CN111812518B - Battery state monitoring method, storage medium and system - Google Patents
Battery state monitoring method, storage medium and system Download PDFInfo
- Publication number
- CN111812518B CN111812518B CN201911359532.3A CN201911359532A CN111812518B CN 111812518 B CN111812518 B CN 111812518B CN 201911359532 A CN201911359532 A CN 201911359532A CN 111812518 B CN111812518 B CN 111812518B
- Authority
- CN
- China
- Prior art keywords
- battery
- data
- historical
- state monitoring
- model
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/382—Arrangements for monitoring battery or accumulator variables, e.g. SoC
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L58/00—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles
- B60L58/10—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries
- B60L58/12—Methods or circuit arrangements for monitoring or controlling batteries or fuel cells, specially adapted for electric vehicles for monitoring or controlling batteries responding to state of charge [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]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/60—Other road transportation technologies with climate change mitigation effect
- Y02T10/70—Energy storage systems for electromobility, e.g. batteries
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Power Engineering (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Secondary Cells (AREA)
Abstract
The invention provides a battery state monitoring method, a storage medium and a system, wherein the method comprises the following steps: acquiring a corresponding relation between historical operation parameters of the battery and the battery core range; acquiring the operation parameters of the battery in a preset time period, and determining the estimated battery core range of the battery at the current moment according to the operation parameters of the battery and the corresponding relation; the interval between the preset time period and the current time is smaller than a set duration; determining whether the battery is in a safe state according to the estimated battery core range of the battery; the battery cell electrode difference is the difference between the maximum battery cell voltage value and the minimum battery cell voltage value in the battery. According to the scheme, after the electric vehicle leaves the factory, the consistency of the battery cells of the battery can be continuously monitored, battery faults caused by the problem of the consistency of the battery cells can be found in time, and the safety and stability of the electric vehicle in the running process are ensured.
Description
Technical Field
The invention relates to the technical field of battery safety monitoring in electric vehicles, in particular to a battery state monitoring method, a storage medium and a system.
Background
With the popularization of electric vehicles and the application of internet of vehicles, more and more electric vehicles enter the consumer market. In the faults of the electric vehicle at present, the consistency of the battery core is poor, the SOC (system on chip) is low, the single battery is in under-voltage alarm, the rechargeable energy storage system is not matched, the type of the vehicle-mounted energy storage device is under-voltage, which is the main fault of the electric vehicle, the faults can be ranked five before in all fault types of the electric vehicle, and the five types of faults account for 78% of the total fault quantity.
As the power source of the electric vehicle, the capacity of the battery is continuously attenuated and the potential safety hazard of the battery is increased along with the increase of the charge and discharge times and the driving mileage, and the possibility of faults of the battery is also higher. At present, in an electric vehicle, the influence of the consistency of the battery cells in the use process of the battery on the safety and the failure rate of the battery is obvious.
In the prior art, a method for evaluating the consistency of battery cells mainly adopts a static test method to measure and obtain the voltage, the capacity, the internal resistance and the SOC distribution value of battery cells in the battery, then adds an evaluation weight value for each measured value, carries out weighted average calculation on each measured value to obtain the evaluation coefficient of the consistency of the battery cells, adopts the method to have more test values required to be measured, has long operation complexity time, can only be used for experimental operation of battery production enterprises in the battery production process, and is difficult to effectively monitor the consistency of the battery cells once the battery leaves the factory. In addition, lithium batteries are often used in batteries, and the lithium batteries are typical dynamic nonlinear electrochemical systems, and internal parameters (cell capacity, cell internal resistance, cell internal temperature, etc.) of the lithium batteries are difficult to measure in real time when the lithium batteries are applied, so that the state recognition and state estimation of battery change still have great challenges. In addition, although the electric vehicle needs to upload the vehicle operation data to the cloud server according to the regulations for the manufacturer to monitor the safety performance of the electric vehicle in the operation process, the data uploading is not performed in real time, and has serious hysteresis, so that the battery fault problem is difficult to find at the first time.
Therefore, there is a need for a solution that can continuously monitor the safety of a battery in an electric vehicle in real time after the electric vehicle leaves the factory, to solve the above problems.
Disclosure of Invention
The embodiment of the invention aims to provide a battery state monitoring method, a storage medium and a system, which are used for solving the technical problem that the battery cannot be safely monitored after an electric vehicle leaves a factory in the prior art.
To this end, the present invention provides a battery state monitoring method, comprising the steps of:
acquiring a corresponding relation between historical operation parameters of the battery and the battery core range;
acquiring the operation parameters of the battery in a preset time period, and determining the estimated battery core range of the battery at the current moment according to the operation parameters of the battery and the corresponding relation; the interval between the preset time period and the current time is smaller than a set duration;
determining whether the battery is in a safe state according to the estimated battery core range of the battery;
the battery cell electrode difference is the difference between the maximum battery cell voltage value and the minimum battery cell voltage value in the battery.
Optionally, in the battery state monitoring method described above:
the historical operation parameters of the battery are historical charging parameters when the battery is in a charging process in a historical time period;
and the operation parameters of the battery in the preset time period are charging parameters when the battery is in a charging process in the preset time period.
Optionally, in the battery state monitoring method described above:
the step of obtaining the corresponding relation between the historical charging parameters and the battery cell electrode difference according to the historical charging data of the battery comprises the following steps: extracting a battery current value, a battery temperature, a battery voltage value and/or a battery SOC value in the historical charging data as the historical charging parameters, and taking the historical charging parameters as input samples; extracting battery cell electrode difference in the historical charging data as an output sample, wherein the acquisition time of the output sample is later than the acquisition time of the input sample; training a machine learning model according to the input sample and the output sample, and taking the trained machine learning model as a state monitoring model; the state monitoring model is used for representing the corresponding relation;
in the step of determining the estimated battery cell range of the battery at the current moment according to the corresponding relation between the charging parameter of the battery and the corresponding relation: extracting a battery current value, a battery temperature, a battery voltage value and/or a battery SOC value in the battery charging parameters as input parameters to be input into the state monitoring model, and taking the output of the state monitoring model as the estimated battery core range of the battery; and taking the time interval between the collection time of the output sample and the collection time of the input sample as the set duration.
Optionally, in the above method for monitoring a battery state, the step of obtaining the correspondence between the historical charging parameter and the battery cell range according to the historical charging data of the battery includes:
data grouping: dividing the historical charging data into two groups, wherein the data volume of the first group of historical charging data is larger than that of the second group of historical charging data;
model training: extracting an input sample and an output sample in a first group of historical charging data, and training the machine learning model to obtain an initial state monitoring model;
and (3) model verification: and extracting an input sample and an output sample in the second set of historical charging data to verify the initial state monitoring model, if the initial state monitoring model passes the verification, taking the initial state monitoring model as the state monitoring model, otherwise, increasing the data volume of the first set of historical charging data, and returning to the model training step.
Optionally, in the method for monitoring a battery state, the step of obtaining the correspondence between the historical charging parameter and the battery cell range according to the historical charging data of the battery includes:
data grouping: dividing the historical charging data into K groups, taking one group of historical charging data as test sample data, and taking the rest groups of historical charging data as training sample data;
model training: extracting an input sample and an output sample in training sample data to train the machine learning model to obtain an initial state monitoring model;
and (3) model verification: extracting an input sample and an output sample in test sample data to verify the initial state monitoring model, and taking the initial state monitoring model as an alternative state monitoring model if the initial state monitoring model passes the verification;
and a cross verification step: taking the other set of historical charging data as test sample data, taking the rest set of historical charging data as training sample data, and repeating the model training step and the model verification step until all K sets of historical charging data have been selected as test sample data;
model determining: and determining the state monitoring model according to all the alternative state monitoring models.
Optionally, in the above battery state monitoring method, in the model determining step:
taking the alternative state monitoring model with the highest verification result accuracy as a state monitoring model; or taking the average value of the adjustment coefficients in all the alternative state monitoring models as an actual adjustment coefficient to obtain the state monitoring model.
Optionally, in the method for monitoring a battery state, the step of determining whether the battery is in a safe state according to the estimated battery core limit of the battery includes:
the cell consistency parameter of the battery is obtained according to the following mode:
wherein V is gap_1 For the estimated battery cell range, V gap_0 A battery cell electrode difference threshold value is preset;
if R_V > R_V_0, judging that the battery is in an unsafe state, wherein R_V_0 is a preset battery cell consistency parameter threshold value.
Optionally, in the above method for monitoring a battery state, before the step of obtaining the correspondence between the historical charging parameter and the battery cell range according to the historical charging data of the battery, the method further includes:
acquiring a historical state data set of the battery in a historical time period, wherein each historical state data in the historical state data set is configured with a state identifier;
and if the state identifier represents a charging state, taking the historical state data as the historical charging data.
Optionally, in the method for monitoring a battery state, the step of obtaining a set of historical state data of the battery in a historical period, where each of the historical state data in the set of historical state data is configured with a state identifier further includes:
and screening out repeated historical state data, error historical state data and incomplete historical state data in the historical state data set.
The present invention also provides a computer-readable storage medium having stored therein program instructions, the computer executing the battery state monitoring method of any one of the above after reading the program instructions.
The invention also provides a battery charging process state monitoring system, which comprises at least one processor and at least one memory, wherein program instructions are stored in at least one memory, and the battery state monitoring method is executed after the program instructions are read by at least one processor.
Compared with the prior art, the technical scheme provided by the invention has at least the following beneficial effects:
according to the battery state monitoring method, the storage medium and the system, the influence of the historical operation data of the battery on the future battery cell range can be obtained according to the historical operation data of the battery, so that the corresponding relation between the battery parameters and the battery cell range in the historical operation process of the battery is determined; therefore, the estimated battery cell range of the battery at the current moment can be predicted according to the corresponding relation and the battery operation data in a period of time before the current moment, so that whether the battery is in a safe state or not can be determined according to the estimated battery cell range of the battery at the current moment. Therefore, after the electric vehicle leaves the factory, the consistency of the battery cells of the battery can be continuously monitored, battery faults caused by the consistency problem of the battery cells can be timely found, and the safety and stability of the electric vehicle in the running process are ensured.
Drawings
FIG. 1 is a flow chart of a method for monitoring battery status according to an embodiment of the present invention;
FIG. 2 is a flow chart of a battery condition monitoring method according to another embodiment of the present invention;
FIG. 3 is a schematic diagram of a data processing flow for predicting cell consistency using a linear regression model according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of deviation values between a cell consistency result predicted by a linear regression model and an actual cell consistency result according to an embodiment of the present invention;
fig. 5 is a schematic diagram of hardware connection relation of a battery charging process state monitoring system according to an embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. The components of the embodiments of the present invention generally described and provided in the figures herein may be arranged and designed in a wide variety of different configurations.
In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings, are merely for convenience of description of the present invention, and are not to indicate or imply that the apparatus or component to be referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. Wherein the terms "first position" and "second position" are two different positions.
In the description of the present invention, it should be noted that, unless explicitly specified and limited otherwise, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be either fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between the two components. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
Example 1
The present embodiment provides a battery state monitoring method, which can be applied to a vehicle management platform, as shown in fig. 1, and includes the following steps:
s101: and obtaining the corresponding relation between the historical operation parameters of the battery and the battery core range. The vehicle management platform can monitor and manage a large number of electric vehicles and can collect and store all data in the running process of the electric vehicles. Each managed electric vehicle has a unique vehicle identifier, and accordingly, a battery of each electric vehicle and related data of the battery in the running process are allocated with a specific data identifier and stored, so that the battery can be uniquely determined by identifying the data identifier or the vehicle identifier in the step. And when a certain time difference exists between the historical operation data of the battery and the battery cell range, for example, the historical operation data is the battery operation data in two months from 10 months to 11 months in 2019, in the time period, the battery operation parameters obtained by pushing the battery cell range of a certain battery and the corresponding time of the battery cell range forward for one week are taken as a group of data, so that the corresponding relation between the battery operation parameters and the battery cell range after one week is obtained. Accordingly, the time period can be adjusted according to actual needs.
S102: acquiring the operation parameters of the battery in a preset time period, and determining the estimated battery core range of the battery at the current moment according to the operation parameters of the battery and the corresponding relation; the interval between the preset time period and the current time is smaller than a set duration; the set time length in this step may be referred to according to the time interval between the history data and the battery cell electrode difference selected at the time of acquiring the correspondence in step S101.
S103: determining whether the battery is in a safe state according to the estimated battery core range of the battery; a standard cell range may be preset, and if the estimated cell range is not within the standard range, the battery may be considered unsafe. Wherein the battery cell electrode difference is the difference between the maximum battery cell voltage value and the minimum battery cell voltage value in the battery
According to the technical scheme provided by the embodiment, the safety state of the battery system is evaluated by establishing the historical operation parameters (current/SOC/temperature and the like) and the voltage range prediction algorithm of the battery, and the vehicle management platform can acquire battery data of the electric vehicle in a real-time long historical period, so that hidden battery state information and evolution rules thereof are mined from rated information and state monitoring data (voltage, current, temperature or SOC and the like) of the battery, the battery core consistency prediction of the battery in the charging process is realized, and the vehicle management platform is assisted in dynamic monitoring and predicting of the battery fault condition of the vehicle. In addition, because the battery historical operation data is adopted to predict the future battery state, the problems of data delay and hysteresis are avoided, the method is not limited by the test stage, and the battery state can be continuously monitored in real time after the electric vehicle leaves the factory.
Preferably, in the above scheme, the historical operation parameter of the battery is a historical charging parameter when the battery is in a charging process in a historical period of time; and the operation parameters of the battery in the preset time period are charging parameters when the battery is in a charging process in the preset time period. Referring to fig. 2, the above method may include the steps of:
s201: and obtaining the corresponding relation between the battery core range and the charging data of the battery according to the historical charging data of the battery. According to national standards, the components of battery operation data in an electric vehicle need to be implemented according to national standards, the data in the battery charging process need to be configured with charging identifiers, and the identifiers in the two charging conditions of the fast charging state and the slow charging state should be distinguished. The method may further comprise the steps of, prior to performing the step: obtaining a set of historical state data over a historical time period, each of the historical state data in the set of historical state data configured with a state identifier; and if the state identifier represents a charging state, taking the historical state data as the historical charging data. For the vehicle management platform, historical charging data can be conveniently extracted from battery operation data of the electric vehicle.
S202: obtaining a charging parameter of the battery in a charging process within a preset time period, and obtaining estimated battery core range of the battery according to the charging parameter and the corresponding relation;
s203: and determining whether the battery is in a safe state according to the estimated battery core range of the battery.
Namely, the data base of the battery cell consistency prediction is performed by the state parameters of the battery in the charging process. Because the vehicle keeps a static state in the charging process, the state parameter change of the battery is the most stable, the influence from the external environment is the least, and the prediction of the consistency of the battery cells by adopting the data in the charging process has higher accuracy.
Preferably, in the above method, the step S201 includes: extracting a battery current value, a battery temperature, a battery voltage value and/or a battery SOC value in the historical charging data as the historical charging parameters, and taking the historical charging parameters as input samples; extracting battery cell electrode difference in the historical charging data as an output sample, wherein the acquisition time of the output sample is later than the acquisition time of the input sample; training a machine learning model according to the input sample and the output sample, and taking the trained machine learning model as a state monitoring model; the state monitoring model is used for representing the corresponding relation.
The step S202 includes: extracting a battery current value, a battery temperature, a battery voltage value and/or a battery SOC value in the battery charging parameters as input parameters to be input into the state monitoring model, and taking the output of the state monitoring model as the estimated battery core range of the battery; and taking the time interval between the collection time of the output sample and the collection time of the input sample as the set duration.
The machine learning model can be realized by adopting the existing machine learning algorithm in the prior art, and the MATLAB machine learning linear regression model is adopted in the scheme. The method specifically comprises the following steps:
(1) Acquiring historical charging data of a battery
Directly downloading BMS national standard data uploaded by an electric vehicle in operation within a preset time period from the existing electric vehicle battery operation data in the vehicle management platform, wherein:
1. and carrying out data downloading and sorting use by taking month as a unit. For example: 2018.2.01-2.28.
2. The data saving format is not limited, for example: * Xls, xlsx, csv, etc.
(2) Performing data cleansing operations on historical charging data
The data cleaning is a process of rechecking and checking the data, and aims to delete repeated information and correct errors and ensure data consistency. The source data uploaded to the vehicle management platform in the BMS system comprises information such as current, voltage and SOC value of the battery pack, on one hand, the data are huge and complex, and the data which can be used for evaluating the consistency of the battery cells in the scheme are only part of the data, so that useless data need to be removed. On the other hand, the source data has the phenomena of dislocation, repetition and null value, and the wrong or conflicted data needs to be washed out, and only the useful data needs to be reserved.
(3) Screening useful data according to the obtained data characteristics to form a new data sample
The sampling period of the voltage and the current of the battery is 10-30s, the dynamic working condition of the electric vehicle causes frequent change of the battery load current, the voltage change is large, stable and reliable data cannot be obtained, and secondly, the safety problems of overcharging, thermal runaway, lithium precipitation and the like are very easy to occur when the lithium battery is charged. The charging process can be divided into two cases of fast charging and slow charging, and the battery cell consistency prediction model can be respectively established for the data in the two types of charging states during actual selection. The steps in the scheme are also available for predicting the consistency of battery cells during slow charge using slow charge data. In this scheme, three data samples are created, as shown in fig. 3, sample data a, sample data B, and sample data C. The sample data is successfully collected in sequence that the sample data A is earlier than the sample data B, and the sample data B is earlier than the sample data C. In this example, the sample data a is a two month sample before march, the sample data B is a last month sample before march, and the sample data C is a last week sample. Referring to fig. 3, the battery current I, the SOC value, the temperature T, and the voltage value V in the sample data are extracted as characteristic parameters for use as independent variable inputs in the machine learning model, and the machine learning model is trained with the battery cell difference as an output of the machine learning model. Firstly, training a machine learning model by using sample data A, checking the trained machine learning model by using sample data B, if the checking is passed, directly estimating the current battery cell range by using sample data C, otherwise, training and checking the machine learning model by using sample data A and sample data B again. In the above scheme, the verification process for the linear regression model may be as follows:
step A1, data grouping step: dividing the historical charging data into two groups, wherein the data volume of the first group of historical charging data is larger than that of the second group of historical charging data; i.e. the data as training verification samples are directly divided into groups a and B.
Step A2, model training step: extracting an input sample and an output sample in a first group of historical charging data, and training the machine learning model to obtain an initial state monitoring model; and training the machine learning model by using the data of the group A to obtain an initial state monitoring model.
Step A3, a model verification step: and extracting an input sample and an output sample in the second group of historical charging data to verify the initial state monitoring model, if the initial state monitoring model passes the verification, taking the initial state monitoring model as the state monitoring model, otherwise, increasing the data quantity of the first group of historical charging data, and returning to the data grouping step. And verifying the initial state model by using the B group data, and obtaining a state monitoring model after verification.
As another approach, the machine learning model may be trained and validated by:
step B1, data grouping step: dividing the historical charging data into K groups, taking one group of historical charging data as test sample data, and taking the rest groups of historical charging data as training sample data;
step B2, model training step: extracting an input sample and an output sample in training sample data to train the machine learning model to obtain an initial state monitoring model;
step B3, a model verification step: extracting an input sample and an output sample in test sample data to verify the initial state monitoring model, and taking the initial state monitoring model as an alternative state monitoring model if the initial state monitoring model passes the verification;
step B4, a cross verification step: taking the other set of historical charging data as test sample data, taking the other set of historical charging data as training sample data, and repeating the model training step and the model verification step until all K sets of historical charging data are selected as test sample data;
step B5, a model determining step: and determining the state monitoring model according to all the alternative state monitoring models.
In the scheme, a cross-validation mode is preferred, and the parameter frames of all models are optimized in the cross-validation method. The reliability of the algorithm depends on the parameter framework, that is, which battery data is most efficient for the results produced. The lithium battery is used as a system for providing electric energy by electrochemical reaction, and the working process of the lithium battery relates to a plurality of physical and chemical processes such as porous solid phase diffusion, ion migration in liquid, surface electrochemical reaction, solid conduction and the like, so that the comprehensive performance of the lithium battery is obviously influenced by factors such as battery design, working temperature, working current, SOC value and the like. For a specific vehicle, the distribution of the cell voltage of the lithium battery is affected by a plurality of factors, mainly including total current, total voltage, cell temperature, SOC and the like. Using small amounts of data that are related to model target values increases the accuracy and learning efficiency of the learning model. In this embodiment, to improve the quality of the parameter framework, the original data is first randomly divided into K parts. Among the K parts, one part is selected as test data, and the rest K-1 parts are used as training data to obtain corresponding experimental results. Then, another part is selected as test data, the remaining K-1 parts are used as training data, and the like, and the K times of cross-checking are repeated. And selecting a different part from the K parts as test data in each experiment, ensuring that the data of the K parts are respectively tested, and taking the rest K-1 parts as training data for performing the experiment.
In the step B5, the alternative state monitoring model with the highest verification result accuracy is used as the state monitoring model; or taking the average value of the adjustment coefficients in all the alternative state monitoring models as an actual adjustment coefficient to obtain the state monitoring model. That is, the K obtained experimental results are preferably averaged, and the experimental results may be the difference between the predicted value and the check value, so that the smaller the difference is, the better the difference is, and thus the best classification is determined, and the training of the model is realized. Referring to fig. 3, in the scheme, regression Learner APP built in MATLAB is used to train and predict data, and K takes a value of 4. The regression effect is evaluated using the determinant coefficients (Rs). The determinable coefficient refers to the specific gravity of the sum of squares regression (ESS-explained sum of squares) in the total variation (TSS-total sum of squares), and the calculation formula is as follows:
wherein y is i ' to predict the cell identity result, y i Is the actual cell consistency result.
Referring also to fig. 3, using data sample a as the data input sample, using K-time cross-check training and verification models, calculating Rs1 value, the training sample may have poor regression effect due to small data volume, i.e. Rs1 value cannot reach the required value (denoted as rs_std), in this case using rs_std=0.99, a longer time of input of the recorded training sample is needed to train the model, i.e. the flow returns to the beginning of "data input"; if the Rs1 value passes verification, continuing to proceed downwards: using sample B as input, calculating Rs2, judging whether Rs2 passes verification (can not reach the requirement Rs_std), if Rs2 fails verification, re-returning to the beginning of data input, re-increasing the data input quantity of data A, and re-executing the flow; if the Rs2 value passes the verification, the next step is continued to obtain a finally available regression model. In this example, one month of vehicle operation history data is used as a sample A to train the sample, one month of data is used as a sample B to verify, rs1 is found to reach the standard, rs2 is found to be not reach the standard, and two months of vehicle operation history data is used as a sample A to train the model, and both Rs1 and Rs2 reach the standard. Referring to fig. 4, if the relationship between the SOC value and the battery cell electrode difference value is used as a consistency criterion, the error values between the predicted result and the actual result by using the machine learning model in the scheme are all within ±0.3%, and the accuracy is very high. In the scheme, as long as the average absolute value Error (ERR) is within 0.90%, the average absolute value Error (ERR) can be considered to be smaller than a preset cell consistency parameter threshold value, and the model training result is indicated to be available.
Further, the battery state monitoring method in the above aspect determines whether the battery is in a safe state during charging by:
the method comprises the steps of obtaining the real-time cell consistency parameter of the battery according to the following mode:
wherein V is gap_1 To the real-time battery cell is extremely poor, V gap_0 A battery cell electrode difference threshold value is preset; if R_V > R_V_0, judging that the battery is in an unsafe state, wherein R_V_0 is a preset battery cell consistency parameter threshold value.
In the step, battery operation data of a vehicle to be predicted is input into a trained state monitoring model, and the state of the consistency of the battery cells of the battery is predicted by using the model, so that the safety state of the battery is monitored. Sample C (i.e. data before one week) can be used as an input sample to calculate the voltage range of the current vehicle battery, and then calculate the cell consistency parameter of the battery, if R_V is greater than R_V_0, the vehicle charging behavior is good, and the safety risk is avoided; if R_V is less than or equal to R_V_0, reporting that the vehicle is at safety risk, and sending a safety warning or carrying out later important monitoring. In this example, the r_v_0 takes a value of 0.9, and the calculation result of r_v is 0.98, so the conclusion is: the battery of the electric vehicle is in a safe state in this example.
The scheme provided by the embodiment provides a method for training and optimizing the model by using the machine learning linear regression model and using the historical operation data of the battery, and the battery safety state can be accurately predicted by using the trained model, so that the purpose of monitoring the battery safety state is achieved. Therefore, the safety state of the vehicle is supervised on the vehicle management platform, the accident rate of the electric vehicle is reduced, and the driver and passenger safety is guaranteed.
Example 2
The present embodiment also provides a computer-readable storage medium having stored therein program instructions, and after the computer reads the program instructions, executing the battery state monitoring method according to any of the aspects of embodiment 1.
Example 4
Fig. 5 is a schematic hardware structure diagram of a battery charging process state monitoring system according to the present embodiment, where the apparatus includes: one or more processors 501 and a memory 502, one processor 501 being illustrated in fig. 5. The apparatus for performing the battery state monitoring method may further include: an input device 503 and an output device 504. The processor 501, memory 502, input devices 503 and output devices 504 may be connected by a bus or otherwise, for example in fig. 5.
The memory 502, as a non-volatile computer readable storage medium, may be used to store non-volatile software programs, non-volatile computer executable programs, and modules. The processor 501 executes various functional applications of the server and data processing, i.e., implements the battery state monitoring method of the above-described method embodiment, by running nonvolatile software programs, instructions, and modules stored in the memory 502.
The product can execute the method provided by the embodiment of the application, and has the corresponding functional modules and beneficial effects of the execution method. Technical details not described in detail in this embodiment may be found in the methods provided in the embodiments of the present application.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (9)
1. A battery condition monitoring method, comprising the steps of:
the method for obtaining the corresponding relation between the historical operating parameter of the battery and the battery core range, wherein the historical operating parameter is a historical charging parameter when the battery is in a charging process in a historical time period, and the step of obtaining the corresponding relation between the historical charging parameter and the battery core range according to the historical charging data of the battery comprises the following steps:
extracting a battery current value, a battery temperature, a battery voltage value and/or a battery SOC value in the historical charging data as the historical charging parameters, and taking the historical charging parameters as input samples;
extracting battery cell electrode difference in the historical charging data as an output sample, wherein the acquisition time of the output sample is later than the acquisition time of the input sample;
training a machine learning model according to the input sample and the output sample, and taking the trained machine learning model as a state monitoring model; the state monitoring model is used for representing the corresponding relation;
acquiring the operation parameters of the battery in a preset time period, and determining the estimated battery core range of the battery at the current moment according to the operation parameters of the battery and the corresponding relation; the interval between the preset time period and the current time is smaller than a set duration, and the running parameter of the battery in the preset time period is a charging parameter when the battery is in a charging process in the preset time period; the step of determining the estimated battery core range of the battery at the current moment according to the corresponding relation between the charging parameter of the battery and the corresponding relation comprises the following steps:
extracting a battery current value, a battery temperature, a battery voltage value and/or a battery SOC value in the battery charging parameters as input parameters to be input into the state monitoring model, and taking the output of the state monitoring model as the estimated battery core range of the battery;
taking the time interval between the collection time of the output sample and the collection time of the input sample as the set duration;
determining whether the battery is in a safe state according to the estimated battery core range of the battery;
the battery cell electrode difference is the difference between the maximum battery cell voltage value and the minimum battery cell voltage value in the battery.
2. The battery state monitoring method according to claim 1, wherein the step of obtaining the correspondence between the historical charging parameter and the battery cell range from the historical charging data of the battery includes:
data grouping: dividing the historical charging data into two groups, wherein the data volume of the first group of historical charging data is larger than that of the second group of historical charging data;
model training: extracting an input sample and an output sample in a first group of historical charging data, and training the machine learning model to obtain an initial state monitoring model;
and (3) model verification: and extracting an input sample and an output sample in the second set of historical charging data to verify the initial state monitoring model, if the initial state monitoring model passes the verification, taking the initial state monitoring model as the state monitoring model, otherwise, increasing the data volume of the first set of historical charging data, and returning to the model training step.
3. The battery state monitoring method according to claim 1, wherein the step of obtaining the correspondence between the historical charging parameter and the battery cell range from the historical charging data of the battery includes:
data grouping: dividing the historical charging data into K groups, taking one group of historical charging data as test sample data, and taking the rest groups of historical charging data as training sample data;
model training: extracting an input sample and an output sample in training sample data to train the machine learning model to obtain an initial state monitoring model;
and (3) model verification: extracting an input sample and an output sample in test sample data to verify the initial state monitoring model, and taking the initial state monitoring model as an alternative state monitoring model if the initial state monitoring model passes the verification;
and a cross verification step: taking the other set of historical charging data as test sample data, taking the rest set of historical charging data as training sample data, and repeating the model training step and the model verification step until all K sets of historical charging data have been selected as test sample data;
model determining: and determining the state monitoring model according to all the alternative state monitoring models.
4. The battery state monitoring method according to claim 3, wherein in the model determining step:
taking the alternative state monitoring model with the highest verification result accuracy as a state monitoring model; or taking the average value of the adjustment coefficients in all the alternative state monitoring models as an actual adjustment coefficient to obtain the state monitoring model.
5. The battery state monitoring method according to any one of claims 1 to 4, wherein the step of determining whether the battery is in a safe state based on the estimated battery cell level difference of the battery includes:
the cell consistency parameter of the battery is obtained according to the following mode:
wherein (1)>
For the estimated battery cell range, +.>
A battery cell electrode difference threshold value is preset;
if R_V > R_V_0, judging that the battery is in an unsafe state, wherein R_V_0 is a preset battery cell consistency parameter threshold value.
6. The battery state monitoring method according to any one of claims 1 to 4, wherein the step of obtaining the correspondence between the historical charging parameter and the battery cell range from the historical charging data of the battery further comprises:
acquiring a historical state data set of the battery in a historical time period, wherein each historical state data in the historical state data set is configured with a state identifier;
and if the state identifier represents a charging state, taking the historical state data as the historical charging data.
7. The battery state monitoring method of claim 6, wherein the step of obtaining a set of historical state data for the battery over a historical period of time, each of the historical state data in the set of historical state data configured with a state identifier further comprises:
and screening out repeated historical state data, error historical state data and incomplete historical state data in the historical state data set.
8. A computer-readable storage medium, wherein program instructions are stored in the storage medium, and a computer executes the battery state monitoring method according to any one of claims 1 to 7 after reading the program instructions.
9. A battery state of charge monitoring system comprising at least one processor and at least one memory, at least one of said memories having program instructions stored therein, at least one of said processors executing the battery state monitoring method of any of claims 1-7 after reading said program instructions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911359532.3A CN111812518B (en) | 2019-12-25 | 2019-12-25 | Battery state monitoring method, storage medium and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911359532.3A CN111812518B (en) | 2019-12-25 | 2019-12-25 | Battery state monitoring method, storage medium and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111812518A CN111812518A (en) | 2020-10-23 |
| CN111812518B true CN111812518B (en) | 2023-04-28 |
Family
ID=72844508
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911359532.3A Active CN111812518B (en) | 2019-12-25 | 2019-12-25 | Battery state monitoring method, storage medium and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111812518B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113022378B (en) * | 2021-03-01 | 2023-03-07 | 中国第一汽车股份有限公司 | Temperature consistency prediction method, temperature consistency prediction device, prediction equipment and storage medium |
| CN113884961B (en) * | 2021-09-23 | 2023-09-19 | 中国第一汽车股份有限公司 | SOC calibration method, modeling device, computer equipment and medium |
| CN113771632A (en) * | 2021-09-28 | 2021-12-10 | 南京禄协信息科技有限公司 | High-performance vehicle-mounted electric drive real-time monitoring and control system |
| CN114325405A (en) * | 2021-12-31 | 2022-04-12 | 中国第一汽车股份有限公司 | Battery pack consistency analysis method, modeling method, device, equipment and medium |
| CN114683964B (en) * | 2022-03-29 | 2024-04-26 | 北京芯虹科技有限责任公司 | Battery state information determining method and charging equipment |
| CN116224091B (en) * | 2022-12-01 | 2024-02-02 | 伏瓦科技(苏州)有限公司 | Battery cell fault detection method and device, electronic equipment and storage medium |
| 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 |
| CN117755150B (en) * | 2023-11-28 | 2024-08-02 | 北京潞电电力建设有限公司 | Quick charging method based on machine learning |
| CN118249473B (en) * | 2024-05-24 | 2024-08-06 | 江苏中兴派能电池有限公司 | Processing method and device based on pressure difference in battery pack |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103116135A (en) * | 2012-12-14 | 2013-05-22 | 惠州市亿能电子有限公司 | Automatic diagnosis system of battery system fault |
| CN103513190A (en) * | 2013-10-24 | 2014-01-15 | 国家电网公司 | Method and system for testing consistency of performances of batteries |
| US20140079969A1 (en) * | 2012-09-19 | 2014-03-20 | Apple Inc. | Estimating state of charge (soc) and uncertainty from relaxing voltage measurements in a battery |
| EP2770605A1 (en) * | 2013-02-25 | 2014-08-27 | RP-Technik e. K. | Power supply device with battery monitoring and battery monitoring method |
| CN106033113A (en) * | 2015-03-19 | 2016-10-19 | 国家电网公司 | Health state evaluation method for energy-storage battery pack |
| CN106154165A (en) * | 2015-03-27 | 2016-11-23 | 国家电网公司 | The appraisal procedure of a kind of high capacity cell energy-storage system performance and assessment system |
| US20170054317A1 (en) * | 2015-08-20 | 2017-02-23 | International Business Machines Corporation | Modeling a change in battery degradation |
| CN106556802A (en) * | 2016-11-01 | 2017-04-05 | 东软集团股份有限公司 | A kind of accumulator battery exception cell recognition methodss and device |
| CN108983103A (en) * | 2018-06-29 | 2018-12-11 | 上海科列新能源技术有限公司 | A kind of data processing method and device of power battery |
| CN109116242A (en) * | 2018-06-29 | 2019-01-01 | 上海科列新能源技术有限公司 | A kind of data processing method and device of power battery |
| WO2019042410A1 (en) * | 2017-08-31 | 2019-03-07 | 比亚迪股份有限公司 | Battery equalization method and system, vehicle, storage medium, and electronic device |
| CN109799463A (en) * | 2019-01-18 | 2019-05-24 | 上海卡鲁自动化科技有限公司 | The estimation and prediction technique of power battery SOC/SOH/SOP under actual operating mode based on big data method |
| CN110045298A (en) * | 2019-05-06 | 2019-07-23 | 重庆大学 | A kind of diagnostic method of power battery pack parameter inconsistency |
| CN110069810A (en) * | 2019-03-11 | 2019-07-30 | 北京百度网讯科技有限公司 | Battery failures prediction technique, device, equipment and readable storage medium storing program for executing |
-
2019
- 2019-12-25 CN CN201911359532.3A patent/CN111812518B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140079969A1 (en) * | 2012-09-19 | 2014-03-20 | Apple Inc. | Estimating state of charge (soc) and uncertainty from relaxing voltage measurements in a battery |
| CN103116135A (en) * | 2012-12-14 | 2013-05-22 | 惠州市亿能电子有限公司 | Automatic diagnosis system of battery system fault |
| EP2770605A1 (en) * | 2013-02-25 | 2014-08-27 | RP-Technik e. K. | Power supply device with battery monitoring and battery monitoring method |
| CN103513190A (en) * | 2013-10-24 | 2014-01-15 | 国家电网公司 | Method and system for testing consistency of performances of batteries |
| CN106033113A (en) * | 2015-03-19 | 2016-10-19 | 国家电网公司 | Health state evaluation method for energy-storage battery pack |
| CN106154165A (en) * | 2015-03-27 | 2016-11-23 | 国家电网公司 | The appraisal procedure of a kind of high capacity cell energy-storage system performance and assessment system |
| US20170054317A1 (en) * | 2015-08-20 | 2017-02-23 | International Business Machines Corporation | Modeling a change in battery degradation |
| CN106556802A (en) * | 2016-11-01 | 2017-04-05 | 东软集团股份有限公司 | A kind of accumulator battery exception cell recognition methodss and device |
| WO2019042410A1 (en) * | 2017-08-31 | 2019-03-07 | 比亚迪股份有限公司 | Battery equalization method and system, vehicle, storage medium, and electronic device |
| CN108983103A (en) * | 2018-06-29 | 2018-12-11 | 上海科列新能源技术有限公司 | A kind of data processing method and device of power battery |
| CN109116242A (en) * | 2018-06-29 | 2019-01-01 | 上海科列新能源技术有限公司 | A kind of data processing method and device of power battery |
| CN109799463A (en) * | 2019-01-18 | 2019-05-24 | 上海卡鲁自动化科技有限公司 | The estimation and prediction technique of power battery SOC/SOH/SOP under actual operating mode based on big data method |
| CN110069810A (en) * | 2019-03-11 | 2019-07-30 | 北京百度网讯科技有限公司 | Battery failures prediction technique, device, equipment and readable storage medium storing program for executing |
| CN110045298A (en) * | 2019-05-06 | 2019-07-23 | 重庆大学 | A kind of diagnostic method of power battery pack parameter inconsistency |
Non-Patent Citations (2)
| Title |
|---|
| 基于ARIMA模型的动力电池不一致性预警研究;宋超等;《控制与信息技术》;20191005;第2019卷(第5期);全文 * |
| 基于运行数据融合的储能电池一致性评估方法研究;贾学翠等;《供用电》;20170405;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111812518A (en) | 2020-10-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111812518B (en) | Battery state monitoring method, storage medium and system | |
| WO2021169486A1 (en) | Method, system and apparatus for monitoring battery impedance abnormality on basis of charging process | |
| CN108544925B (en) | Battery management system | |
| EP4083643A1 (en) | Soh test method and apparatus | |
| CN107870306A (en) | A kind of lithium battery charge state prediction algorithm based under deep neural network | |
| CN109116242B (en) | Data processing method and device for power battery | |
| CN108037463A (en) | A kind of lithium ion battery life-span prediction method | |
| CN103698713A (en) | Method for assessing SOH (state of health) of lithium ion battery | |
| CN108205114B (en) | Method and system for predicting service life of battery | |
| CN114675188A (en) | Battery health state information determination method and device and battery system | |
| CN109031138A (en) | A kind of safety evaluation method and device of power battery | |
| EP3988952A1 (en) | Method for detecting abnormal battery cell | |
| CN113657360A (en) | Lithium battery health state estimation method, device, equipment and readable storage medium | |
| US20230366937A1 (en) | Computer-implemented method for diagnosing states of a battery | |
| WO2021052540A1 (en) | Condition value for rechargeable batteries | |
| Zhang et al. | An application‐oriented multistate estimation framework of lithium‐ion battery used in electric vehicles | |
| CN115616425A (en) | Battery performance analysis method, electronic equipment and energy storage system | |
| CN115219932A (en) | Method and device for evaluating the relative aging state of a battery of a device | |
| CN111308352A (en) | Method for estimating battery attenuation of lithium ions | |
| Cai et al. | D-ukf based state of health estimation for 18650 type lithium battery | |
| CN115469226B (en) | Real-time safety early warning method for power battery of electric automobile based on operation big data | |
| CN105759217A (en) | Lead-acid battery pack online fault diagnosis method based on measurable data | |
| CN116256651A (en) | Battery thermal incapacitation early warning method, system and readable storage medium | |
| CN113665436B (en) | Battery management method and device | |
| CN115951225A (en) | Battery equalization optimization capacity estimation method and device |
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 |