CN108427080A - The state-of-charge computational methods of the power battery pack of hybrid power ship - Google Patents
The state-of-charge computational methods of the power battery pack of hybrid power ship Download PDFInfo
- Publication number
- CN108427080A CN108427080A CN201810639464.5A CN201810639464A CN108427080A CN 108427080 A CN108427080 A CN 108427080A CN 201810639464 A CN201810639464 A CN 201810639464A CN 108427080 A CN108427080 A CN 108427080A
- Authority
- CN
- China
- Prior art keywords
- battery pack
- state
- power battery
- equivalent
- charge
- 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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
- Tests Of Electric Status Of Batteries (AREA)
Abstract
The present invention relates to a kind of state-of-charge computational methods of the power battery pack of hybrid power ship, comprise the steps of：S1, Thevenin equivalent-circuit models are based on, equivalent-circuit model is established for the power battery pack of hybrid power ship；S2, expanded Kalman filtration algorithm is improved using Sage Husa adaptive algorithms, include in expanded Kalman filtration algorithm process noise and measurement noise be modified；S3, using improved expanded Kalman filtration algorithm, the state-of-charge of the equivalent-circuit model of the power battery pack of hybrid power ship under charging and discharging state is estimated.The present invention is improved expanded Kalman filtration algorithm using Sage Husa adaptive algorithms, is modified to process noise and measurement noise, improves accuracy, adaptivity and the stability of expanded Kalman filtration algorithm estimation state-of-charge.
Description
Technical field
The present invention relates to a kind of state-of-charge computational methods of power battery pack, in particular to for hybrid power ship
The state-of-charge computational methods of lithium iron phosphate dynamic battery group belong to the technical field that power battery pack uses.
Background technology
Sustainable development has become the grand strategy of world today's socio-economic development, and traffic shipping industry is also gradually to low
Energy consumption, zero-emission direction are developed.The advantages of hybrid power ship existing pure electric vehicle ship low energy consumption, and have diesel engine motor vessel concurrently
The preferable advantage of oceangoing ship cruising ability, plays an important role in terms of energy-saving and emission-reduction.Therefore, domestic and international traffic shipping industry will also mix
Close the main direction of development one of of the powered ship as the following marine technology.
Power resources one of of the power battery pack as hybrid power ship have the power performance of ship very heavy
The effect wanted, state-of-charge (State of Charge, SOC) are not only to weigh the important indicator of power battery pack performance, more
It is the key factor for pushing hybrid power ship sane development.It is therefore desirable to the SOC to power battery pack to make more accurately
Estimation, in this way for improve hybrid power ship power performance, and extend power battery pack service life have it is non-
Often important meaning.
Based on above-mentioned, the present invention proposes a kind of state-of-charge computational methods of the power battery pack of hybrid power ship, with
Solve limitation existing in the prior art and demand.
Invention content
The object of the present invention is to provide a kind of state-of-charge computational methods of the power battery pack of hybrid power ship, use
Sage-Husa adaptive algorithms are improved expanded Kalman filtration algorithm, are modified to process noise and measurement noise,
Improve accuracy, adaptivity and the stability of expanded Kalman filtration algorithm estimation state-of-charge.
To achieve the above object, the present invention provides a kind of state-of-charge calculating side of the power battery pack of hybrid power ship
Method comprises the steps of：
S1, Thevenin equivalent-circuit models are based on, equivalent circuit is established for the power battery pack of hybrid power ship
Model；
S2, expanded Kalman filtration algorithm is improved using Sage-Husa adaptive algorithms, including to extending karr
Process noise and measurement noise in graceful filtering algorithm are modified；
S3, using improved expanded Kalman filtration algorithm, to the equivalent electricity of the power battery pack of hybrid power ship
State-of-charge of the road model under charging and discharging state is estimated.
In the S1, power battery pack by multiple single batteries by being connected in series or in parallel, adopt by each single battery
With Thevenin equivalent-circuit models.
When the power battery pack is composed in series by multiple single batteries, which can be equivalent to a list
Only single battery, the i.e. power battery pack can be equivalent to Thevenin equivalent-circuit models.
When the power battery pack is composed in parallel by multiple single batteries, which can be equivalent to a list
Only single battery, the i.e. power battery pack can be equivalent to Thevenin equivalent-circuit models.
In the S2, specifically comprise the steps of：
S21, to the process noise w in expanded Kalman filtration algorithm iterative processkAnd process noise covariance QkIt is repaiied
Just：
Wherein, ekNewly to cease；dkFor weighting coefficient；Kk-1Indicate the iteration coefficient at k-1 moment；Ak-1Indicate etching system when k-1
State matrix；Symbol "+" indicates correction value；Indicate the optimal estimation value of k-1 moment system modes；For system shape
State evaluated error covariance；
S22, to the measurement noise v in expanded Kalman filtration algorithm iterative processkAnd measurement noise covariance RkIt is repaiied
Just：
Wherein, Yk-1Indicate the measured value of etching system when k-1；Symbol "-" indicates discreet value；Indicate etching system when k-1
Measure discreet value；CkIndicate the corresponding state matrix of k moment observational equations.
Further, in the S21 and S22, ekAnd dkValue be respectively：
dk=(1-b)/(1-bk+1)；
Wherein, b is forgetting factor, meets 0<b<1.
In the S3, using expanded Kalman filtration algorithm, to the equivalent-circuit model of power battery pack in constant current arteries and veins
The state-of-charge for rushing charging and discharging state is estimated；Using improved expanded Kalman filtration algorithm, to power battery pack
Equivalent-circuit model is estimated in the state-of-charge of cycle charge discharge electricity condition.
In the S3, using improved expanded Kalman filtration algorithm, to the equivalent-circuit model of power battery pack
It is estimated respectively in the state-of-charge of constant-current pulse charging and discharging state and cycle charge discharge electricity condition.
In conclusion the state-of-charge computational methods of the power battery pack of hybrid power ship provided by the present invention, root
It is adaptive using Sage-Husa according to the particularity of hybrid power ship and the use environment of hybrid power ship power battery pack
It answers algorithm to be improved expanded Kalman filtration algorithm, process noise and measurement noise is modified.Pass through emulation experiment
And hybrid power ship experiment porch demonstrates the algorithm estimated accuracy, improves the adaptivity of expanded Kalman filtration algorithm
And stability, the estimation for practical ship power battery pack state-of-charge provide theoretical foundation, have Practical Project with reference to meaning
Justice.
Description of the drawings
Fig. 1 is the schematic diagram of the Thevenin equivalent-circuit models of the lithium iron phosphate dynamic battery group in the present invention；
Fig. 2 is the schematic diagram of the equivalent circuit after two single batteries series connection in the present invention；
Fig. 3 is the schematic diagram of the equivalent circuit after two single battery parallel connections in the present invention；
Fig. 4 is the experiment simulation model of the equivalent circuit of the lithium iron phosphate dynamic battery group in the present invention；
Fig. 5 and Fig. 6 is respectively SOC of the lithium iron phosphate dynamic battery group under constant-current pulse charging and discharging state in the present invention
Estimate experimental result；
Fig. 7 is that SOC of the lithium iron phosphate dynamic battery group under cycle charge discharge electricity condition in the present invention estimates experimental result；
Fig. 8 is improved SOC estimation of the lithium iron phosphate dynamic battery group under cycle charge discharge electricity condition in the present invention
Experimental result；
Fig. 9 be before improvement of the lithium iron phosphate dynamic battery group under cycle charge discharge electricity condition in the present invention with it is improved
SOC estimates the comparison diagram of experimental result；
Figure 10 is the improved expanded Kalman filtration algorithm of use in the present invention to hybrid power ship power battery
The experimental result of the SOC estimations of group；
Figure 11 is the flow chart of the state-of-charge computational methods of the power battery pack of the hybrid power ship in the present invention.
Specific implementation mode
Below in conjunction with Fig. 1~Figure 11, the preferred embodiment that the present invention will be described in detail.
As shown in figure 11, it is the state-of-charge computational methods of the power battery pack of hybrid power ship provided by the invention,
Specifically comprise the steps of：
S1, Thevenin equivalent-circuit models are based on, equivalent circuit is established for the power battery pack of hybrid power ship
Model；
S2, expanded Kalman filtration algorithm is improved using Sage-Husa adaptive algorithms, including to extending karr
Process noise and measurement noise in graceful filtering algorithm are modified；
S3, using improved expanded Kalman filtration algorithm, to the equivalent electricity of the power battery pack of hybrid power ship
State-of-charge of the road model under charging and discharging state is estimated.
Due to the rated capacity of single battery and the finiteness of power, have very when under applied to high power load operating mode
Big limitation, therefore many times need single battery forming power battery pack by way of serial or parallel connection.
By taking the experiment porch of hybrid power ship as an example, which is equipped with high-powered lithium ferric phosphate dynamic battery
Group, the power battery pack is in series by 160 LiFePO4 single batteries, and performance parameter is as shown in table 1：
Table 1, experiment porch power battery pack parameter
Power battery pack is generally connected in series or in parallel by multiple single batteries by a series of, and single battery model is built
It is vertical relatively simple, therefore the foundation of battery pack equivalent-circuit model can be regarded as the equivalent-circuit model of similar single battery,
So as to analyze the relationship between single battery and power battery pack.
Thevenin (Dai Weinan) equivalent-circuit models are comprehensive more excellent in precision, structural complexity etc..Therefore, originally
The equivalent-circuit model of single battery uses Thevenin equivalent-circuit models in invention, specific as shown in Figure 1.It is basic herein
On, it will carry out relevant deriving analysis in terms of single battery is connected with two in parallel below.
1, when two single batteries are connected, it is assumed that two single battery consistency are identical, according to circuital law and circuit
Rudimentary knowledge is it is found that all-in resistance after the series connection of two single batteries and total voltage are original twice, two single batteries warps
Cross series connection after to entire circuit structure, there is no essential influences.Therefore it may only be necessary to consider that the RC of two single batteries is in parallel
Circuit is after series connection, if it can be reduced to the shunt circuits RC of another equivalent single battery, i.e., as shown in Figure 2 two
Whether the equivalent circuit transformation after a single battery series connection is true, and analytical derivation is as detailed below.
According to electric circuit knowledge it is found that the definition of capacitance capacitive reactance is in circuit：
In the shunt circuits RC, according to parallel circuit impedance computation formula it is found that caused by after resistance is in parallel with capacitance
Total impedance Z is：
As shown in Fig. 2, total impedance Z of two shunt circuits RC (R1C1, R2C2) after series connection1For：
As shown in Fig. 2, and the total impedance Z of the single shunt circuits RC (R3C3)2For：
If two shunt circuits RC (R1C1, R2C2) in Fig. 2 can be equivalent to a shunt circuit RC after series connection
(R3C3), then there must be Z1=Z2, i.e.,：
Carrying out equation transformation to above formula can obtain：
Further it is equivalent to：
Think equal principle according to the real part of both members and imaginary part are corresponding respectively, obtains following equation equation
Group：
Wherein, it enables：
It can thus be concluded that：
Last solution obtains：
According to above-mentioned derivation it is found that when the shunt circuits RC of two single batteries are after series connection, it is equivalent to one
A new equivalent shunt circuits RC, therefore circuit overall structure can still be returned with the RC parallel connections of an individual single battery
Road indicates.And the parameter of resistance and capacitance is by resistance, the electricity in two single batteries in the new equivalent shunt circuits RC
Hold parameter to be determined.
Further, according to above-mentioned derivation it is found that when the shunt circuits RC for having multiple (at least three) single batteries pass through
After crossing series connection, it can be also equivalent to a new equivalent shunt circuit RC, therefore circuit overall structure can still use one
The shunt circuits RC of individual single battery indicate, specifically can above-mentioned two single battery is concatenated to be pushed away by repeatedly recycling
The process of leading can be realized.And the parameter of resistance and capacitance is by multiple single batteries in the new equivalent shunt circuits RC
Resistance, capacitance parameter are determined.
2, when two single battery parallel connections, it is envisaged that whether the integrated circuit structure after parallel connection can use one
A equivalent single battery indicates whether the equivalent circuit transformation after that is, as shown in Figure 3 two single battery parallel connections is true.With
Equivalent circuit after single battery directly series connection is compared, and the equivalent-simplification process of the integrated circuit structure after parallel connection is more complex.By
Conversion of circuits after above-mentioned single battery series connection derives it is found that the integrated circuit structure after series connection still can be with one newly
Equivalent single battery indicate that only the parameter of integrated circuit structure can occur to change accordingly.It is given below shown in Fig. 3
The derivation of equivalent circuit transformation after single battery parallel connection proves.
According to circuit base knowledge it is found that the voltage at parallel circuit both ends is equal.It is set by the total of R1C1 parallel circuits
Electric current is I1；Total current by R2C2 parallel circuits is I2；Entire parallel circuit both end voltage is V；R1C1 parallel branches
Total impedance is Z1；The total impedance of R2C2 parallel branches is Z2；Therefore following equation can be obtained according to circuital law：
Carrying out corresponding equation transformation can obtain：
The above-mentioned equation group of simultaneous can obtain：
V (I1-I2)=I1V1-I2V1+I1I2 (R02-R01)+I1I2 (Z2-Z1)；
Being deformed to above formula can obtain：
Further it is equivalent to：
After two single battery parallel connections, if integrated circuit structure can be equivalent to a single battery, i.e. in Fig. 3
When conversion of circuits is set up, then have：
V=V3+R03 (I1+I2)+Z3(I1+I2)；
Therefore it can obtain：
Equivalent-circuit model and electric circuit knowledge according to fig. 3 is it is found that V3, R03 and Z3It is nonnegative value, to V3, R03, Z3
Expression formula is analyzed as follows：Since single battery during by composing in parallel power battery pack accordingly, it is assumed that each
The consistency of single battery is identical, and performance indicator is also essentially identical；So each single battery in actual use, is opened
The impedance of road voltage, the internal resistance of cell and the shunt circuits RC is also essentially identical；Therefore, in two RC parallel branches (R1C1, R2C2)
Voltage V2 be approximately equal to V1.Furthermore the electric current in two RC parallel branches (R1C1, R2C2) is I1, I2 respectively, it is assumed that I1 is big
It is V since the voltage at two RC parallel branches both ends is equal, according to series-parallel circuit rule in equal to I2：Where R1C1
RC parallel branches total impedance Z1Less than the total impedance Z of the RC parallel branches where R2C22, therefore the internal resistance R02 of single battery
More than the internal resistance R01 of another single battery.It can be obtained according to above-mentioned analysis and circuital law：V3, R03 and Z3It is positive value.Together
Reason can prove to obtain when I1 is less than I2, V3, R03 and Z3Also it is positive value.
Above-mentioned derivation conversion process is pushed away according to above-mentioned analysis based on Thevenin equivalent-circuit models as shown in Figure 1
It leads it is found that the equivalent-circuit model of single battery is only there are one the shunt circuits RC, similarly it is found that when increasing the equivalent of single battery
The number (at least increasing by 1, can also increase multiple, i.e. at least two single batteries parallel connection) of the shunt circuits RC in circuit model
When, the circuit that single battery is formed after parallel connection still can be indicated with a new equivalent single battery.
In conclusion in the present invention, the single battery for constituting power battery pack uses Thevenin etc. as shown in Figure 1
Circuit model is imitated, therefore the equivalent-circuit model of the power battery pack constituted is also based on the Thevenin equivalent circuit moulds
Type.One new individual single battery can be equivalent to by multiple single batteries power battery pack in series, by multiple lists
The power battery pack that body cell parallel is constituted can also be equivalent to a new individual single battery.
Next, building the phosphoric acid based on Thevenin equivalent-circuit models using Matlab/Simulink emulation tools
The experiment simulation model (as shown in Figure 4) of the equivalent circuit of iron lithium power battery pack, and estimated using expanded Kalman filtration algorithm
Calculate its SOC.
In the experiment simulation model of the equivalent-circuit model of lithium iron phosphate dynamic battery group shown in Fig. 4, LiFePO4
The capacity of power battery pack is 20Ah, at 20 DEG C of constant temperature, simulates the constant-current pulse charge and discharge of lithium iron phosphate dynamic battery group respectively
Electric process and cycle charge discharge electric process, and SOC estimations are carried out under both states respectively.
1, the SOC estimations under constant-current pulse charging and discharging state：
In the case where the SOC initial values of lithium iron phosphate dynamic battery group have error, expanded Kalman filtration algorithm is in perseverance
Flowing has preferable convergence under pulse charging and discharging state.It is dynamic that LiFePO4 is simulated by simulating constant-current pulse charging and discharging state
The initial value of SOC is respectively set to several different situations, to verify spreading kalman by the charge and discharge process of power battery pack
Filtering algorithm estimates that the validity of SOC, experimental result difference are as shown in Figure 5 and Figure 6 under constant-current pulse charging and discharging state.
According to Fig. 5 and Fig. 6 it is found that by initial values different setting lithium iron phosphate dynamic battery group SOC, one of them
SOC initial values are actual value, and a SOC initial value is 0.8, and a SOC initial value is 0.2；In simulation constant-current pulse charge and discharge
Under state, there is preferable calibration result, Ke Yixun to the error present in SOC initial values using expanded Kalman filtration algorithm
Speed converges to the estimated value of SOC near actual value.SOC estimation value stabilization near actual value after, estimation result
Worst error will not be over 5%, and mean error is 3.36% or so.Become according to the convergence of the expanded Kalman filtration algorithm
Gesture can accurately estimate the SOC of lithium iron phosphate dynamic battery group with the progress of time.
2, the SOC estimations under cycle charge-discharge (dynamic operation condition) state：
The charge and discharge process that lithium iron phosphate dynamic battery group is simulated by simulation loop charging and discharging state, by the first of SOC
Initial value is respectively set to several different situations, is estimated under cycle charge discharge electricity condition to verify expanded Kalman filtration algorithm
The validity of SOC, experimental result are as shown in Figure 7.
As can be seen from FIG. 7, the initial value different by the way that lithium iron phosphate dynamic battery group SOC is arranged, one of SOC are initial
Value is actual value, and a SOC initial value is 0.8, and a SOC initial value is 0.2；Under cycle charge discharge electricity condition, using extension
Kalman filtering algorithm can not effectively correct the error present in SOC initial values, the estimated value and actual value of obtained SOC
Between there is relatively large error always.Compared with constant-current pulse charging and discharging state, SOC estimated values are in precision, stability side
Face is all short of, and basic reason is that the nonlinearity that lithium iron phosphate dynamic battery group shows under dynamic operation condition is special
Sign, simultaneously because in terms of the foundation of power battery group model, there is also error components.
In view of this, in order to enable expanded Kalman filtration algorithm to be preferably applied to lithium iron phosphate dynamic battery
The estimation of group SOC value, it is therefore desirable to some improvement be carried out to expanded Kalman filtration algorithm, improve expanded Kalman filtration algorithm
For the estimated accuracy of SOC, especially during the cycle charge-discharge of dynamic operation condition.Certainly, for constant-current pulse charge and discharge
State can choose whether to be improved and correct according to actual conditions.
According to above-mentioned analysis it is found that due to the more complicated electrochemical reaction in power battery pack inside, power battery pack exists
Outside shows the phenomenon that nonlinearity.Thevenin equivalent-circuit models mainly are used for describing by a shunt circuit RC
The polarity effect of power battery pack, therefore the model simplification of power battery pack and the Parameters variation of internal model, can cause at certain
A little aspects can not reflect the comprehensive characteristic of power battery pack, and it is inevitable to generate error.If by adding more exponent number
The shunt circuits RC describe the nonlinear characteristic of power battery pack, although can be greatly improved in terms of model accuracy,
Also bring along the complexity of equivalent-circuit model structure and the operand of SOC estimation process.Based on this, used in the present invention
Thevenin equivalent-circuit models only consider limited influence factor, mainly from the system mode noise of equivalent-circuit model and survey
It predicted, corrected in terms of amount noise, to improve the estimation accuracy and tracking effect of SOC.
According to the analysis for realizing step to expanded Kalman filtration algorithm, estimate being extended Kalman filtering state optimization
Timing, kalman gain are affected by process noise, measurement noise.Such as：When measurement noise covariance increases, karr
Graceful gain can become smaller therewith, filter effect unobvious；When measurement noise covariance is smaller, kalman gain is very big, filtering
Significant effect.
Using expanded Kalman filtration algorithm during estimating SOC, process noise and measurement noise are by artificially setting
Fixed, variance, the mean value to be generated by Matlab meet the white Gaussian noise of Gaussian Profile, have certain regularity.And it moves
When power battery pack is under actual condition, process noise, measurement noise can occur to change accordingly with load current, and process is made an uproar
The accuracy that the inaccuracy of sound and measurement noise will will have a direct impact on expanded Kalman filtration algorithm and be estimated for SOC.It is based on
This, the present invention by using Sage-Husa adaptive algorithms, in expanded Kalman filtration algorithm process noise and measurement make an uproar
Sound is modified and improves.The specific method is as follows：
To the process noise w in expanded Kalman filtration algorithm iterative processkAnd process noise covariance QkIt is modified：
Wherein, ekNewly to cease, dkFor weighting coefficient；Kk-1Indicate the iteration coefficient at k-1 moment；Ak-1Indicate etching system when k-1
State matrix；Symbol "+" indicates correction value；Indicate the optimal estimation value of k-1 moment system modes；For system
State estimation error covariance；
To the measurement noise v in expanded Kalman filtration algorithm iterative processkAnd measurement noise covariance RkIt is modified：
Wherein, Yk-1Indicate the measured value of etching system when k-1；Symbol "-" indicates discreet value；Indicate etching system when k-1
Measure discreet value；CkIndicate the corresponding state matrix of k moment observational equations.
Wherein, ekAnd dkValue be respectively：
dk=(1-b)/(1-bk+1)；
Wherein, b is forgetting factor, meets 0<b<1, general value range is [0.95,0.99].In the preferred reality of the present invention
It applies in example, takes b=0.96.
According to above-mentioned analysis it is found that the amendment of process noise and measurement noise is by experiment porch DC voltage and power battery
The terminal voltage of group equivalent-circuit model determines, is calculated according to new breath.Although the influence of model error can not be eliminated thoroughly,
It is that prediction can be modified process noise and measurement noise by Sage-Husa adaptive algorithms, improves spreading kalman
Filtering algorithm increases the real-time adjustment capability and stability of Extended Kalman filter to the estimation precision of SOC.
Next, the charge and discharge process of lithium iron phosphate dynamic battery group is simulated by simulation loop charging and discharging state, it will
The initial value of SOC is respectively set to several different situations (i.e. there are errors for SOC initial values), to verify improved extension
Kalman filtering algorithm estimates the validity of SOC under cycle charge discharge electricity condition, and improved experimental results are shown in figure 8, and
The comparison for improving front and back experimental result is as shown in Figure 9.According to Fig. 7, Fig. 8 and Fig. 9, improved Extended Kalman filter
Algorithm effectively improves to the estimation precision of SOC under cycle charge discharge electricity condition.
Below by way of specific embodiment, the lotus of the power battery pack for the hybrid power ship that the present invention will be described in detail is provided
Electricity condition computational methods.The experiment porch of hybrid power ship is mainly by hybrid power source system, marine propeller, load simulation system
The compositions such as system, control platform and collecting computer.Hybrid power source system is mainly by diesel generating set and LiFePO4 power electric
Pond group two large divisions forms, and experiment porch is simulated ship actual condition by load controller and loaded, and diesel generating set is selected
With the operating mode of lithium iron phosphate dynamic battery group.Wherein operating mode can be divided into diesel generating set individually power, ferric phosphate
Lithium power battery pack individually powers, both three kinds of modes of hybrid power supply.The experiment porch is by simulating the negative of hybrid power ship
It carries, under the independent powering mode of lithium iron phosphate dynamic battery group, the SOC of lithium iron phosphate dynamic battery group is carried out by experiment porch
It is practical to measure, and improved expanded Kalman filtration algorithm is applied to the SOC estimations of hybrid power ship power battery pack.
It is real that hybrid power ship experiment porch can carry out electric discharge according to vessel power demand to lithium iron phosphate dynamic battery group
It tests, discharge time 6min, sample frequency 50HZ, records the SOC data of lithium iron phosphate dynamic battery group as actual value.Root
According to the equivalent-circuit model of the lithium iron phosphate dynamic battery group of foundation, using improved expanded Kalman filtration algorithm to its SOC
Estimation prediction is carried out, power battery pack SOC estimation of the improved Kalman filtering algorithm applied to hybrid power ship is verified
Feasibility and reliability.
By simulating under ship actual condition loading condition in experiment, corresponding load is set, experiment porch postscript is started
The experimental data of the hybrid power system operating mode part of record is as shown in table 2：
Table 2 tests sampling section data
According to hybrid power ship fictitious load operation data, electric current, voltage and the true SOC in experimentation are carried out
Record, estimates the SOC of hybrid power ship lithium iron phosphate dynamic battery group using improved Kalman filtering algorithm.Together
Sample, there are error, the SOC estimated results of improved expanded Kalman filtration algorithm as shown in Figure 10 in SOC initial values.
According to Figure 10 experimental results it is found that improved expanded Kalman filtration algorithm is applied to hybrid power ship phosphoric acid
The SOC of iron lithium power battery pack estimates that method has feasibility, and estimation effect is preferable, improved expanded Kalman filtration algorithm
It is capable of the SOC of accurate estimation lithium iron phosphate dynamic battery group, there is certain reference significance in Practical Project utilization.
In conclusion the state-of-charge computational methods of the power battery pack of hybrid power ship proposed by the present invention, from mixed
The actual conditions of power battery pack for closing powered ship are set out, analyze the influence of hybrid power ship power battery pack SOC because
Element, and establish power battery pack equivalent-circuit model.By the analysis to Extended Kalman filter principle, for conventional Extension card
The deficiency of Kalman Filtering algorithm is modified process noise and measurement noise using Sage-Husa adaptive algorithms, and verification changes
The validity of expanded Kalman filtration algorithm after.On this basis, which is applied to the experiment of hybrid power ship to put down
The SOC of the power battery pack of platform estimates, SOC estimation is compared analysis with actual value, demonstrates algorithm estimation mixing
The validity of powered ship SOC.
Therefore, the state-of-charge computational methods of the power battery pack of hybrid power ship provided by the invention, according to mixing
The particularity of powered ship and the use environment of hybrid power ship power battery pack, using Sage-Husa adaptive algorithms
Expanded Kalman filtration algorithm is improved, process noise and measurement noise are modified.Pass through emulation experiment and mixing
Powered ship experiment porch demonstrates the algorithm estimated accuracy, improves the adaptivity and stabilization of expanded Kalman filtration algorithm
Property, theoretical foundation is provided for the estimation of practical ship power battery pack SOC, there is Practical Project reference significance.
Although present disclosure is discussed in detail by above preferred embodiment, but it should be appreciated that above-mentioned
Description is not considered as limitation of the present invention.After those skilled in the art have read the above, for the present invention's
A variety of modifications and substitutions all will be apparent.Therefore, protection scope of the present invention should be limited to the appended claims.
Claims (8)
1. a kind of state-of-charge computational methods of the power battery pack of hybrid power ship, which is characterized in that comprise the steps of：
S1, Thevenin equivalent-circuit models are based on, equivalent-circuit model is established for the power battery pack of hybrid power ship；
S2, expanded Kalman filtration algorithm is improved using Sage-Husa adaptive algorithms, including spreading kalman is filtered
Process noise and measurement noise in wave algorithm are modified；
S3, using improved expanded Kalman filtration algorithm, to the equivalent circuit mould of the power battery pack of hybrid power ship
State-of-charge of the type under charging and discharging state is estimated.
2. the state-of-charge computational methods of the power battery pack of hybrid power ship as described in claim 1, which is characterized in that
In the S1, by multiple single batteries by being connected in series or in parallel, each single battery uses power battery pack
Thevenin equivalent-circuit models.
3. the state-of-charge computational methods of the power battery pack of hybrid power ship as claimed in claim 2, which is characterized in that
When the power battery pack is composed in series by multiple single batteries, which can be equivalent to an individual monomer
Battery, the i.e. power battery pack can be equivalent to Thevenin equivalent-circuit models.
4. the state-of-charge computational methods of the power battery pack of hybrid power ship as claimed in claim 2, which is characterized in that
When the power battery pack is composed in parallel by multiple single batteries, which can be equivalent to an individual monomer
Battery, the i.e. power battery pack can be equivalent to Thevenin equivalent-circuit models.
5. the state-of-charge computational methods of the power battery pack of hybrid power ship as described in claim 3 or 4, feature exist
In in the S2, specifically comprising the steps of：
S21, to the process noise w in expanded Kalman filtration algorithm iterative processkAnd process noise covariance QkIt is modified：
Wherein, ekNewly to cease；dkFor weighting coefficient；Kk-1Indicate the iteration coefficient at k-1 moment；Ak-1Indicate the shape of etching system when k-1
State matrix；Symbol "+" indicates correction value；Indicate the optimal estimation value of k-1 moment system modes；For system mode
Evaluated error covariance；
S22, to the measurement noise v in expanded Kalman filtration algorithm iterative processkAnd measurement noise covariance RkIt is modified：
Wherein, Yk-1Indicate the measured value of etching system when k-1；Symbol "-" indicates discreet value；Indicate k-1 moment systematic surveys
Discreet value；CkIndicate the corresponding state matrix of k moment observational equations.
6. the state-of-charge computational methods of the power battery pack of hybrid power ship as claimed in claim 5, which is characterized in that
In the S21 and S22, ekAnd dkValue be respectively：
dk=(1-b)/(1-bk+1)；
Wherein, b is forgetting factor, meets 0<b<1.
7. the state-of-charge computational methods of the power battery pack of hybrid power ship as claimed in claim 5, which is characterized in that
In the S3, using expanded Kalman filtration algorithm, to the equivalent-circuit model of power battery pack in constant-current pulse charge and discharge
The state-of-charge of state is estimated；Using improved expanded Kalman filtration algorithm, to the equivalent circuit of power battery pack
Model is estimated in the state-of-charge of cycle charge discharge electricity condition.
8. the state-of-charge computational methods of the power battery pack of hybrid power ship as claimed in claim 5, which is characterized in that
In the S3, using improved expanded Kalman filtration algorithm, to the equivalent-circuit model of power battery pack respectively in perseverance
The state-of-charge of stream pulse charging and discharging state and cycle charge discharge electricity condition is estimated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810639464.5A CN108427080A (en) | 2018-06-20 | 2018-06-20 | The state-of-charge computational methods of the power battery pack of hybrid power ship |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810639464.5A CN108427080A (en) | 2018-06-20 | 2018-06-20 | The state-of-charge computational methods of the power battery pack of hybrid power ship |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108427080A true CN108427080A (en) | 2018-08-21 |
Family
ID=63164840
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810639464.5A Pending CN108427080A (en) | 2018-06-20 | 2018-06-20 | The state-of-charge computational methods of the power battery pack of hybrid power ship |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108427080A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112470017A (en) * | 2019-02-07 | 2021-03-09 | 株式会社Lg化学 | Battery management device, battery management method and battery pack |
| CN113533995A (en) * | 2021-07-05 | 2021-10-22 | 上海电享信息科技有限公司 | Power battery consistency detection method |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102981125A (en) * | 2012-11-30 | 2013-03-20 | 山东省科学院自动化研究所 | SOC (Stress Optical Coefficient) estimation method for power batteries based on RC (Remote Control) equivalent model |
| CN104502851A (en) * | 2014-12-12 | 2015-04-08 | 广西科技大学 | SOC (Stress Optical Coefficient) estimation method based on AUKF (Adaptive Unscented Kalman Filter) algorithm |
| CN106597304A (en) * | 2016-12-08 | 2017-04-26 | 盐城工学院 | Battery system charge of state estimation method using improved noise estimator |
| CN106772094A (en) * | 2017-01-09 | 2017-05-31 | 成都理工大学 | A kind of SOC methods of estimation of the battery model based on parameter adaptive |
| CN106767900A (en) * | 2016-11-23 | 2017-05-31 | 东南大学 | A kind of online calibration method of the optical fibre SINS system based on integrated navigation technology |
| CN107192961A (en) * | 2017-07-12 | 2017-09-22 | 江苏维科新能源科技有限公司 | Novel power battery SOC estimation method |
| WO2017179175A1 (en) * | 2016-04-14 | 2017-10-19 | 富士通株式会社 | Estimation device, estimation program, and charging control device |
| CN108020791A (en) * | 2017-12-04 | 2018-05-11 | 上海海事大学 | A kind of hybrid power ship lithium iron phosphate dynamic battery group state-of-charge method of estimation |
-
2018
- 2018-06-20 CN CN201810639464.5A patent/CN108427080A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102981125A (en) * | 2012-11-30 | 2013-03-20 | 山东省科学院自动化研究所 | SOC (Stress Optical Coefficient) estimation method for power batteries based on RC (Remote Control) equivalent model |
| CN104502851A (en) * | 2014-12-12 | 2015-04-08 | 广西科技大学 | SOC (Stress Optical Coefficient) estimation method based on AUKF (Adaptive Unscented Kalman Filter) algorithm |
| WO2017179175A1 (en) * | 2016-04-14 | 2017-10-19 | 富士通株式会社 | Estimation device, estimation program, and charging control device |
| CN106767900A (en) * | 2016-11-23 | 2017-05-31 | 东南大学 | A kind of online calibration method of the optical fibre SINS system based on integrated navigation technology |
| CN106597304A (en) * | 2016-12-08 | 2017-04-26 | 盐城工学院 | Battery system charge of state estimation method using improved noise estimator |
| CN106772094A (en) * | 2017-01-09 | 2017-05-31 | 成都理工大学 | A kind of SOC methods of estimation of the battery model based on parameter adaptive |
| CN107192961A (en) * | 2017-07-12 | 2017-09-22 | 江苏维科新能源科技有限公司 | Novel power battery SOC estimation method |
| CN108020791A (en) * | 2017-12-04 | 2018-05-11 | 上海海事大学 | A kind of hybrid power ship lithium iron phosphate dynamic battery group state-of-charge method of estimation |
Non-Patent Citations (1)
| Title |
|---|
| 程晨 等: "基于Sage-Husa自适应滤波算法的锂电池荷电状态估计", 《南京工业大学学报(自然科学版)》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112470017A (en) * | 2019-02-07 | 2021-03-09 | 株式会社Lg化学 | Battery management device, battery management method and battery pack |
| CN112470017B (en) * | 2019-02-07 | 2023-12-01 | 株式会社Lg新能源 | Battery management device, battery management method, and battery pack |
| US11923710B2 (en) | 2019-02-07 | 2024-03-05 | Lg Energy Solution, Ltd. | Battery management apparatus, battery management method and battery pack |
| CN113533995A (en) * | 2021-07-05 | 2021-10-22 | 上海电享信息科技有限公司 | Power battery consistency detection method |
| CN113533995B (en) * | 2021-07-05 | 2023-10-20 | 上海电享信息科技有限公司 | Consistency detection method for power battery |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kwak et al. | Parameter identification and SOC estimation of a battery under the hysteresis effect | |
| CN103926538B (en) | Change exponent number RC equivalent-circuit model based on AIC criterion and implementation method | |
| CN105548901B (en) | A kind of rail traffic lithium titanate battery power rating prediction technique | |
| CN110146816B (en) | Method, device and equipment for determining remaining charging time of battery and storage medium | |
| Sun et al. | Overview of the types of battery models | |
| CN103632018B (en) | A kind of fuel cell modelling method based on Simscape platform | |
| CN103616647B (en) | A kind of estimation method of battery dump energy for cell management system of electric automobile | |
| CN107576919A (en) | Power battery charged state estimating system and method based on ARMAX models | |
| CN107589379A (en) | A kind of On-line Estimation lithium battery SOC and the method for impedance | |
| CN110824363B (en) | Lithium battery SOC and SOE joint estimation method based on improved CKF | |
| CN106019164A (en) | Lithium battery SOC estimation algorithm based on dual adaptive unscented Kalman filter | |
| CN103293485A (en) | Model-based storage battery SOC (state of charge) estimating method | |
| Moura | Estimation and control of battery electrochemistry models: A tutorial | |
| CN109633472A (en) | A kind of state-of-charge algorithm for estimating of lithium battery monomer | |
| Sun et al. | Study of parameters identification method of Li-ion battery model for EV power profile based on transient characteristics data | |
| CN108020791A (en) | A kind of hybrid power ship lithium iron phosphate dynamic battery group state-of-charge method of estimation | |
| Enache et al. | Comparative study for generic battery models used for electric vehicles | |
| Plakhtii et al. | The analysis of mathematical models of charge-discharge characteristics in lithium-ion batteries | |
| CN105116338A (en) | Parallel type battery system modeling method based on SOC compensator | |
| Kazhamiaka et al. | Li-ion storage models for energy system optimization: the accuracy-tractability tradeoff | |
| CN108427080A (en) | The state-of-charge computational methods of the power battery pack of hybrid power ship | |
| CN109635446B (en) | Method for fitting open-circuit potential curve of composite electrode and application thereof | |
| CN110850322A (en) | Method for estimating relative state of health of battery based on wavelet signal decomposition | |
| CN106126798A (en) | lithium iron phosphate storage battery SOC algorithm | |
| Huang et al. | State of charge estimation of lithium-ion battery based on second-order extended Kalman filter |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180821 |