrangements comprising digital calculation means, e.g. for performing an algorithm
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/378—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC] specially adapted for the type of battery or accumulator
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/382—Arrangements for monitoring battery or accumulator variables, e.g. SoC
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/385—Arrangements for measuring battery or accumulator variables
- G01R31/387—Determining ampere-hour charge capacity or SoC
- G01R31/388—Determining ampere-hour charge capacity or SoC involving voltage measurements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/392—Determining battery ageing or deterioration, e.g. state of health
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The application discloses a method for pre-judging and positioning faults of a battery compartment of an energy storage power station, which comprises the steps of analyzing the health degree of a single battery, predicting the variation trend of the performance of the battery, accurately pre-judging and positioning the faults of the battery, and acquiring the upper and lower limit thresholds of the running parameters of the battery and the health state of the battery in the current state through analyzing the health degree of the single battery; predicting the battery performance change trend in a future period based on the historical data of the past period, and performing pre-judgment and positioning on possible faults in the future by combining an upper threshold and a lower threshold; meanwhile, if the sudden change of a certain performance or certain performances of the single battery, the battery pack or the battery cluster possibly occurs in a very short time is detected, the fault possibly occurring in a short time can be predicted through the accurate battery fault prediction and positioning function, and the corresponding fault is positioned and identified through the fault expert database. The comprehensive implementation of the application can obviously improve the operation and maintenance technical support level and the system safety and stability level of the energy storage power station.
Description
Technical Field
The application relates to an electrochemical energy storage fault processing method, in particular to a method for pre-judging and positioning faults of a battery compartment of an energy storage power station.
Background
Along with the continuous expansion of the scale and application range of the energy storage power station, power station accidents of framework and operation by adopting lithium ion batteries occur. The lithium ion energy storage battery has high energy density and high thermal runaway risk, and has extremely high fire extinguishing difficulty after fire disaster, and serious direct loss and indirect loss can be caused. Therefore, the development of the fire early warning function of the energy storage power station is not delayed, the fire early warning function is discovered in advance, and the occurrence possibility and the hazard of the fire can be greatly reduced by timely treatment. Meanwhile, as the cycle times of the battery are increased, the consistency of the battery is reduced, and the electric performance, the thermal performance and the mechanical performance of the battery are all attenuated. Therefore, it is critical to predict the battery fault of the energy storage power station and locate and trace the source in time effectively, but the energy storage power station lacks an effective active battery fault pre-judging system at present.
Disclosure of Invention
The application aims to: aiming at the problems, the application provides a method for pre-judging and positioning the battery faults of the battery compartment of the energy storage power station, which is used for pre-alarming and diagnosing the battery faults of the energy storage power station, combines accurate pre-judging and fuzzy trend analysis of the faults, realizes pre-judging and tracing positioning of multi-level faults of battery monomers, battery modules, battery clusters and the like, and improves the safe operation level of the energy storage power station.
The technical scheme is as follows: the technical scheme adopted by the application is a method for predicting and positioning battery faults of a battery compartment of an energy storage power station, which comprises the following steps:
(1) Obtaining upper and lower limit thresholds of battery operation parameters and the battery health state in the current state through analyzing the health degree of the single battery; the single battery health degree analysis comprises the following steps:
(11) Selecting statistics by principal component analysis;
(12) The selected historical data is subjected to kernel density estimation, a probability density function is obtained by combining a statistical method, and a corresponding threshold value is calculated according to the probability density function; wherein the selected historical data includes voltage, temperature, and SOC data for the battery within two weeks.
(13) The overrun frequency of each statistic is calculated as a health degree based on the selected historical data to evaluate the health status of each individual.
(2) Predicting the battery performance change trend in a future period based on the historical data of the past period, and performing pre-judgment and positioning on possible faults in the future by combining an upper threshold and a lower threshold; meanwhile, the performance of the battery monomer, the battery pack or the battery cluster is subjected to abrupt change prediction in a short time, the possible faults of the performance of the battery monomer, the battery pack or the battery cluster in a short time are predicted through accurate prediction and positioning of the faults, and the fault positioning is performed.
Wherein, the battery performance change trend comprises the following steps:
(21) Analyzing the similarity degree of the development trend of each factor of the feature quantity to be measured and the capacity two sequences by a gray level correlation analysis method, and measuring the correlation between the two sequences; the degree of linear correlation of two variables is measured through the Pearson correlation coefficient; finally, whether the two variables are strictly monotonous or not is measured through the spearman grade correlation coefficient, and then the health factor is determined; and (3) the health factor determined in the step (21) is an equal-time-interval voltage difference sequence, and the voltage difference is obtained according to the monitoring time, the monitoring voltage and the capacity value in the constant-current charging mode.
(22) Establishing a health state estimation model by adopting an echo state network algorithm, and training the health state estimation model by taking the health factors determined in the step (21) as an input data set and the battery capacity as an output data set to obtain a battery capacity lithium ion battery health state estimation curve; the step (22) includes the following: dividing the obtained time interval to-be-measured parameter sequence and the corresponding capacity sequence into a training set and a testing set, wherein the model parameters of the health state estimation model comprise: the number N of the reserve pool processing units, the spectrum radius SR, the reserve pool input telescopic scale IS and the input unit displacement IF; the training process is to train the output weight of the health state estimation model through logistic regression, so that the root mean square error of the output result and the actual result is minimum, and optimize each parameter value through a cross verification method, so that the output result reaches the optimal state; and finally, inputting the parameters to be tested at equal time intervals of the test data into a model, estimating the capacity value of the battery, comparing the capacity estimated value of the battery with a true value, and measuring the accuracy of the model.
(23) And predicting the state of health of the lithium ion battery by adopting a degradation model algorithm, predicting the future state of health factors of the battery by using a long-short-term memory neural network algorithm, inputting the predicted result into the degradation model algorithm to obtain a capacity estimated value, and predicting the residual life of the lithium ion battery.
The long-term and short-term memory neural network algorithm comprises the following contents: sequentially recursively inputting m data by selecting the isochronous voltage difference data from the 1 st to (n-m) th cycle periods, sequentially recursively inputting the m data by selecting the data from the m to the n th cycle periods, and sequentially recursively outputting the m data; dividing all data into a training set and a test verification set, training a long-short-term memory neural network model through the training set, firstly normalizing the data during training, and then carrying out inverse normalization after prediction is finished to obtain a true value of a predicted result.
The accurate pre-judging and positioning of the battery fault in the step (2) comprises the following steps:
(31) Analyzing historical data within a period of time; the historical data mainly comprises a stack number, a cluster number, a stack voltage, a stack SOC, a stack temperature difference, a cluster voltage, a cluster SOC, a cluster pressure difference, a cluster temperature difference, an insulation resistance-, an insulation resistance+, a single cell number, a single cell voltage, a single cell temperature and a single cell SOC;
(32) Predicting a data value of the battery in a future period of time through an elastic network model; the step (32) includes the steps of:
(321) Initializing a model;
(322) Model training: training a model by using three-dimensional input data at the time of 0-t-1 in the past and SOC output data at the time of 0-t-1, wherein the three-dimensional input data comprises single battery voltage, SOC and single battery temperature;
(323) SOC prediction: estimating the SOC at the t moment by using the three-dimensional input data at the t moment by using the elastic network model obtained through training;
(324) Returning to execute model initialization, model training and SOC prediction, wherein the model training is to use three-dimensional input at the time of 0-t-2 and an SOC training model at the time of 1-t-1, and the SOC prediction is to use three-dimensional input data at the time of t to predict the SOC at the time of t+1;
(325) And (324) repeatedly executing each time the model training input is unchanged, outputting and advancing one time, and predicting one time by the SOC prediction, until the SOC prediction is gradually advanced to predict the SOC at the t+delta t moment by using the input at the t moment, and obtaining the predicted value of the SOC of the single battery in delta t after the current moment.
(33) Judging whether a fault occurs in a short time in the future according to a threshold value;
(34) And performing fault positioning and fault tracing of the faults according to the fault tree and the incidence matrix of the faults and the monitoring signals.
Firstly, combing faults possibly occurring in hardware and software aspects of an energy storage power station, combing faults possibly occurring in functional aspects of the energy storage power station, and constructing a typical fault tree of the energy storage power station; secondly, analyzing characterization parameters of main fault types of the energy storage power station through construction of a fault tree, and further establishing influence relations of the main fault types and relevant characterization parameters of the main fault types; thirdly, establishing an influence weight analysis of a main fault type and related characterization parameters thereof based on a Delphi method, and constructing a fault expert database of a battery compartment of the energy storage power station; finally, the type and weight relation of the characterization parameters are pre-judged through a battery fault accurate pre-judging procedure, and accurate pre-judging and positioning are carried out on the battery fault.
The method for pre-judging and positioning the battery fault of the battery compartment of the energy storage power station is stored in a computer readable storage medium in the form of a computer program, and the computer program enables a computer to execute the steps in the method for pre-judging and positioning the battery fault.
The beneficial effects are that: compared with the prior art, the application has the following advantages: firstly, an empirical fault expert database is constructed based on a fault tree and a Delphi method, so that the corresponding relation between the hardware fault of the battery compartment of the energy storage power station and the characterization parameters is defined, and the judgment and the positioning of the fault are realized. And secondly, a big data analysis method based on data driving is adopted to realize the health state assessment of the key equipment of the energy storage power station, so that the risk early warning, fault tracing and intelligent decision making capability are improved. The comprehensive implementation of the technology can obviously improve the operation and maintenance technical support level and the system safety and stability level of the energy storage power station.
Drawings
FIG. 1 is a flow chart of a cell health analysis according to the present application;
FIG. 2 is a battery cell health score criteria according to the present application;
FIG. 3 is a cell voltage health threshold line according to the present application;
FIG. 4 is a monomer temperature health threshold line according to the present application;
FIG. 5 is a monomer SOC health threshold line according to the present application;
FIG. 6 is a flow chart of the performance trend prediction according to the present application;
FIG. 7 is a health assessment model according to the present application;
FIG. 8 is a model of performance trend prediction according to the present application;
fig. 9 is a flowchart for accurately predicting and locating a battery fault according to the present application.
Detailed Description
The technical scheme of the application is further described below with reference to the accompanying drawings and examples.
The method for pre-judging and positioning the battery faults of the battery compartment of the energy storage power station comprises the steps of single battery health degree analysis, battery performance change trend prediction and accurate pre-judging and positioning of the battery faults. The method comprises the steps of obtaining upper and lower limit thresholds of single batteries and evaluating the health state of the single batteries in the current state through analyzing the health degree of the single batteries, and replacing corresponding battery modules of a battery compartment of the energy storage power station if the health degree of one or more single batteries is too low; and meanwhile, the single battery health threshold value, a module set by a Battery Management System (BMS) and the voltage, current, temperature and SOC threshold value of a battery cluster are taken as the criteria for battery fault pre-judgment. Predicting the change trend of battery performance in a future period (hour level or day level) based on historical data of a past period by predicting the change trend of battery performance, and if predicting that a certain performance and/or certain performance of a battery monomer and/or a battery pack and/or a battery cluster are about to exceed a threshold value, performing fault pre-judgment and positioning by accurately pre-judging and positioning the battery fault; meanwhile, if detecting that some performance and/or some performance of the battery cells and/or the battery pack and/or the battery cluster may be suddenly changed in a very short time, the battery fault accurate pre-judging and positioning function can be used for predicting faults possibly occurring in a short time (in minutes), and the fault expert database is used for positioning and identifying the corresponding faults.
(1) And (5) analyzing the health degree of the single battery.
The battery health degree analysis comprises three links of historical data analysis, threshold value determination and health degree analysis.
Historical data analysis includes selecting statistics by building a principal component model. The collection and preprocessing of data is the basis for establishing a statistical model, and data which is most effective for monitoring the production process and has influence on the quality of the product are selected, and certain correlation should be formed between the data. And screening the data meeting the conditions through the principal component model. In this embodiment, the voltage, temperature and SOC data of the battery are finally screened out and analyzed.
The threshold determination includes: for two weeks of monomer data, a kernel density estimation is used, and a probability density function is obtained by combining a statistical method, so that corresponding control limits (namely, threshold values) are calculated, as shown in fig. 1.
As shown in fig. 2, the standard of health score of the single battery according to the present application is shown. The health degree analysis comprises the steps of obtaining the overrun frequency of the latest several days of voltage, temperature and SOC as health degree according to the control limits of the first two weeks to evaluate the health state of each monomer; the health degree reflects the possibility that the performance indexes (mainly including voltage, SOC and temperature) of the single battery are in the upper and lower limits of the normal working range for a long time in the future. Wherein voltage health is defined as the frequency of occurrence of exceeding alarm values in the voltage data acquisition information over the past two weeks, as shown in fig. 3. The SOC health is defined as how often the SOC data acquisition information occurs beyond the alarm value over the past two weeks, as shown in fig. 4. Temperature health is defined as the frequency of occurrence in the past two weeks of temperature data acquisition information that exceeds an alarm value, as shown in fig. 5.
(2) And predicting the variation trend of the battery performance.
The battery performance trend prediction includes: the historical data analysis searches for health factors, establishes a health state evaluation model and predicts trend, as shown in figure 6.
The historical data analysis and searching for health factors comprise analyzing the similarity degree of development trends of the two factors of the feature quantity to be detected and the capacity by a gray level correlation analysis method, and measuring the correlation between the two sequences; the degree of linear correlation of two variables is measured through the Pearson correlation coefficient; finally, whether the two variables are strictly monotonous or not is measured through the spearman grade correlation coefficient, and then the health factor is determined;
as shown in fig. 7, the establishment of the health state estimation model refers to establishing the health state estimation model by using a health factor having a strong correlation with the battery capacity through the gray coefficient correlation analysis, training the model and obtaining a lithium ion battery health state estimation curve, and the health state estimation model is realized by using an Echo State Network (ESN) algorithm.
The construction flow of the health state evaluation model based on the Echo State Network (ESN) algorithm is as follows: (1) data preparation; (2) ESN modeling; and (3) model prediction.
The data preparation includes: extracting monitoring time, monitoring voltage and capacity value in a constant current charging mode; the voltage at a certain time interval from the beginning is selected, and the time interval voltage difference sequence for voltage difference construction is obtained.
The ESN modeling includes: dividing the obtained time interval parameter sequence to be measured and the corresponding capacity sequence into a training set and a testing set. The model parameters of ESN are four in total, the number N of the reserve pool processing units, the spectrum radius SR, the reserve pool input telescopic scale IS and the input unit displacement IF. The training process is to train the output weight of ESN model through logistic regression to minimize the root mean square error between the output result and the actual result, and optimize the four parameter values through the cross verification method to make the output result reach the optimal state.
Modeling predictions include: and inputting the parameters to be tested at equal time intervals of the test data into a model, estimating the capacity value of the battery, comparing the capacity estimated value of the battery with a true value, and measuring the accuracy of the model.
As shown in fig. 8, the trend prediction includes predicting the state of health factor of the future cycle number of the battery through a long short time memory network (LSTM) algorithm, inputting the predicted result into a health state estimation model to obtain a capacity estimation value, and predicting the remaining life of the lithium ion battery.
The operation strategy of the LSTM neural network model is as follows: the experiment sequentially recursively extracts 10 data as input by selecting the isochronous voltage difference data from the 1 st to (n-10) th cycle period, sequentially recursively extracts the data from the 10 th to n th cycle period, and sequentially recursively extracts the 10 data as output. All data is divided into training and test validation sets. Because LSTM neural network is sensitive to the value, normalize at first, convert all data to between 0 and 1, after predicting, normalize inversely, get the true value of the predicted result.
(3) And (5) accurately pre-judging and positioning the battery faults.
Accurate prejudgement and location of battery trouble include: analyzing historical data of the past 1 hour; analyzing contribution degree of historical data through multivariate statistics by an elastic network method; based on the above, whether a fault occurs in the future 10min is judged according to a threshold value (namely, fault identification is carried out); based on failure mode analysis, performing hazard degree analysis on the pre-judging failure; and then performing fault positioning and fault tracing of the faults according to the fault tree and the incidence matrix of the faults and the monitoring signals, as shown in fig. 9.
The historical data mainly comprises a stack number, a cluster number, a stack voltage, a stack SOC, a stack temperature difference, a cluster voltage, a cluster SOC, a cluster pressure difference, a cluster temperature difference, an insulation resistance, an insulation resistance+, a single cell number, a single cell voltage, a single cell temperature, a single cell SOC and the like. Wherein-sum+ represents the insulation resistance measured at the total positive and total negative of the battery system
The contribution analysis includes: and predicting the numerical value of the battery in a future period through an elastic network model EN, analyzing the contribution degree through multivariate statistics, analyzing which time point is most likely to generate faults in the period, and performing pre-judgment on the faults.
The elastic network model EN prediction method comprises the following steps: .
(1) EN is initialized.
(2) And training a model. The EN model is fitted with three-dimensional input data (cell voltage, SOC, cell temperature) over a total of one hour from time 0-59 (minutes) past and SOC output data from time 0-59 (minutes) past.
(3) And (5) SOC prediction. By using the EN model obtained by training, the SOC data at the 60 th moment is estimated by using three-dimensional input data (monomer voltage, SOC and monomer temperature) at the 60 th moment (in the order of minutes), so that the monomer SOC value at the 1 st minute after the current moment can be obtained.
(4) The SOC of the next time is predicted. The first to third steps are performed back, except that the second step is fitted with three-dimensional inputs (cell voltage, SOC, cell temperature) at times 0-58 (minutes) and SOC at times 1-59 (minutes). And thirdly, predicting the SOC at the 61 st moment by using the input at the 60 th moment, so that the monomer SOC value at the 2 nd minute after the current moment can be obtained.
(5) Repeating the fourth step, and returning to execute model initialization, model training and SOC prediction, wherein the model training input is unchanged, the output is advanced for one moment, and each time forward prediction is performed for 1 minute, until the third step is to predict the SOC at the 70 th moment by using the input at the 60 th moment. Thus, the predicted value of the SOC of the monomer 10 minutes after the current moment can be obtained.
The accurate positioning of the battery faults comprises the construction of a typical fault tree of the energy storage power station, the analysis of the influence relation of the main fault type of the power station and relevant characterization parameters thereof, and the analysis of the influence weights of the main fault type of the power station and the relevant characterization parameters thereof. Firstly, combing faults possibly occurring in the aspects of hardware and software of an energy storage power station, combing faults possibly occurring in the aspects of functions of the energy storage power station, and constructing a typical fault tree of the energy storage power station; secondly, through construction of a fault tree, analysis of the occurrence of the main fault type of the energy storage power station can be represented through characterization parameters (including but not limited to voltage, current, temperature, SOC, insulation resistance and the like), and further influence relation analysis of the main fault type and related characterization parameters is established; and thirdly, establishing an influence weight analysis of the main fault type and related characterization parameters thereof based on the Delphi method, and constructing a fault expert database of the battery compartment of the energy storage power station. Finally, the type and weight relation of the characterization parameters are pre-judged through a battery fault accurate pre-judging procedure, and accurate pre-judging and positioning are carried out on the battery fault.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application and not for limiting the same, although the present application is described in detail with reference to the above embodiments, it should be understood that: all methods or products that may be modified or substituted by one or more of ordinary skill in the art based on the described embodiments of the application without inventive effort are intended to be within the scope of the claims.
Claims (8)
1. The method for pre-judging and positioning the battery faults of the battery compartment of the energy storage power station is characterized by comprising the following steps:
(1) Obtaining upper and lower limit thresholds of battery operation parameters and the battery health state in the current state through analyzing the health degree of the single battery;
(2) Predicting the battery performance change trend in a future period based on the historical data of the past period, and performing pre-judgment and positioning on possible faults in the future by combining an upper threshold and a lower threshold; meanwhile, the performances of the battery monomer, the battery pack or the battery cluster are subjected to abrupt change prediction in a short time, the faults possibly occurring in the performances of the battery monomer, the battery pack or the battery cluster in a short time are predicted through accurate battery fault prediction and positioning, and fault positioning is performed;
the accurate pre-judging and positioning of the battery fault comprises the following steps:
(31) Analyzing historical data within a period of time; the historical data comprises stack numbers, cluster numbers, stack voltages, stack SOCs, stack temperature differences, cluster voltages, cluster SOCs, cluster pressure differences, cluster temperature differences, insulation resistances-, insulation resistances+, single cell numbers, single cell voltages, single cell temperatures and single cell SOCs;
(32) Predicting a data value of the battery in a future period of time through an elastic network model;
(33) Judging whether a fault occurs in a short time in the future according to a threshold value;
(34) Performing fault positioning and fault tracing of faults according to the fault tree and the incidence matrix of the faults and the monitoring signals;
the step (32) includes the steps of:
(321) Initializing a model;
(322) Model training: by 0 totThree-dimensional input data at-1 moment and 0-t-SOC output data training model at time 1, the three-dimensional input data comprising cell voltage, SOC and cell temperature;
(323) SOC prediction: using the elastic network model obtained by training, usingtEstimation of time three-dimensional input datatSOC at time;
(324) Returning to execute model initialization, model training and SOC prediction, wherein the model training is performed by 0-0% t-2 three-dimensional input at time and 1- tSOC training model at time-1, and SOC prediction using the firsttThree-dimensional input data prediction of time of daytSOC at +1;
(325) Repeating executionStep (324), each time the model training input is unchanged, the output advances one time, the SOC prediction advances one time until the SOC prediction advances gradually to use the first timetInput prediction of time of dayt+ΔtUntil the SOC at the moment, delta after the current moment is obtainedtAnd (5) predicting the SOC value of the inner single battery.
2. The method for predicting and locating a battery fault in a battery compartment of an energy storage power station according to claim 1, wherein the analyzing the health of the single battery comprises the following steps:
(11) Selecting statistics by principal component analysis;
(12) The selected historical data is subjected to kernel density estimation, a probability density function is obtained by combining a statistical method, and a corresponding threshold value is calculated according to the probability density function; wherein the selected historical data includes voltage, temperature and SOC data of the battery within two weeks;
(13) The overrun frequency of each statistic is calculated as a health degree based on the selected historical data to evaluate the health status of each individual.
3. The method for predicting and locating a battery fault in a battery compartment of an energy storage power station according to claim 1, wherein the battery performance trend in the step (2) comprises the following steps:
(21) Analyzing the similarity degree of the development trend of each factor of the feature quantity to be measured and the capacity two sequences by a gray level correlation analysis method, and measuring the correlation between the two sequences; the degree of linear correlation of two variables is measured through the Pearson correlation coefficient; finally, whether the strict monotonous correlation index between the two variables is measured through the spearman grade correlation coefficient, and the characteristic quantity with the strict monotonous correlation index is used as a health factor;
(22) Establishing a health state estimation model by adopting an echo state network algorithm, and training the health state estimation model by taking the health factors determined in the step (21) as an input data set and the battery capacity as an output data set to obtain a lithium ion battery health state estimation curve;
(23) And predicting the state of health of the lithium ion battery by adopting a degradation model algorithm, predicting the future state of health factors of the battery by using a long-short-term memory neural network algorithm, inputting the predicted result into the degradation model algorithm to obtain a capacity estimated value, and predicting the residual life of the lithium ion battery.
4. The method for predicting and locating battery faults of battery compartment of energy storage power station according to claim 3, wherein the method comprises the following steps: and (3) the health factor in the step (21) is an equal-time-interval voltage difference sequence, and the voltage difference is obtained according to the monitoring time, the monitoring voltage and the capacity value in the constant-current charging mode.
5. The method of claim 3, wherein the step (22) comprises the steps of: dividing the obtained time interval parameter sequence to be measured and the corresponding capacity sequence into a training set and a testing set, wherein model parameters of a health state estimation model comprise the number N of storage pool processing units, the spectrum radius SR, and the storage pool input telescopic scale IS and the input unit displacement IF; the training process is to train the output weight of the health state estimation model through logistic regression, so that the root mean square error of the output result and the actual result is minimum, and optimize each parameter value through a cross verification method, so that the output result reaches the optimal state; and finally, inputting the parameters to be tested at equal time intervals of the test data into a model, estimating the capacity value of the battery, comparing the capacity estimated value of the battery with a true value, and measuring the accuracy of the model.
6. The method for predicting and locating a battery fault in a battery compartment of an energy storage power station according to claim 3, wherein the long-short-term memory neural network algorithm in step (23) comprises the following steps: sequentially recursively inputting m data by selecting the isochronous voltage difference data from the 1 st to (n-m) th cycle periods, sequentially recursively inputting the m data by selecting the data from the m to the n th cycle periods, and sequentially recursively outputting the m data; dividing all data into a training set and a test verification set, training a long-short-term memory neural network model through the training set, firstly normalizing the data during training, and then carrying out inverse normalization after prediction is finished to obtain a true value of a predicted result.
7. The method for predicting and locating battery faults of battery compartment of energy storage power station according to claim 1, wherein the method comprises the following steps: the step (34) includes: firstly, a typical fault tree of the energy storage power station is constructed by combing faults possibly occurring in the energy storage power station; secondly, analyzing characterization parameters of main fault types of the energy storage power station through construction of a fault tree, and further establishing influence relations of the main fault types and relevant characterization parameters of the main fault types; thirdly, establishing an influence weight analysis of a main fault type and related characterization parameters thereof based on a Delphi method, and constructing a fault expert database of a battery compartment of the energy storage power station; finally, the type and weight relation of the characterization parameters are pre-judged through a battery fault accurate pre-judging procedure, and accurate pre-judging and positioning are carried out on the battery fault.
8. A computer readable storage medium storing a computer program, wherein the program causes a computer to perform the steps of the method for predicting and locating battery faults in a battery compartment of an energy storage power station according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011543908.9A CN112731159B (en) | 2020-12-23 | 2020-12-23 | Method for pre-judging and positioning battery faults of battery compartment of energy storage power station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011543908.9A CN112731159B (en) | 2020-12-23 | 2020-12-23 | Method for pre-judging and positioning battery faults of battery compartment of energy storage power station |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112731159A CN112731159A (en) | 2021-04-30 |
| CN112731159B true CN112731159B (en) | 2023-09-22 |
Family
ID=75604819
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011543908.9A Active CN112731159B (en) | 2020-12-23 | 2020-12-23 | Method for pre-judging and positioning battery faults of battery compartment of energy storage power station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112731159B (en) |
Families Citing this family (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113255764A (en) * | 2021-05-21 | 2021-08-13 | 池测(上海)数据科技有限公司 | Method, system and device for detecting electrochemical energy storage system fault by using machine learning |
| CN113282886B (en) * | 2021-05-26 | 2021-12-14 | 北京大唐神州科技有限公司 | Bank loan default judgment method based on logistic regression |
| CN113219343A (en) * | 2021-06-02 | 2021-08-06 | 上海玫克生储能科技有限公司 | Lithium battery health state prediction method, system, equipment and medium based on elastic network |
| CN113554224B (en) * | 2021-07-20 | 2022-11-25 | 上海航天测控通信研究所 | Fault diagnosis method and system combining multipoint statistics with health trend prediction |
| CN113791350B (en) * | 2021-08-06 | 2024-05-14 | 陕西汽车集团股份有限公司 | Battery fault prediction method |
| CN113625692B (en) * | 2021-08-23 | 2023-03-31 | 公安部交通管理科学研究所 | Electric automobile battery security inspection system based on fault injection |
| CN113655391A (en) * | 2021-08-26 | 2021-11-16 | 江苏慧智能源工程技术创新研究院有限公司 | Energy storage power station battery fault diagnosis method based on LightGBM model |
| JP7366975B2 (en) * | 2021-09-30 | 2023-10-23 | 本田技研工業株式会社 | Battery deterioration estimation device, battery deterioration estimation system, battery deterioration estimation method, and program |
| CN114252103B (en) * | 2021-11-02 | 2023-12-22 | 国网吉林省电力有限公司电力科学研究院 | Fusion power station operation fault prediction method |
| CN114236409B (en) * | 2021-11-09 | 2024-01-19 | 长安大学 | Power battery pack fault diagnosis device, method and system |
| CN114297255B (en) * | 2021-12-17 | 2024-04-19 | 中电信数智科技有限公司 | Network quality work order fault early warning method based on log analysis |
| CN114490836B (en) * | 2022-04-15 | 2022-08-02 | 国网天津市电力公司电力科学研究院 | Data mining processing method suitable for electric vehicle charging fault |
| CN116150666B (en) * | 2022-10-08 | 2023-10-31 | 深圳先进技术研究院 | Energy storage system fault detection method and device and intelligent terminal |
| CN115805810A (en) * | 2022-10-31 | 2023-03-17 | 宁德时代新能源科技股份有限公司 | Battery failure prediction method, apparatus, device, storage medium, and program product |
| CN115542172A (en) * | 2022-12-01 | 2022-12-30 | 湖北工业大学 | Power battery fault detection method, system, device and storage medium |
| CN115963408B (en) * | 2022-12-19 | 2024-04-16 | 北京双登慧峰聚能科技有限公司 | Energy storage power station single battery fault early warning system and method |
| CN115877222A (en) * | 2023-02-14 | 2023-03-31 | 国网浙江省电力有限公司宁波供电公司 | Energy storage power station fault detection method and device, medium and energy storage power station |
| CN115951230B (en) * | 2023-03-10 | 2023-06-16 | 苏州时代华景新能源有限公司 | Abnormality detection method and system for lithium battery energy storage box |
| CN115993540A (en) * | 2023-03-22 | 2023-04-21 | 钛玛科（北京）工业科技有限公司 | Comprehensive evaluation method and equipment for lithium battery health |
| CN116027206A (en) * | 2023-03-29 | 2023-04-28 | 安徽中科中涣智能装备股份有限公司 | Multi-element perception grading early warning intelligent monitoring system based on station-level energy storage |
| CN116736134B (en) * | 2023-06-20 | 2024-01-05 | 深圳市赛格瑞电子有限公司 | Real-time energy storage battery data monitoring method and device |
| CN116500460B (en) * | 2023-06-29 | 2023-08-22 | 北京云控安创信息技术有限公司 | Cloud computing-based battery health state diagnosis and prediction system for Internet of things |
| CN116736142B (en) * | 2023-08-14 | 2023-10-24 | 新誉集团有限公司 | Method, system and device for early warning of health condition of battery pack |
| CN116819345B (en) * | 2023-08-29 | 2023-11-21 | 武汉船用电力推进装置研究所(中国船舶集团有限公司第七一二研究所) | Battery system fault identification method and device, electronic equipment and storage medium |
| CN116973769B (en) * | 2023-09-22 | 2023-12-26 | 中国华能集团清洁能源技术研究院有限公司 | Method, device, electronic equipment and medium for detecting state of energy storage battery |
| CN117013583B (en) * | 2023-09-28 | 2024-01-16 | 烟台开发区德联软件有限责任公司 | Fault early warning method and system for electrochemical energy storage power station |
| CN117117356B (en) * | 2023-10-24 | 2024-03-12 | 内蒙古中电储能技术有限公司 | Energy storage battery pack thermal runaway monitoring and tracing method and system |
| CN117216603B (en) * | 2023-11-07 | 2024-02-06 | 张家港长寿工业设备制造有限公司 | Method and system for predicting faults of tubular falling film evaporator |
| CN117289145B (en) * | 2023-11-27 | 2024-04-19 | 宁德时代新能源科技股份有限公司 | Fault analysis method, data acquisition method, device, equipment, system and medium |
| CN118068234A (en) * | 2024-04-19 | 2024-05-24 | 内蒙古中电储能技术有限公司 | Wiring abnormity monitoring method and system based on battery compartment state analysis |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102636740A (en) * | 2012-04-18 | 2012-08-15 | 南京航空航天大学 | Method for predicting faults of power electronic circuit based on FRM-RVM (fuzzy rough membership-relevant vector machine) |
| CN105241680A (en) * | 2015-08-26 | 2016-01-13 | 电子科技大学 | Health state assessment method for rotary machine based on probability density function |
| CN106844775A (en) * | 2017-03-02 | 2017-06-13 | 上海交通大学 | Spacecraft malfunction rapid detection system |
| CN109460574A (en) * | 2018-10-09 | 2019-03-12 | 南京航空航天大学 | A kind of prediction technique of aero-engine remaining life |
| CN110363339A (en) * | 2019-07-05 | 2019-10-22 | 南京简睿捷软件开发有限公司 | A kind of method and system carrying out predictive maintenance based on the parameter of electric machine |
| CN110763929A (en) * | 2019-08-08 | 2020-02-07 | 浙江大学 | Intelligent monitoring and early warning system and method for convertor station equipment |
| CN111222702A (en) * | 2020-01-09 | 2020-06-02 | 中国电力科学研究院有限公司 | Method and system for predicting operation change trend of regulation and control system |
| CN111443294A (en) * | 2020-04-10 | 2020-07-24 | 华东理工大学 | Method and device for indirectly predicting remaining life of lithium ion battery |
| CN111614504A (en) * | 2020-06-02 | 2020-09-01 | 国网山西省电力公司电力科学研究院 | Power grid regulation and control data center service characteristic fault positioning method and system based on time sequence and fault tree analysis |
| CN111832221A (en) * | 2020-06-23 | 2020-10-27 | 华中科技大学无锡研究院 | Lithium battery life prediction method based on feature screening |
| CN111880099A (en) * | 2020-06-30 | 2020-11-03 | 国网浙江省电力有限公司电力科学研究院 | Method and system for predicting service life of battery monomer in energy storage power station |
-
2020
- 2020-12-23 CN CN202011543908.9A patent/CN112731159B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102636740A (en) * | 2012-04-18 | 2012-08-15 | 南京航空航天大学 | Method for predicting faults of power electronic circuit based on FRM-RVM (fuzzy rough membership-relevant vector machine) |
| CN105241680A (en) * | 2015-08-26 | 2016-01-13 | 电子科技大学 | Health state assessment method for rotary machine based on probability density function |
| CN106844775A (en) * | 2017-03-02 | 2017-06-13 | 上海交通大学 | Spacecraft malfunction rapid detection system |
| CN109460574A (en) * | 2018-10-09 | 2019-03-12 | 南京航空航天大学 | A kind of prediction technique of aero-engine remaining life |
| CN110363339A (en) * | 2019-07-05 | 2019-10-22 | 南京简睿捷软件开发有限公司 | A kind of method and system carrying out predictive maintenance based on the parameter of electric machine |
| CN110763929A (en) * | 2019-08-08 | 2020-02-07 | 浙江大学 | Intelligent monitoring and early warning system and method for convertor station equipment |
| CN111222702A (en) * | 2020-01-09 | 2020-06-02 | 中国电力科学研究院有限公司 | Method and system for predicting operation change trend of regulation and control system |
| CN111443294A (en) * | 2020-04-10 | 2020-07-24 | 华东理工大学 | Method and device for indirectly predicting remaining life of lithium ion battery |
| CN111614504A (en) * | 2020-06-02 | 2020-09-01 | 国网山西省电力公司电力科学研究院 | Power grid regulation and control data center service characteristic fault positioning method and system based on time sequence and fault tree analysis |
| CN111832221A (en) * | 2020-06-23 | 2020-10-27 | 华中科技大学无锡研究院 | Lithium battery life prediction method based on feature screening |
| CN111880099A (en) * | 2020-06-30 | 2020-11-03 | 国网浙江省电力有限公司电力科学研究院 | Method and system for predicting service life of battery monomer in energy storage power station |
Non-Patent Citations (1)
| Title |
|---|
| 郭东亮等.储能电站用锂离子电池热失控早期预警参数研究.《消防科学与技术》.2020,第39卷(第8期),第1156-1159页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112731159A (en) | 2021-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112731159B (en) | Method for pre-judging and positioning battery faults of battery compartment of energy storage power station | |
| CN112763929B (en) | Method and device for predicting health of battery monomer of energy storage power station system | |
| CN114114049B (en) | Lithium ion battery life prediction method based on sample migration | |
| CN112327167B (en) | Battery capacity diving risk assessment method and system | |
| CN112561736A (en) | Fault diagnosis system and method for relay protection device of intelligent substation | |
| CN114371409B (en) | Training method of battery state prediction model, battery state prediction method and device | |
| CN111614491A (en) | Power monitoring system oriented safety situation assessment index selection method and system | |
| CN113447828B (en) | Lithium battery temperature estimation method and system based on Bayesian neural network | |
| CN116401585B (en) | Energy storage battery failure risk assessment method based on big data | |
| CN113866642A (en) | Lithium ion battery fault diagnosis method based on gradient lifting tree | |
| CN113821976A (en) | Lithium battery fault diagnosis modeling method based on integrated algorithm | |
| CN111983474A (en) | Lithium ion battery life prediction method and system based on capacity decline model | |
| CN114545234A (en) | Method for monitoring battery state based on battery temperature gradient change | |
| CN115480180A (en) | New energy battery health diagnosis and analysis method | |
| CN114462259A (en) | SOC-based battery remaining life prediction method and system, automobile and medium | |
| CN114720883A (en) | Lithium battery fault identification method based on isolated forest algorithm | |
| CN116754984A (en) | Battery consistency detection method and related device | |
| CN117113232A (en) | Thermal runaway risk identification method for lithium ion battery pack of electric automobile | |
| CN117330987B (en) | Method, system, medium and apparatus for time-based battery state of health assessment | |
| CN114879070A (en) | Battery state evaluation method and related equipment | |
| Diao et al. | Research on Electric Vehicle Charging Safety Warning Based on A-LSTM Algorithm | |
| Sohn et al. | CNN-based online diagnosis of knee-point in Li-ion battery capacity fade curve | |
| CN116901707A (en) | Power battery pack fault early warning method, system and vehicle | |
| CN115128468A (en) | Chemical energy storage battery PHM undervoltage fault prediction method | |
| CN117686935B (en) | Battery RUL prediction method based on voltage probability density |
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 |