CN106772065A - Micro-capacitance sensor energy storage SOC estimation method and system based on least square method supporting vector machine - Google Patents
Micro-capacitance sensor energy storage SOC estimation method and system based on least square method supporting vector machine Download PDFInfo
- Publication number
- CN106772065A CN106772065A CN201611066845.6A CN201611066845A CN106772065A CN 106772065 A CN106772065 A CN 106772065A CN 201611066845 A CN201611066845 A CN 201611066845A CN 106772065 A CN106772065 A CN 106772065A
- Authority
- CN
- China
- Prior art keywords
- vector machine
- supporting vector
- square method
- soc
- method supporting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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]
- G01R31/385—Arrangements for measuring battery or accumulator variables
- G01R31/387—Determining ampere-hour charge capacity or SoC
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Secondary Cells (AREA)
Abstract
The present invention discloses a kind of micro-capacitance sensor energy storage SOC estimation method and system based on least square method supporting vector machine, first with the micro-capacitance sensor experiment porch comprising energy-storage system, measurement obtains voltage of the energy-storage system in charge and discharge process, electric current, temperature and SOC data sequences for it；The training set and test set sample of least square method supporting vector machine are obtained further according to measurement data；Then the optimal training parameter of the training set of least square method supporting vector machine is selected using cross validation method, and then obtains training pattern；Test set is recycled to be tested training pattern and evaluated result, SOC estimations are finally carried out to the energy-storage system of micro-capacitance sensor respective type using training pattern determines.The inventive method can overcome the shortcoming of traditional SOC computational methods, for the charge and discharge control of energy-storage system provides foundation, prevent electric power storage tank depth discharge and recharge, so as to improve the security of the life of storage battery and micro-capacitance sensor operation.
Description
Technical field
The present invention relates to micro-capacitance sensor technical field of energy storage, it is particularly a kind of suitable for multiple-energy-source micro-capacitance sensor based on a most young waiter in a wineshop or an inn
Multiply the micro-capacitance sensor energy storage SOC estimation method and system of SVMs.
Background technology
As new round electricity changes progressively implementing for scheme, sales market will be decontroled, and will significantly expand the development of micro-capacitance sensor
Space.Micro-capacitance sensor is usually in parallel with bulk power grid, the richness by control system, the various distributed energies of energy-storage system coordinated balance
Remaining electric power is input into bulk power grid or energy-storage system, and electric power or the electric power using storage are bought from bulk power grid during generation deficiency, when running into
Then quick off-the-line when bulk power grid breaks down, isolated power grid ensures main loads.Energy storage is relatively costly in micro-grid system,
In view of the financial cost of microgrid operation, in the case where micro-grid system safe operation is ensured, should as far as possible extend the life-span of energy storage.
Battery is the most frequently used stored energy form in current micro-grid system, and battery charge state (SOC) is energy-storage system charge and discharge control
In an important references amount, affect safe operation and the service life of energy-storage system.
Because energy storage SOC is related to several factors and with very strong non-linear, energy storage SOC real-time estimations have very big
It is difficult.At present, the measuring method of battery SOC has discharge test method, ampere-hour method, open circuit voltage method, internal resistance measurement method, linear mould
Type method and neural network etc..Wherein discharge test method and internal resistance measurement method are not applied in Practical Project；Ampere-hour method application is wide
It is general, but need to know initial SOC sizes, and accumulation over time, error can be increasing；Open circuit voltage method high precision, letter
It is single, however it is necessary that the open-circuit voltage values that can be just stablized after standing the long period；Internal resistance measurement method only considers putting for battery
Two basic factors of electric current and internal resistance, computational accuracy is limited, it is difficult to which accumulator capacity is accurately estimated；Linear mould
Type method is applied to the gradual situation of low current, SOC, and the estimation effect of time-dependent current situation will be studied further；Neural network base
In empirical risk minimization, local minimum is absorbed in sometimes, structural parameters are difficult to determine, lack appropriate theoretical direction.
The content of the invention
The technical problem to be solved in the present invention is：Using least square method supporting vector machine, by the rule for excavating given data
Rule, realizes determining the estimation of unknown energy storage SOC, reduces calculation error, while life-span and the micro-capacitance sensor fortune of battery can be improved
Capable security.
The technical scheme that the present invention takes is specially：Micro-capacitance sensor energy storage SOC estimations based on least square method supporting vector machine
Method, comprises the following steps：
The experimental data that energy-storage system is produced in charge and discharge process is obtained using micro-capacitance sensor energy storage experiment porch；
Based on the experimental data for obtaining, the training set sample and test set of least square method supporting vector machine training pattern are obtained
Sample；
Based on the training set sample, the optimal training parameter of least square method supporting vector machine is obtained；
Based on the optimized parameter for obtaining, least square method supporting vector machine is trained using training set sample, obtained most
A young waiter in a wineshop or an inn multiplies the training pattern of SVMs；
The training pattern for obtaining is tested using test set sample, evaluate is using above-mentioned training pattern calculating SOC
It is no effective；
If training pattern test is effective, using the least square method supporting vector machine training pattern for obtaining to phase in micro-capacitance sensor
The SOC of type energy-storage system is answered to be calculated, to determine micro-capacitance sensor energy storage SOC.
The micro-capacitance sensor battery SOC value obtained using the inventive method estimation, can be the battery in follow-up micro-capacitance sensor
Control provides foundation, prevents electric power storage tank depth discharge and recharge, so as to improve the security of the life of storage battery and micro-capacitance sensor operation.Right
When actual micro-capacitance sensor energy storage SOC is estimated, the method that energy storage experiment porch obtains data is referred to, obtain energy storage in micro-capacitance sensor
System includes the related data of voltage, electric current, temperature in actual charge and discharge process, is trained as least square method supporting vector machine
The input quantity of model, then obtains output quantity, i.e. SOC result of calculations by least square method supporting vector machine training pattern.Energy storage
Experimental subjects energy-storage system type in experiment porch is with to be actually needed the micro-grid energy storage system type estimated consistent.
Further, in the present invention, the charge and discharge process of energy-storage system includes constant-current charge state procedure, constant-current discharge shape
State process and alternately constant current charge-discharge state procedure；The experimental data produced in charge and discharge process include it is above-mentioned it is each during respectively
The voltage of generation, electric current, temperature and SOC data.
Specifically, using micro-capacitance sensor energy storage experiment porch, obtaining experimental data of the energy-storage system in charge and discharge process, bag
Include step：
After will be filled with the static first setting duration of battery of electricity, make battery-operated in constant current electric discharge operation shape
State, with sampling period t1, sampling obtains voltage data sequence, current data sequence, temperature number of the battery under current state
According to sequence and SOC data sequences；
After by the static second setting duration of battery, make battery-operated in constant current charge running status, to sample
Cycle t2, sampling obtain voltage data sequence of the battery under current state, current data sequence, temperature data sequence and
SOC data sequences；
Will after battery is fully charged static 3rd setting duration, battery-operated alternately charged, put in constant current
The running status of electricity, with the sampling period as t3, obtains voltage data sequence, current data sequence of the battery under current state
Row, temperature data sequence and SOC data sequences.
In process of the test, above-mentioned first sets duration, the second setting duration to battery, the 3rd setting duration can be according to warp
Test or need to set, need to only cause battery condition stabilization, be prior art.Sampling period t1, t2, t3 are also as needed
Setting, such as 1s.When carrying out discharge and recharge to battery, make battery-operated in the state of constant current discharge and recharge, be such as operated in perseverance
Determine 0.25c charge operations or constant 0.25c electric discharge operations, or constant 0.25c charges, discharge alternate run.The electricity of discharge and recharge
Flow big I to set as needed, can be 0.25c, 0.3c or other numerical value.
The data acquisition experimentation that above energy-storage system is in each state can be carried out continuously, and can also carry out respectively, state
Standing between conversion contributes to the test data stablized, and then can improve the final degree of accuracy for calculating data.If respectively
Carry out, cause that energy-storage system standing a period of time similarly causes that experimental result data can more react truth before experiment.
Further, in the present invention, the data set of the training set and test set sample is：
Wherein, M is data sampling points,
The form for being write as matrix is：
In formula, x1N () is contact potential series, x2N () is current sequence, x3N () is temperature sequence；Y (n) is SOC sequences；
And there is mapping function：
Y (n)=F (X (n)) 4)
Meet the relation between battery tension, electric current, temperature and corresponding SOC data.
In the present invention, the training objective of least square method supporting vector machine is to build mapping function y=f (x), it is preferred that fixed
Benefit film showing penetrates type function for f (x)=(ω x)+b, introduces slack variable ξiThe least square of >=0 and regularization parameter γ is supported
Vector machine is described as follows：
Wherein, (xi,yi) it is training set sample data, l is training set sample data number；
The kernel function of least square method supporting vector machine chooses RBF kernel functions, and RBF kernel functions are normally defined any in space
Point x to a certain center xcBetween Euclidean distance monotonic function：
Wherein, xcIt is kernel function center, σ2It is kernel functional parameter.
Preferably, in the present invention, using n times cross validation method, the optimal training of least square method supporting vector machine is obtained
Parameter；In the n times cross validation method, N=10；Before cross validation, the regularization ginseng of least square method supporting vector machine is given
Number γ scopes are γ ∈ [γmin,γmax], kernel functional parameter σ2Scope is σ2∈[σ2 min,σ2 max], when cross validation is trained,
γ and σ2Above range in traversal value, to each parameter combination (γ, σ2) n times cross validation is carried out, help to obtain
The training pattern of the more reliable least square method supporting vector machine of SOC result of calculations.
Determine system invention additionally discloses a kind of micro-capacitance sensor energy storage based on least square method supporting vector machine, it includes：
Experimental data acquisition module, using the micro-capacitance sensor energy storage experiment porch comprising energy-storage system, obtains energy-storage system and exists
The experimental data produced in charge and discharge process；The instruction of least square method supporting vector machine training pattern is set up based on the experimental data
Practice collection sample and test set sample；
Least square method supporting vector machine training module, using the training set sample, by n times cross validation method, obtains
Obtain the optimal training parameter of least square method supporting vector machine；Based on the optimal training parameter, using the training set sample pair
Least square method supporting vector machine is trained, and obtains the training pattern of least square method supporting vector machine；
Least square method supporting vector machine test module, is trained using the test set sample to least square method supporting vector machine
Model is tested, to evaluate the validity of least square method supporting vector machine training pattern；
Micro-capacitance sensor energy storage SOC determining modules, the SOC relevant state datas of energy-storage system are used as defeated using in actual micro-capacitance sensor
Enter, using effective least square method supporting vector machine training pattern is tested, it is determined that respective type energy-storage system in actual micro-capacitance sensor
SOC；The SOC relevant state datas include voltage, electric current and the temperature data of energy-storage system.
In experimental data acquisition module, the charge and discharge process of energy-storage system includes constant-current charge state procedure, constant-current discharge
State procedure and alternately constant current charge-discharge state procedure；The experimental data produced in charge and discharge process include it is above-mentioned it is each during point
Voltage, electric current, temperature and the SOC data not produced.
In present system, voltage, electric current and the temperature data produced in energy-storage system charge and discharge process are defined, with SOC
Mapping function type between data is f (x)=(ω x)+b, introduces slack variable ξi>=0 and the minimum of regularization parameter γ
Two multiply being described as follows for SVMs：
Wherein, (xi,yi) it is training set sample data, l is training set sample data number；
The kernel function of least square method supporting vector machine chooses RBF kernel functions, and RBF kernel functions are defined as any point x in space
To a certain center xcBetween Euclidean distance monotonic function：
Wherein, xcIt is kernel function center, σ2It is kernel functional parameter.
In least square method supporting vector machine training module, using n times cross validation method, least square supporting vector is obtained
The optimal training parameter of machine；In the n times cross validation method, N=10；Before cross validation, least square supporting vector is given
The regularization parameter γ scopes of machine are γ ∈ [γmin,γmax], kernel functional parameter σ2Scope is σ2∈[σ2 min,σ2 max], intersection is tested
During card training, in γ and σ2Above range in traversal value, to each parameter combination (γ, σ2) carry out n times intersect test
Card.
Beneficial effect
Compared with the prior art, the present invention has advantages below and progress：
(1) present invention can take into full account the factor related to energy storage SOC, be put down by building the experiment comprising energy-storage system
Platform is measured to the running parameter in energy-storage system charge and discharge process, and then is excavated using least square method supporting vector machine
The inherent law of primary data, finally can effectively realize the estimation of actual micro-capacitance sensor energy storage SOC；
(2) present invention can overcome the initial SOC of ampere-hour method that prior art is used to be difficult to determine, and open circuit voltage method needs
Shortcoming static for a long time, but least square method supporting vector machine training pattern is obtained by multiple training, and by test
The training pattern that set pair is obtained carries out efficiency evaluation, to ensure reliability when estimating actual micro-capacitance sensor energy storage SOC, does not deposit
In cumulative errors；
(3) actual micro-grid energy storage system SOC value is obtained using present invention estimation, can is that the charge and discharge of energy-storage system is automatically controlled
System provides foundation, prevents battery from carrying out depth discharge and recharge, so as to improve the security of the life of storage battery and micro-capacitance sensor operation.
Brief description of the drawings
Fig. 1 is the inventive method schematic flow sheet.
Specific embodiment
The present invention is further described with specific embodiment below in conjunction with the accompanying drawings.
With reference to Fig. 1, the present invention is based on the micro-capacitance sensor energy storage SOC estimation method of least square method supporting vector machine, including following
Step：
The micro-capacitance sensor energy storage experiment porch comprising energy-storage system is built, obtaining energy-storage system using energy storage experiment porch is filling
The experimental data produced in discharge process, experimental data includes voltage, electric current, temperature and SOC；
Based on the experimental data for obtaining, the training set sample and test set of least square method supporting vector machine training pattern are obtained
Sample；
Based on the training set sample, the optimal training parameter of least square method supporting vector machine is obtained；
Based on the optimized parameter for obtaining, least square method supporting vector machine is trained using training set sample, obtained most
A young waiter in a wineshop or an inn multiplies the training pattern of SVMs；
The training pattern for obtaining is tested using test set sample, evaluate is using above-mentioned training pattern calculating SOC
It is no effective；
If training pattern test is effective, using the foregoing least square method supporting vector machine training pattern for obtaining to micro-capacitance sensor
The SOC of middle respective type energy-storage system is calculated, to determine micro-capacitance sensor energy storage SOC.
The micro-capacitance sensor battery SOC value obtained using the inventive method estimation, can be the battery in follow-up micro-capacitance sensor
Control provides foundation, prevents electric power storage tank depth discharge and recharge, so as to improve the security of the life of storage battery and micro-capacitance sensor operation.Right
When actual micro-capacitance sensor energy storage SOC is estimated, the method that energy storage experiment porch obtains data is referred to, obtain energy storage in micro-capacitance sensor
System includes the related data of voltage, electric current, temperature in actual charge and discharge process, is trained as least square method supporting vector machine
The input quantity of model, then obtains output quantity, i.e. SOC result of calculations by least square method supporting vector machine training pattern.
Embodiment
The micro-grid energy storage system SOC estimation method of the present embodiment is concretely comprised the following steps：
(1) the micro-capacitance sensor experiment porch comprising energy-storage system is built：
The micro-grid energy storage system type that can be calculated according to actual needs carries out building for experiment porch.In current micro-capacitance sensor
Using it is most be lead-acid accumulator, rated voltage 12V is selected in the present embodiment, battery capacity is made for the lead-acid accumulator of 50Ah
It is experimental subjects.The voltage in battery running, electric current and temperature signal are collected and stored using data acquisition equipment, and
Actual numerical value is reduced within a processor.Being obtained for the measurement of energy-storage system experimental data can use prior art.
(2) measured respectively using experiment porch obtain voltage of the energy-storage system in charge and discharge process, electric current, temperature and
SOC, concretely comprises the following steps：
(2-1) will be filled with battery static a period of time of electricity, then make battery-operated in constant 0.25C electric discharge operation shapes
State, the sampling period is 1s, obtains battery contact potential series in this case, current sequence, temperature sequence and SOC sequences；
(2-2) makes battery static a period of time of completion (2-1) step, then battery-operated is filled in constant 0.25C
Electric running status, the sampling period is 1s, obtains battery contact potential series in this case, current sequence, temperature sequence and SOC
Sequence；
(2-3) will be filled with electricity battery static a period of time, then make battery-operated constant 0.25C alternately charge,
The running status of electric discharge, the sampling period is 1s, obtains battery contact potential series, current sequence, temperature sequence in this case
And SOC sequences；
(3) according to voltage, electric current, temperature and SOC sequences, obtain training set in least square method supporting vector machine model and
Test set sample.
Based on above-mentioned steps, the data set for obtaining is：
Wherein,
The form for being write as matrix is：
In formula, x1N () is contact potential series, x2N () is current sequence, x3N () is temperature sequence；Y (n) is SOC sequences.
According to acquired battery tension, electric current and the corresponding SOC data of temperature, a mapping function can be found,
So that
Y (n)=F (X (n)) 4)
Define voltage, the mapping between electric current and temperature data, with SOC data produced in energy-storage system charge and discharge process
Type function is f (x)=(ω x)+b, introduces slack variable ξi>=0 and the least square method supporting vector machine of regularization parameter γ
Be described as follows：
Wherein, (xi,yi) it is training set sample data, l is training set sample data number；
Because SOC and battery tension, electric current and temperature exist very strong non-linear, for nonlinear regression problem, need
Kernel-Based Methods are introduced to solve.Select RBF kernel functions in the present invention, RBF kernel functions be defined as in space any point x to certain
One center xcBetween Euclidean distance monotonic function：Such as formula 5) shown in.Therefore have two in least square method supporting vector machine forecast model
Individual parameter needs selection, i.e. regularization parameter γ and kernel functional parameter σ2。
Wherein, xcIt is kernel function center, σ2It is kernel functional parameter.
(4) training parameter of the training set of least square method supporting vector machine is selected using n times cross validation method：
Training set is divided into N number of subset, N is generally higher than equal to 2, one of subset is kept as verifying model
Data, other N-1 subset are used for training.Cross validation repeats n times, and each subset is verified once.Before training, canonical is given
Change parameter γ and kernel functional parameter σ2Scope, i.e. γ ∈ [γmin,γmax], σ2∈[σ2 min,σ2 max], make γ and σ2In this model
Interior traversal value is enclosed, n times cross validation is carried out to each combination parameter.Choose mean square error MSE (Mean Square
Error) as evaluation index, the MSE of n times test result is averaged index corresponding as parameter, last selective goal
One group of parameter combination of highest (i.e. MSE is minimum) is used as final optimized parameter.
In the present embodiment, γ ∈ [1,10000], σ are chosen2∈ [0.01,10000], N=10.Equivalent to having carried out n times instruction
Practice and n times test, and then obtain required best parameter group (γ, σ2)。
(5) using the sample data of training set, by the optimal training parameter that has selected to least square method supporting vector machine
It is trained, and then obtains training pattern.
(6) training pattern for obtaining is tested using test set sample, calculates SOC's to evaluate above-mentioned training pattern
Validity.
In this step, the SOC data that will be obtained using training pattern are entered with the actual SOC data obtained using experiment porch
Row compares, and the size according to mean square error and relative error judges to calculate SOC using least square method supporting vector machine training pattern
Method validity.
If the SOC for calculating is small with the SOC errors of reality, prove that the method calculates SOC effective, reliable；If
Error is very big then to prove that the foregoing least square method supporting vector machine for obtaining is unreliable inapplicable.
(7) if the least square method supporting vector machine training pattern obtained by verifying afterwards after tested is reliable effectively, using preceding
The least square method supporting vector machine training pattern for obtaining is stated to count the SOC of respective type energy-storage system in actual micro-capacitance sensor
Calculate.
When actual calculating is carried out, the method that experimental data is obtained with reference to energy storage experiment porch obtains energy storage in micro-capacitance sensor
System includes the related data of voltage, electric current, temperature in charge and discharge process, used as least square method supporting vector machine training pattern
Input quantity, output quantity, i.e. SOC result of calculations are then obtained by least square method supporting vector machine training pattern.
Embodiment
The present embodiment is that the micro-capacitance sensor energy storage based on least square method supporting vector machine determines system, and it includes：
Experimental data acquisition module, using the micro-capacitance sensor energy storage experiment porch comprising energy-storage system, obtains energy-storage system and exists
The experimental data produced in charge and discharge process, the experimental data includes voltage, electric current, temperature and SOC；Based on the experiment number
According to the training set sample and test set sample of setting up least square method supporting vector machine training pattern；
Least square method supporting vector machine training module, chooses the kernel function of least square method supporting vector machine training pattern, profit
With the training set sample, the optimal training parameter of least square method supporting vector machine is obtained；Based on the optimal training parameter, profit
Least square method supporting vector machine is trained with the training set sample, obtains the training mould of least square method supporting vector machine
Type；
Least square method supporting vector machine test module, is trained using the test set sample to least square method supporting vector machine
Model is tested, to evaluate the validity of least square method supporting vector machine training pattern；
Micro-capacitance sensor energy storage SOC determining modules, the SOC relevant state datas of energy-storage system are used as defeated using in actual micro-capacitance sensor
Enter, using effective least square method supporting vector machine training pattern is tested, it is determined that respective type energy-storage system in actual micro-capacitance sensor
SOC；The SOC relevant state datas include voltage, electric current and the temperature data of energy-storage system.
In experimental data acquisition module, the charge and discharge process of energy-storage system includes constant-current charge state procedure, constant-current discharge
State procedure and alternately constant current charge-discharge state procedure；The experimental data produced in charge and discharge process include it is above-mentioned it is each during point
Voltage, electric current, temperature and the SOC data not produced.
In the present embodiment, voltage, electric current and the temperature data produced in energy-storage system charge and discharge process are defined, with SOC numbers
Mapping function type between is f (x)=(ω x)+b, introduces slack variable ξiA most young waiter in a wineshop or an inn of >=0 and regularization parameter γ
Multiply being described as follows for SVMs：
Wherein, (xi,yi) it is training set sample data, l is training set sample data number；
The kernel function of least square method supporting vector machine chooses RBF kernel functions, and RBF kernel functions are defined as any point x in space
To a certain center xcBetween Euclidean distance monotonic function：
Wherein, xcIt is kernel function center, σ2It is kernel functional parameter.
In least square method supporting vector machine training module, using n times cross validation method, least square supporting vector is obtained
The optimal training parameter of machine；In the n times cross validation method, N=10；Before cross validation, least square supporting vector is given
The regularization parameter γ scopes of machine are γ ∈ [γmin,γmax], kernel functional parameter σ2Scope is σ2∈[σ2 min,σ2 max], intersection is tested
During card training, in γ and σ2Above range in traversal value, to each parameter combination (γ, σ2) carry out n times intersect test
Card.
The above is only the preferred embodiment of the present invention, it should be pointed out that：For the ordinary skill people of the art
For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
It should be understood by those skilled in the art that, embodiments herein can be provided as method, system or computer program
Product.Therefore, the application can be using the reality in terms of complete hardware embodiment, complete software embodiment or combination software and hardware
Apply the form of example.And, the application can be used and wherein include the computer of computer usable program code at one or more
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) is produced
The form of product.
The application is the flow with reference to method, equipment (system) and computer program product according to the embodiment of the present application
Figure and/or block diagram are described.It should be understood that every first-class during flow chart and/or block diagram can be realized by computer program instructions
The combination of flow and/or square frame in journey and/or square frame and flow chart and/or block diagram.These computer programs can be provided
The processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that produced for reality by the instruction of computer or the computing device of other programmable data processing devices
The device of the function of being specified in present one flow of flow chart or multiple one square frame of flow and/or block diagram or multiple square frames.
These computer program instructions may be alternatively stored in can guide computer or other programmable data processing devices with spy
In determining the computer-readable memory that mode works so that instruction of the storage in the computer-readable memory is produced and include finger
Make the manufacture of device, the command device realize in one flow of flow chart or multiple one square frame of flow and/or block diagram or
The function of being specified in multiple square frames.
These computer program instructions can be also loaded into computer or other programmable data processing devices so that in meter
Series of operation steps is performed on calculation machine or other programmable devices to produce computer implemented treatment, so as in computer or
The instruction performed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in individual square frame or multiple square frames.
Claims (10)
1. a kind of micro-capacitance sensor energy storage SOC estimation method based on least square method supporting vector machine, it is characterized in that, including：
The experimental data that energy-storage system is produced in charge and discharge process is obtained using micro-capacitance sensor energy storage experiment porch；
Based on the experimental data for obtaining, the training set sample and test set sample of least square method supporting vector machine training pattern are obtained
This；
Based on the training set sample, the optimal training parameter of least square method supporting vector machine is obtained；
Based on the optimized parameter for obtaining, least square method supporting vector machine is trained using training set sample, obtains a most young waiter in a wineshop or an inn
Multiply the training pattern of SVMs；
The training pattern for obtaining is tested using test set sample, is evaluated and is calculated whether SOC has using above-mentioned training pattern
Effect；
If training pattern test is effective, using the least square method supporting vector machine training pattern for obtaining to respective class in micro-capacitance sensor
The SOC of type energy-storage system is calculated, to determine micro-capacitance sensor energy storage SOC.
2. method according to claim 1, it is characterized in that, the charge and discharge process of energy-storage system includes constant-current charge state mistake
Journey, constant-current discharge state procedure and alternating constant current charge-discharge state procedure；The experimental data produced in charge and discharge process includes upper
Voltage, electric current, temperature and the SOC data produced respectively during stating respectively.
3. method according to claim 1, it is characterized in that, experimental data of the energy-storage system in charge and discharge process is obtained,
Including step：
Will be filled with electricity battery it is static first setting duration after, make battery-operated constant current discharge running status, with
Sampling period t1, sampling obtains voltage data sequence, current data sequence, temperature data sequence of the battery under current state
And SOC data sequences；
After by the static second setting duration of battery, make battery-operated in constant current charge running status, with the sampling period
T2, sampling obtains voltage data sequence of the battery under current state, current data sequence, temperature data sequence and SOC numbers
According to sequence；
Will after battery is fully charged it is static 3rd setting duration, make battery-operated constant current alternately charge, electric discharge
Running status, with the sampling period as t3, obtains voltage data sequence, current data sequence, temperature of the battery under current state
Degrees of data sequence and SOC data sequences.
4. according to the method in claim 2 or 3, it is characterized in that, the data set of the training set and test set sampleFor：
Wherein, M is data sampling points,
The form i.e. X for being write as matrix is：
In formula, x1N () is contact potential series, x2N () is current sequence, x3N () is temperature sequence；Y (n) is SOC sequences；
And there is mapping function：
Y (n)=F (X (n)) 4)
Meet the relation between battery tension, electric current, temperature and corresponding SOC data.
5. method according to claim 4, it is characterized in that, it is f (x)=(ω x)+b to define mapping function type, is introduced
Slack variable ξiThe least square method supporting vector machine of >=0 and regularization parameter γ is described as follows：
Wherein, (xi,yi) it is training set sample data, l is training set sample data number；
The kernel function of least square method supporting vector machine chooses RBF kernel functions, RBF kernel functions be defined as in space any point x to certain
One center xcBetween Euclidean distance monotonic function：
Wherein, xcIt is kernel function center, σ2It is kernel functional parameter.
6. method according to claim 5, it is characterized in that, using n times cross validation method, obtain least square support to
The optimal training parameter of amount machine；In the n times cross validation method, N=10；Before cross validation, given least square support to
The regularization parameter γ scopes of amount machine are γ ∈ [γmin,γmax], kernel functional parameter σ2Scope is σ2∈[σ2 min,σ2 max], intersect
During checking training, in γ and σ2Above range in traversal value, to each parameter combination (γ, σ2) carry out n times intersect test
Card.
7. a kind of micro-capacitance sensor energy storage estimating system based on least square method supporting vector machine, it is characterized in that, including：
Experimental data acquisition module, using the micro-capacitance sensor energy storage experiment porch comprising energy-storage system, obtains energy-storage system in charge and discharge
The experimental data produced in electric process；The training set of least square method supporting vector machine training pattern is set up based on the experimental data
Sample and test set sample；
Least square method supporting vector machine training module, using the training set sample, obtains least square method supporting vector machine most
Excellent training parameter；Based on the optimal training parameter, least square method supporting vector machine is instructed using the training set sample
Practice, obtain the training pattern of least square method supporting vector machine；
Least square method supporting vector machine test module, using the test set sample to least square method supporting vector machine training pattern
Tested, to evaluate the validity of least square method supporting vector machine training pattern；
Micro-capacitance sensor energy storage SOC determining modules, the SOC relevant state datas of energy-storage system are sharp as input using in actual micro-capacitance sensor
With testing effective least square method supporting vector machine training pattern, it is determined that in actual micro-capacitance sensor respective type energy-storage system SOC；
The SOC relevant state datas include voltage, electric current and the temperature data of energy-storage system.
8. the micro-capacitance sensor energy storage estimating system based on least square method supporting vector machine according to claim 7, it is characterized in that,
In experimental data acquisition module, the charge and discharge process of energy-storage system includes constant-current charge state procedure, constant-current discharge state procedure
With alternately constant current charge-discharge state procedure；The experimental data produced in charge and discharge process include it is above-mentioned it is each during produce respectively
Voltage, electric current, temperature and SOC data.
9. the micro-capacitance sensor energy storage estimating system based on least square method supporting vector machine according to claim 8, it is characterized in that,
Define voltage, the mapping function type between electric current and temperature data, with SOC data produced in energy-storage system charge and discharge process
It is f (x)=(ω x)+b, introduces slack variable ξiThe description of the least square method supporting vector machine of >=0 and regularization parameter γ is such as
Under：
Wherein, (xi,yi) it is training set sample data, l is training set sample data number；
The kernel function of least square method supporting vector machine chooses RBF kernel functions, RBF kernel functions be defined as in space any point x to certain
One center xcBetween Euclidean distance monotonic function：
Wherein, xcIt is kernel function center, σ2It is kernel functional parameter.
10. method according to claim 9, it is characterized in that, in least square method supporting vector machine training module, using n times
Cross validation method, obtains the optimal training parameter of least square method supporting vector machine；
In the n times cross validation method, N=10；Before cross validation, the regularization parameter of least square method supporting vector machine is given
γ scopes are γ ∈ [γmin,γmax], kernel functional parameter σ2Scope is σ2∈[σ2 min,σ2 max], when cross validation is trained, in γ
And σ2Above range in traversal value, to each parameter combination (γ, σ2) carry out n times cross validation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611066845.6A CN106772065A (en) | 2016-11-25 | 2016-11-25 | Micro-capacitance sensor energy storage SOC estimation method and system based on least square method supporting vector machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611066845.6A CN106772065A (en) | 2016-11-25 | 2016-11-25 | Micro-capacitance sensor energy storage SOC estimation method and system based on least square method supporting vector machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106772065A true CN106772065A (en) | 2017-05-31 |
Family
ID=58904812
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611066845.6A Pending CN106772065A (en) | 2016-11-25 | 2016-11-25 | Micro-capacitance sensor energy storage SOC estimation method and system based on least square method supporting vector machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106772065A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107132490A (en) * | 2017-07-05 | 2017-09-05 | 福州大学 | A kind of method for realizing the estimation of lithium battery group state-of-charge |
| CN107290679A (en) * | 2017-07-03 | 2017-10-24 | 南京能瑞电力科技有限公司 | The Intelligentized battery method for detecting health status of charging pile is shared for electric automobile |
| CN109962515A (en) * | 2019-03-15 | 2019-07-02 | 湖南工业大学 | Ratio-support vector machines stable state estimates super capacitor charge/discharge control method |
| CN110850315A (en) * | 2019-11-29 | 2020-02-28 | 北京邮电大学 | Method and device for estimating state of charge of battery |
| CN111882155A (en) * | 2020-06-23 | 2020-11-03 | 国网河北省电力有限公司电力科学研究院 | Testing and evaluating method of distributed energy storage system |
| CN111983469A (en) * | 2020-08-24 | 2020-11-24 | 哈尔滨理工大学 | Lithium battery safety degree estimation method and device based on voltage safety boundary and temperature safety boundary |
| CN111983473A (en) * | 2020-08-24 | 2020-11-24 | 哈尔滨理工大学 | Lithium ion power battery safety degree evaluation method and device based on support vector machine |
| CN113406500A (en) * | 2021-06-29 | 2021-09-17 | 同济大学 | Method for estimating residual electric quantity of power lithium battery |
| CN113884905A (en) * | 2021-11-02 | 2022-01-04 | 山东大学 | Power battery state of charge estimation method and system based on deep learning |
-
2016
- 2016-11-25 CN CN201611066845.6A patent/CN106772065A/en active Pending
Non-Patent Citations (2)
| Title |
|---|
| 王斌 等: ""壶瓶枣轻微损伤可见/近红外光谱动态判别模型研究"", 《现代食品科技》 * |
| 高利琴: ""混合动力车用锉离子电池SOC最优预测策略研究"", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107290679A (en) * | 2017-07-03 | 2017-10-24 | 南京能瑞电力科技有限公司 | The Intelligentized battery method for detecting health status of charging pile is shared for electric automobile |
| CN107290679B (en) * | 2017-07-03 | 2019-11-12 | 南京能瑞电力科技有限公司 | The Intelligentized battery method for detecting health status of charging pile is shared for electric car |
| CN107132490A (en) * | 2017-07-05 | 2017-09-05 | 福州大学 | A kind of method for realizing the estimation of lithium battery group state-of-charge |
| CN109962515A (en) * | 2019-03-15 | 2019-07-02 | 湖南工业大学 | Ratio-support vector machines stable state estimates super capacitor charge/discharge control method |
| CN109962515B (en) * | 2019-03-15 | 2023-01-10 | 湖南工业大学 | Proportional-support vector machine steady state estimation super capacitor charging and discharging control method |
| CN110850315A (en) * | 2019-11-29 | 2020-02-28 | 北京邮电大学 | Method and device for estimating state of charge of battery |
| CN111882155B (en) * | 2020-06-23 | 2022-06-24 | 国网河北省电力有限公司电力科学研究院 | Testing and evaluating method of distributed energy storage system |
| CN111882155A (en) * | 2020-06-23 | 2020-11-03 | 国网河北省电力有限公司电力科学研究院 | Testing and evaluating method of distributed energy storage system |
| CN111983469A (en) * | 2020-08-24 | 2020-11-24 | 哈尔滨理工大学 | Lithium battery safety degree estimation method and device based on voltage safety boundary and temperature safety boundary |
| CN111983473A (en) * | 2020-08-24 | 2020-11-24 | 哈尔滨理工大学 | Lithium ion power battery safety degree evaluation method and device based on support vector machine |
| CN111983473B (en) * | 2020-08-24 | 2023-03-10 | 哈尔滨理工大学 | Lithium ion power battery safety degree evaluation method and device based on support vector machine |
| CN111983469B (en) * | 2020-08-24 | 2023-08-22 | 哈尔滨理工大学 | Lithium battery safety degree estimation method and device based on voltage safety boundary and temperature safety boundary |
| CN113406500A (en) * | 2021-06-29 | 2021-09-17 | 同济大学 | Method for estimating residual electric quantity of power lithium battery |
| CN113884905A (en) * | 2021-11-02 | 2022-01-04 | 山东大学 | Power battery state of charge estimation method and system based on deep learning |
| CN113884905B (en) * | 2021-11-02 | 2022-06-14 | 山东大学 | Power battery state of charge estimation method and system based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106772065A (en) | Micro-capacitance sensor energy storage SOC estimation method and system based on least square method supporting vector machine | |
| Ding et al. | An improved Thevenin model of lithium-ion battery with high accuracy for electric vehicles | |
| Ma et al. | State of charge and state of energy estimation for lithium-ion batteries based on a long short-term memory neural network | |
| CN103926538B (en) | Change exponent number RC equivalent-circuit model based on AIC criterion and implementation method | |
| CN105203969B (en) | State-of-charge method of estimation based on modified RC battery models | |
| CN105807226B (en) | Lithium ion battery SOC Forecasting Methodologies and device based on equivalent-circuit model | |
| Forman et al. | Genetic identification and fisher identifiability analysis of the Doyle–Fuller–Newman model from experimental cycling of a LiFePO4 cell | |
| CN106772064A (en) | A kind of health state of lithium ion battery Forecasting Methodology and device | |
| CN103632018B (en) | A kind of fuel cell modelling method based on Simscape platform | |
| Wang et al. | A novel energy management strategy for the ternary lithium batteries based on the dynamic equivalent circuit modeling and differential Kalman filtering under time-varying conditions | |
| EP4083643A1 (en) | Soh test method and apparatus | |
| Cheng et al. | An ensemble prognostic method for lithium-ion battery capacity estimation based on time-varying weight allocation | |
| CN104569844B (en) | Valve-regulated sealed lead-acid batteries health status monitoring method | |
| CN112147530B (en) | Battery state evaluation method and device | |
| CN106055775A (en) | Prediction method for life of secondary battery based on particle filter and mechanism model | |
| Mamo et al. | Long short-term memory with attention mechanism for state of charge estimation of lithium-ion batteries | |
| CN104868180B (en) | Cell method for group matching and system | |
| CN105808914A (en) | Method and device for predicting life of satellite lithium ion battery | |
| CN107632262A (en) | A kind of detection method and device of power battery pack cycle life | |
| CN104965179A (en) | Lithium ion storage battery temperature combinational circuit model and parameter identification method thereof | |
| CN108846227A (en) | A kind of capacity of lithium ion battery degradation prediction appraisal procedure based on random forest and capacity self- recoverage effect analysis | |
| Li et al. | The lithium-ion battery state-of-charge estimation using random forest regression | |
| CN109613440A (en) | Stage division, device, equipment and the storage medium of battery | |
| CN110426639A (en) | A kind of lithium ion battery life-span prediction method and system based on motional impedance spectrum | |
| Wong et al. | Li-ion batteries state-of-charge estimation using deep lstm at various battery specifications and discharge cycles |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170531 |
|
| RJ01 | Rejection of invention patent application after publication |