CN114839550A - UKF-AUKF-based lithium battery SOC joint estimation method - Google Patents
UKF-AUKF-based lithium battery SOC joint estimation method Download PDFInfo
- Publication number
- CN114839550A CN114839550A CN202210395241.5A CN202210395241A CN114839550A CN 114839550 A CN114839550 A CN 114839550A CN 202210395241 A CN202210395241 A CN 202210395241A CN 114839550 A CN114839550 A CN 114839550A
- Authority
- CN
- China
- Prior art keywords
- soc
- lithium battery
- battery
- model
- estimation
- 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.)
- Granted
Links
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 title claims abstract description 75
- 229910052744 lithium Inorganic materials 0.000 title claims abstract description 75
- 238000000034 method Methods 0.000 title claims description 31
- 238000001914 filtration Methods 0.000 claims abstract description 15
- 238000002474 experimental method Methods 0.000 claims description 17
- 230000010287 polarization Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 239000003990 capacitor Substances 0.000 claims description 4
- 238000007599 discharging Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 3
- 238000010835 comparative analysis Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 4
- 230000003044 adaptive effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/385—Arrangements for measuring battery or accumulator variables
- G01R31/387—Determining ampere-hour charge capacity or SoC
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/60—Other road transportation technologies with climate change mitigation effect
- Y02T10/70—Energy storage systems for electromobility, e.g. batteries
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Tests Of Electric Status Of Batteries (AREA)
Abstract
The invention discloses a guarantee that the accurate estimation of the state of charge (SOC) of a lithium battery is safe in driving of an electric automobile. In order to reduce errors caused by the fact that a battery model does not conform to the time-varying characteristics of actual battery parameters under actual complex working conditions, an unscented Kalman filtering algorithm (UKF) is adopted to carry out online parameter identification on a lithium battery second-order equivalent model, the SOC of a lithium battery is estimated by combining with a self-adaptive unscented Kalman filtering Algorithm (AUKF), time-varying parameters are fed back to a model for SOC estimation, the SOC estimation precision and the adaptability to various working conditions are improved, comparative analysis is carried out on the battery model and an offline single extended Kalman filtering algorithm (EKF) and an online double extended Kalman filtering algorithm (DEKF) under the UDDS working conditions, and the accuracy and the robustness of the UKF-AUKF are verified through experimental results.
Description
Technical Field
The invention relates to the field of lithium battery state of charge estimation, in particular to a method for estimating SOC based on UKF (unscented Kalman Filter) online parameter identification and AUKF (adaptive unscented Kalman Filter).
Background
In order to meet the requirement of the era of sustainable development, new energy electric automobiles are rapidly developed. The lithium battery is the core of the working operation of the electric automobile, so the estimation of the state of the lithium battery becomes a central importance. A great deal of research is carried out on the estimation of the SOC of the lithium battery at home and abroad, at present, most of the research is based on models, firstly, the SOC of the battery is estimated by analyzing and calculating the chemical reaction of the lithium battery through an electrochemical model, and the method needs certain related chemical theory and has more complex calculation; secondly, a black box model is built through a large amount of data drive, such as a neural network and the like, but the method needs a large amount of data to train the model, and has poor effect under complex working conditions and environments; and thirdly, an equivalent circuit model, wherein the SOC is estimated by combining a circuit model formed by resistors and capacitors with various filtering algorithms, the method has small calculated amount and high accuracy, and the estimation accuracy under various complex working conditions can be guaranteed, so that the method is widely applied.
The model accuracy directly influences the estimation accuracy of the SOC, and Cao et al propose offline parameter identification based on RLS, so that the identification accuracy of the model is ensured; RLS based on constraint conditions carries out online parameter identification, and battery SOC is estimated by UKF, so that model accuracy is further improved, but the RLS algorithm adopts fewer data points and has relatively lower accuracy; parameter identification is carried out by adopting a recursive least square method (FFRLS) with a forgetting factor, the SOC of the battery is estimated by using the EKF, historical data is abandoned by setting the forgetting factor, the problem of data redundancy is effectively solved, but the forgetting factor of the algorithm is fixed and possibly suitable for the current working condition, but errors can be increased sharply for other working conditions, and the method does not have universality for different working conditions; parameter and SOC joint estimation are carried out to two extension kalman filtering algorithm (DEKF), have realized online parameter identification, improve the adaptability of model to different operating modes greatly, but linearization process among the EKF causes the error to continuously exist because of omitting partial high order term, and the noise influence of operating condition can't be handled moreover.
Disclosure of Invention
The method aims at the problems that the time-varying characteristics of lithium battery parameters and noise have large influence on SOC estimation under actual working conditions. The invention provides a method for identifying parameters of an equivalent circuit model by adopting an UKF algorithm under a macroscopic time scale, and estimating the SOC of a lithium battery by combining the AUKF algorithm under a microscopic time scale, thereby solving the problems of larger error caused by fixing a traditional offline parameter identification model and low algorithm precision caused by omitting higher-order terms of EKF, further filtering environmental and working condition noise, and improving the robustness and the accuracy of the algorithm.
The technical scheme adopted by the invention is as follows:
step 1: the experimental object for building the experimental platform adopts a battery pack formed by connecting ten 18650 ternary lithium batteries in parallel, the experimental platform is as shown in figure 1, a temperature control box is used for setting the working temperature of the lithium batteries, the constant temperature is set to be 25 ℃, and a PC (personal computer) end sends instructions to a programmable electronic load and a programmable DC (direct current) power supply through a serial port to perform charging and discharging experiments on the batteries. Meanwhile, the data acquisition card is used for acquiring lithium battery data and transmitting the lithium battery data to the PC terminal in real time.
Step 2: the lithium battery model is the basis of SOC estimation, balances calculated amount and precision, and selects a second-order equivalent circuit model. And establishing a lithium battery loop equation, and establishing a state space equation of the lithium battery online parameters according to Laplace transform.
And step 3: the parameter identification of the model is mainly completed through two experiments, namely, the relationship between the Uoc and the SOC is determined through a constant current discharge experiment, and R0, R1, C1, R2 and C2 are identified on line through a UDDS experiment. And determining the functional relation between each parameter of the equivalent circuit and the SOC.
And 4, step 4: and (3) under the macroscopic time scale, performing lithium battery parameter identification by using an unscented Kalman filtering algorithm.
And 5: and after the parameters of the lithium battery model are updated, switching to a microscopic time scale, and estimating the SOC of the lithium battery by using a self-adaptive unscented Kalman filtering algorithm.
Step 6: and (4) carrying out verification on the joint estimation precision under a pulse discharge working condition and a UDDS working condition (urban road circulation).
Further, the step 2 comprises:
step 2.1: establishing a second-order equivalent circuit model of the lithium battery, and listing a loop equation and an observation equation according to the second-order equivalent circuit model:
U 0 ＝U oc (SOC)-R 0 I-U 1 -U 2 (2)
in the formula: c1 and C2 are polarization capacitors, Uoc is open-circuit voltage, U0 is terminal voltage, Ts is sampling time, Qn is battery capacity, R1 and R2 are polarization resistors, and R0 is ohmic internal resistance of the battery.
Step 2.2: carrying out differential discretization on the formulas (1) and (2) to obtain:
step 2.3: for a non-linear system:
the system noise value is a measurement noise value.
Further, the step 3 comprises:
step 3.1: discharging the fully charged lithium battery with 30A constant pulse current for 6 minutes, namely 0.1 SOC value, standing the battery for four hours after the discharge is finished, recording the open-circuit voltage of the battery, and repeating the operation for 10 times;
step 3.2: according to 10 groups of data points of the relationship between the Uoc and the SOC measured by experiments, the SOC is used as a variable, and the data points of the Uoc and the SOC are subjected to eight-order fitting (the eight-order fitting effect is good) through a formula (5), so that a function of the Uoc and the SOC is obtained:
U ocv (SOC)＝p 1 SOC 8 +p 2 SOC 7 +p 3 SOC 6 +p 4 SOC 5 +p 5 SOC 4 +p 6 SOC 3 +p 7 SOC 2 +p 8 SOC+p 9 (5)
step 3.3: and adopting American urban circulation system (UDDS) as an online parameter to identify the actual working condition. And under the condition of meeting the working condition of the battery, circulating the working condition once, reducing the SOC value of the battery by 5 percent, and circulating the working condition 20 times until the SOC is equal to 0.
Step 3.4: the thermostat is set to be 25 ℃, the battery pack is excited by adopting UDDS working condition current, and the actually measured voltage of the battery pack under the working condition is acquired by the data acquisition card to obtain the current and the voltage under the actual working condition.
Further, the step 4 comprises:
step 4.1: firstly, according to the state space equation of formula (3), a state space equation with parameter variables and state variables as independent variables is expressed:
in the above formula, theta is [ R R ] 1 C 1 R 2 C 2 ] T The macro scale L is 60s, and the micro scale sequence L is formed by (1-L) w k v k For process noise and observation noise of the system, p k Model parametric process noise.
Step 4.2: parameter variables and parameter variable covariances are initialized, and parameters alpha, ki, β, and M of the UKF algorithm are determined to be 0.01, 0, 2, and 5, respectively.
Step 4.3: calculating sampling points at the k moment:
step 4.4: calculating the weight:
step 4.5: parameter predicted value and system variance predicted value Pxx:
step 4.6: the prediction parameters are updated, the observations and the observation variance predictions are updated Pyy, and the parameter variables covariance and the unscented kalman gain K are solved simultaneously.
Step 4.7: updating the observation value of the system:
step 4.8: updating parameters of the system:
further, the step 5 comprises:
step 5.1: the UKF-AUKF jointly estimates the SOC of the lithium battery, the UKF is adopted to perform online parameter identification on a macroscopic time scale, the AUKF is used to estimate the SOC of the lithium battery on a microscopic time scale, and online parameter identification and SOC joint estimation of the lithium battery based on the UKF-AUKF are achieved.
Step 5.2: for the initial value of the circuit model parameter, the advantage of accurate offline parameter identification is taken, and the model parameter obtained by offline parameter identification by the recursive least square method is used as the initial value of the lithium battery joint estimation
Step 5.3: establishing battery equivalent circuit model and state space equation of initial state of lithium battery
Step 5.4: and performing SOC estimation by using a 60s micro time scale sequence, switching the time scale to a macro time scale for primary parameter identification after the SOC estimation reaches 60s, updating the identified parameters into a state space equation, switching to the micro time scale for SOC estimation, and circularly and repeatedly realizing the joint estimation of online parameter identification and SOC.
Further, the step 6 comprises:
step 6.1: the accuracy of the model is mainly embodied in two aspects, namely the error between the terminal voltage of the model and the true value, and the other aspect is embodied in the estimation accuracy of the SOC.
Step 6.2: and verifying the online parameter identification precision and the lithium battery SOC estimation precision of the joint estimation algorithm by using the pulse discharge working condition and the UDDS working condition.
Step 6.3: in order to further explore the precision of the joint estimation algorithm, the traditional RLS-EKF algorithm of RLS off-line parameter identification, the DEKF algorithm of EKF on-line parameter identification and the UKF-AUKF algorithm are respectively adopted for comparison.
Compared with the existing lithium battery state of charge estimation, the method has the following advantages and beneficial effects:
1. according to the invention, online parameter identification is carried out on the lithium battery second-order equivalent model through the UKF algorithm, the SOC of the lithium battery is estimated by combining the AUKF algorithm, the average absolute error of the SOC is reduced to 0.0051, the problems of time variation of lithium battery parameters under complex working conditions and noise influence in the SOC estimation process are solved, and the precision of the battery model is greatly improved; the accuracy and stability of the method are further verified through comparison and analysis of various parameter identification methods and SOC estimation algorithms under the pulse discharge working condition and the UDDS working condition.
2. According to the method, as seen from results of RLS-EKF and DEKF, compared with traditional offline parameter identification, the online parameter identification precision is obviously high, and a model for online parameter identification is updated in a self-adaptive manner along with the change of working conditions, so that the method is more suitable for the actual working conditions of the lithium battery; UKF-AUKF precision is the highest, compares in DEKF, and its SOC estimation error is littleer, and along with the sharp change of operating mode, its error fluctuation is little moreover, and is extremely stable, and the robustness is good.
Drawings
Fig. 1 is a flow chart of online parameter identification and SOC joint estimation of a lithium battery according to an embodiment of the present invention.
Fig. 2 is a graph of fitted open circuit voltage and state of charge curves for a lithium battery according to an embodiment of the present invention.
Fig. 3 is a voltage and current curve diagram of a lithium battery according to an embodiment of the present invention under UDDS cycle conditions.
FIG. 4 is a diagram of SOC estimation error curves of algorithms of a lithium battery according to an embodiment of the present invention under a pulse discharge condition.
Fig. 5 is a SOC estimation curve diagram of each algorithm of a lithium battery according to an embodiment of the present invention under a UDDS cycle condition.
Fig. 6 is a SOC estimation error curve diagram of each algorithm of a lithium battery according to an embodiment of the present invention under a UDDS cycle condition.
Detailed Description
According to the invention, online parameter identification is carried out on the lithium battery second-order equivalent model through the UKF algorithm, the SOC of the lithium battery is estimated by combining the AUKF algorithm, the time-varying problem of the lithium battery parameters under complex working conditions and the noise influence in the SOC estimation process are solved, and the error fluctuation is small, extremely stable and good in robustness along with the rapid change of the working conditions.
The present invention will be described in further detail with reference to the accompanying drawings.
As shown in fig. 1, the joint estimation of the state of charge of the lithium battery based on the unscented kalman filter algorithm on-line parameter identification and the adaptive unscented kalman filter mainly comprises the following steps:
step 1: and (5) establishing an experiment platform for experiment.
Step 2: the lithium battery model is the basis of SOC estimation, balances calculated amount and precision, and selects a second-order equivalent circuit model. And establishing a lithium battery loop equation by using the circuit knowledge, and establishing a state space equation of the online parameters of the lithium battery according to Laplace transform.
And step 3: the parameter identification of the model is mainly completed through two experiments, namely, the relationship between the Uoc and the SOC is determined through a constant current discharge experiment, and R0, R1, C1, R2 and C2 are identified on line through a UDDS experiment. And determining the functional relation between each parameter of the equivalent circuit and the SOC.
And 4, step 4: and (3) under the macroscopic time scale, performing lithium battery parameter identification by using an unscented Kalman filtering algorithm.
And 5: and after the parameters of the lithium battery model are updated, switching to a microscopic time scale, and estimating the SOC of the lithium battery by using a self-adaptive unscented Kalman filtering algorithm.
Step 6: and (4) verifying the joint estimation precision under a pulse discharge working condition and a UDDS working condition (urban road circulation).
Further, the step 2 comprises:
step 2.1: establishing a second-order equivalent circuit model of the lithium battery, and listing a loop equation and an observation equation according to the second-order equivalent circuit model:
U 0 ＝U oc (SOC)-R 0 I-U 1 -U 2 (13)
in the formula: c1 and C2 are polarization capacitors, Uoc is open-circuit voltage, U0 is terminal voltage, Ts is sampling time, Qn is battery capacity, R1 and R2 are polarization resistors, and R0 is ohmic internal resistance of the battery.
Step 2.2: the differential discretization of the formulas (12) and (13) is carried out to obtain:
step 2.3: for a non-linear system
The system noise value is a measurement noise value.
Further, the step 3 comprises:
step 3.1: discharging the fully charged lithium battery with 30A constant pulse current for 6 minutes, namely 0.1 SOC value, standing the battery for four hours after the discharge is finished, recording the open-circuit voltage of the battery, and repeating the operation for 10 times; the Uoc-SOC curve obtained by fitting is shown in FIG. 2.
Step 3.2: according to 10 groups of data points of the relationship between the Uoc and the SOC measured by experiments, the SOC is used as a variable, and the data points of the Uoc and the SOC are subjected to eight-order fitting (the eight-order fitting effect is good) through a formula (16), so that a function of the Uoc and the SOC is obtained:
U ocv (SOC)＝p 1 SOC 8 +p 2 SOC 7 +p 3 SOC 6 +p 4 SOC 5 +p 5 SOC 4 +p 6 SOC 3 +p 7 SOC 2 +p 8 SOC+p 9 (16)
step 3.3: and adopting American urban circulation system (UDDS) as an online parameter to identify the actual working condition. And under the condition of meeting the working condition of the battery, circulating the working condition once, reducing the SOC value of the battery by 5 percent, and circulating the working condition 20 times until the SOC is equal to 0.
Step 3.4: the thermostat is set to 25 ℃, the battery pack is excited by adopting UDDS working condition current, the actually measured voltage of the battery pack under the working condition is collected through a data acquisition card, the current and the voltage under the actual working condition are obtained, and the experimental voltage and current curves under the UDDS circulating working condition are obtained and are shown in figure 3.
Further, the step 4 comprises:
step 4.1: firstly, a state space equation with parameter variables and state variables as independent variables is expressed according to the state space equation of formula (14):
in the above formula, theta is [ R R ] 1 C 1 R 2 C 2 ] T The macro scale L is 60s, and the micro scale sequence L is formed by (1-L) w k v k For process and observation noise, ρ, of the system k Model parametric process noise.
And 4.2: parameter variables and parameter variable covariances are initialized, and parameters alpha, ki, β, and M of the UKF algorithm are determined to be 0.01, 0, 2, and 5, respectively.
Step 4.3: calculating k time sampling point
Step 4.4: calculating weights
Step 4.5: parameter prediction value and system variance prediction value
Step 4.6: the prediction parameters are updated in such a way that,
step 4.7: the observation and observation variance prediction values are updated Pyy,
step 4.8: covariance of parameter variables and unscented Kalman gain K
Step 4.9: observation update for a system
Step 4.10: parameter updating of a system
Further, the step 5 comprises:
step 5.1: the UKF-AUKF jointly estimates the SOC of the lithium battery, the UKF is adopted to perform online parameter identification on a macroscopic time scale, the AUKF is used to estimate the SOC of the lithium battery on a microscopic time scale, and online parameter identification and SOC joint estimation of the lithium battery based on the UKF-AUKF are achieved.
Step 5.2: for the initial value of the circuit model parameter, the advantage of accurate offline parameter identification is taken, and the model parameter obtained by offline parameter identification by the recursive least square method is used as the initial value of the lithium battery joint estimation
Step 5.3: establishing battery equivalent circuit model and state space equation of initial state of lithium battery
Step 5.4: and performing SOC estimation by using a 60s micro time scale sequence, switching the time scale to a macro time scale for primary parameter identification after the SOC estimation reaches 60s, updating the identified parameters into a state space equation, switching to the micro time scale for SOC estimation, and circularly and repeatedly realizing the joint estimation of online parameter identification and SOC.
Further, the step 6 comprises:
step 6.1: the accuracy of the model is mainly embodied in two aspects, namely the error between the terminal voltage and the true value of the model, and the estimation accuracy of the SOC is realized on the other aspect.
Step 6.2: and verifying the online parameter identification precision and the lithium battery SOC estimation precision of the joint estimation algorithm by using the pulse discharge working condition and the UDDS working condition.
Step 6.3: in order to further explore the precision of the joint estimation algorithm, the traditional RLS-EKF algorithm of RLS off-line parameter identification, the DEKF algorithm of EKF on-line parameter identification and the UKF-AUKF algorithm are respectively adopted for comparison.
Compared with the estimation of the SOC of the lithium battery by the EKF algorithm under RLS offline parameter identification, the SOC estimation error is shown in FIG. 4, and the UKF-AUKF has smaller SOC error, less fluctuation and good robustness on the whole. The model precision difference obtained by identifying the two parameters in the standing stage is not large, but when the current changes, the UKF online parameter identification can always track the parameter change quickly, and the model identified by the RLS is always constant and unchangeable, so that the error becomes large and the fluctuation occurs. In order to further explore the precision of the joint estimation algorithm, SOC estimation is carried out on a model obtained by identifying each parameter under the UDDS working condition, namely an RLS-EKF algorithm adopting traditional RLS offline parameter identification, an EKF online parameter identification DEKF algorithm and a UKF-AUKF algorithm are adopted, battery SOC estimation is shown in FIG. 5, SOC estimation errors of each joint estimation are shown in FIG. 6, it can be seen from the figure that the UKF-AUKF is closest to a true value, the DEKF algorithm estimation precision is slightly poor, and the RLS-EKF estimation precision is worst; when the SOC algorithms are guaranteed to be EKF algorithms, the DEKF is obviously superior to the RLS-EKF, the accuracy of the online parameter identification model is proved, and the SOC estimation method based on online parameter identification of the DEKF and the UKF-AUKF is higher in accuracy compared with the SOC estimation method based on offline parameter identification. Compared with DEKF, the SOC estimation error of UKF-AUKF is smaller, which further explains that UKF-AUKF solves the process of expanding linearization by Taylor formula in EKF algorithm, and avoids the error caused by omitting high-order terms due to linearization; the error fluctuation of the algorithm is small under the complex working condition, which shows that the UKF-AUKF can well filter the noise.
The above description is only an embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention.
Claims (1)
1. The invention relates to a lithium battery state of charge jointly estimated based on unscented Kalman filtering algorithm on-line parameter identification and self-adaptive unscented Kalman filtering, which mainly comprises the following steps:
step 1: and (5) establishing an experiment platform for experiment.
Step 2: the lithium battery model is the basis of SOC estimation, balances calculated amount and precision, and selects a second-order equivalent circuit model. And establishing a lithium battery loop equation by using the circuit knowledge, and establishing a state space equation of the online parameters of the lithium battery according to Laplace transform.
And step 3: the parameter identification of the model is mainly completed through two experiments, namely, the relationship between the Uoc and the SOC is determined through a constant current discharge experiment, and R0, R1, C1, R2 and C2 are identified on line through a UDDS experiment. And determining the functional relation between each parameter of the equivalent circuit and the SOC.
And 4, step 4: and (3) identifying the lithium battery parameters by using an unscented Kalman filtering algorithm under a macroscopic time scale.
And 5: and after the parameters of the lithium battery model are updated, switching to a microscopic time scale, and estimating the SOC of the lithium battery by using a self-adaptive unscented Kalman filtering algorithm.
Step 6: and (4) verifying the joint estimation precision under a pulse discharge working condition and a UDDS working condition (urban road circulation).
Further, the step 2 comprises:
step 2.1: establishing a second-order equivalent circuit model of the lithium battery, and listing a loop equation and an observation equation according to the second-order equivalent circuit model:
U 0 ＝U oc (SOC)-R 0 I-U 1 -U 2 (2)
in the formula: c1 and C2 are polarization capacitors, Uoc is open-circuit voltage, U0 is terminal voltage, Ts is sampling time, Qn is battery capacity, R1 and R2 are polarization resistors, and R0 is ohmic internal resistance of the battery.
Step 2.2: carrying out differential discretization on the formulas (1) and (2) to obtain:
step 2.3: for a non-linear system
The system noise value is a measurement noise value.
Further, the step 3 includes:
step 3.1: discharging the fully charged lithium battery with 30A constant pulse current for 6 minutes, namely 0.1 SOC value, standing the battery for four hours after the discharge is finished, recording the open-circuit voltage of the battery, and repeating the operation for 10 times; the Uoc-SOC curve obtained by fitting is shown in FIG. 2.
Step 3.2: and carrying out eight-order fitting (the eight-order fitting effect is good) on the data points of the Uoc and the SOC by taking the SOC as a variable according to 10 groups of data points of the relationship between the Uoc and the SOC measured by experiments, thereby obtaining a function of the Uoc and the SOC.
Step 3.3: and adopting American urban circulation system (UDDS) as an online parameter to identify the actual working condition. And under the condition of meeting the working condition of the battery, circulating the working condition once, reducing the SOC value of the battery by 5 percent, and circulating the working condition 20 times until the SOC is equal to 0.
Step 3.4: the thermostat is set to 25 ℃, the battery pack is excited by adopting UDDS working condition current, the actually measured voltage of the battery pack under the working condition is collected through a data acquisition card, the current and the voltage under the actual working condition are obtained, and the experimental voltage and current curves under the UDDS circulating working condition are obtained and are shown in figure 3.
Further, the step 4 comprises:
step 4.1: firstly, according to the state space equation of formula (3), a state space equation with parameter variables and state variables as independent variables is expressed:
in the above formula, theta is [ R R ] 1 C 1 R 2 C 2 ] T The macro scale L is 60s, and the micro scale sequence L is formed by (1-L) w k v k For process noise and observation noise of the system, p k Model parametric process noise.
Step 4.2: parameter variables and parameter variable covariances are initialized, and parameters alpha, ki, β, and M of the UKF algorithm are determined to be 0.01, 0, 2, and 5, respectively.
Step 4.3: calculating k time sampling point
Step 4.4: calculating weights
Step 4.5: parameter prediction value and system variance prediction value
Step 4.6: the prediction parameters are updated in such a way that,
step 4.7: the observation and observation variance prediction values are updated Pyy,
step 4.8: covariance of parameter variables and unscented Kalman gain K
Step 4.9: observation update for a system
Step 4.10: parameter updating of a system
Further, the step 5 comprises:
step 5.1: the UKF-AUKF jointly estimates the SOC of the lithium battery, the UKF is adopted to perform online parameter identification on a macroscopic time scale, the AUKF is used to estimate the SOC of the lithium battery on a microscopic time scale, and online parameter identification and SOC joint estimation of the lithium battery based on the UKF-AUKF are achieved.
Step 5.2: for the initial value of the circuit model parameter, the advantage of accurate offline parameter identification is taken, and the model parameter obtained by offline parameter identification by the recursive least square method is used as the initial value of the lithium battery joint estimation
Step 5.3: establishing battery equivalent circuit model and state space equation of initial state of lithium battery
Step 5.4: and performing SOC estimation by using a 60s micro time scale sequence, switching the time scale to a macro time scale for primary parameter identification after the SOC estimation reaches 60s, updating the identified parameters into a state space equation, switching to the micro time scale for SOC estimation, and circularly and repeatedly realizing the joint estimation of online parameter identification and SOC.
Further, the step 6 comprises:
step 6.1: the accuracy of the model is mainly embodied in two aspects, namely the error between the terminal voltage of the model and the true value, and the other aspect is embodied in the estimation accuracy of the SOC.
Step 6.2: and verifying the online parameter identification precision and the lithium battery SOC estimation precision of the joint estimation algorithm by using the pulse discharge working condition and the UDDS working condition.
Step 6.3: in order to further explore the precision of the joint estimation algorithm, the traditional RLS-EKF algorithm of RLS off-line parameter identification, the DEKF algorithm of EKF on-line parameter identification and the UKF-AUKF algorithm are respectively adopted for comparison.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210395241.5A CN114839550B (en) | 2022-04-14 | 2022-04-14 | UKF-AUKF-based lithium battery SOC joint estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210395241.5A CN114839550B (en) | 2022-04-14 | 2022-04-14 | UKF-AUKF-based lithium battery SOC joint estimation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114839550A true CN114839550A (en) | 2022-08-02 |
| CN114839550B CN114839550B (en) | 2024-05-10 |
Family
ID=82565291
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210395241.5A Active CN114839550B (en) | 2022-04-14 | 2022-04-14 | UKF-AUKF-based lithium battery SOC joint estimation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114839550B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112858928A (en) * | 2021-03-08 | 2021-05-28 | 安徽理工大学 | Lithium battery SOC estimation method based on online parameter identification |
| CN113138344B (en) * | 2021-04-23 | 2024-03-05 | 安徽理工大学 | SOC estimation method based on fractional order extended Kalman filtering algorithm |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103744026A (en) * | 2013-12-19 | 2014-04-23 | 广西科技大学 | Storage battery state of charge estimation method based on self-adaptive unscented Kalman filtering |
| CN107064811A (en) * | 2017-03-01 | 2017-08-18 | 华南理工大学 | A kind of lithium battery SOC On-line Estimation methods |
| WO2019113999A1 (en) * | 2017-12-14 | 2019-06-20 | 江苏大学 | Gas-liquid dynamic model-based accurate lithium-ion battery soc estimation method |
| CN112858928A (en) * | 2021-03-08 | 2021-05-28 | 安徽理工大学 | Lithium battery SOC estimation method based on online parameter identification |
| CN113030752A (en) * | 2021-04-12 | 2021-06-25 | 安徽理工大学 | Online parameter identification and SOC joint estimation method based on forgetting factor |
| US20210278466A1 (en) * | 2020-03-03 | 2021-09-09 | Karma Automotive Llc | Method of determining the state of charge of a battery used in an electric vehicle |
| CN114167298A (en) * | 2021-10-28 | 2022-03-11 | 上海空间电源研究所 | Lithium ion battery SOC estimation method and system based on improved EKF |
-
2022
- 2022-04-14 CN CN202210395241.5A patent/CN114839550B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103744026A (en) * | 2013-12-19 | 2014-04-23 | 广西科技大学 | Storage battery state of charge estimation method based on self-adaptive unscented Kalman filtering |
| CN107064811A (en) * | 2017-03-01 | 2017-08-18 | 华南理工大学 | A kind of lithium battery SOC On-line Estimation methods |
| WO2019113999A1 (en) * | 2017-12-14 | 2019-06-20 | 江苏大学 | Gas-liquid dynamic model-based accurate lithium-ion battery soc estimation method |
| US20210278466A1 (en) * | 2020-03-03 | 2021-09-09 | Karma Automotive Llc | Method of determining the state of charge of a battery used in an electric vehicle |
| CN112858928A (en) * | 2021-03-08 | 2021-05-28 | 安徽理工大学 | Lithium battery SOC estimation method based on online parameter identification |
| CN113030752A (en) * | 2021-04-12 | 2021-06-25 | 安徽理工大学 | Online parameter identification and SOC joint estimation method based on forgetting factor |
| CN114167298A (en) * | 2021-10-28 | 2022-03-11 | 上海空间电源研究所 | Lithium ion battery SOC estimation method and system based on improved EKF |
Non-Patent Citations (3)
| Title |
|---|
| LI WANG 等: "Magnetic and Cryogenic Design of MICE Coupling Solenoid Magnet System", IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, vol. 19, no. 3, 30 June 2019 (2019-06-30) * |
| 卢云帆 等: "基于UKF-AUKF锂电池在线参数辨识和SOC联合估计", 电源技术, vol. 46, no. 10, 31 October 2022 (2022-10-31) * |
| 边东生: "基于AUKF算法的锂电池SOC估算", 现代机械, no. 1, 28 February 2022 (2022-02-28) * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112858928A (en) * | 2021-03-08 | 2021-05-28 | 安徽理工大学 | Lithium battery SOC estimation method based on online parameter identification |
| CN112858928B (en) * | 2021-03-08 | 2024-02-06 | 安徽理工大学 | Lithium battery SOC estimation method based on online parameter identification |
| CN113138344B (en) * | 2021-04-23 | 2024-03-05 | 安徽理工大学 | SOC estimation method based on fractional order extended Kalman filtering algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114839550B (en) | 2024-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shu et al. | Stage of charge estimation of lithium-ion battery packs based on improved cubature Kalman filter with long short-term memory model | |
| CN111505506A (en) | Battery SOC estimation method integrating multi-scale Kalman filtering and unscented Kalman filtering | |
| CN113030752B (en) | Online parameter identification and SOC joint estimation method based on variable forgetting factor | |
| CN113049962B (en) | LSTM-based energy storage device operation situation deduction method | |
| CN111722118B (en) | Lithium ion battery SOC estimation method based on SOC-OCV optimization curve | |
| CN111060824B (en) | Charge state estimation method based on model and algorithm step-by-step fusion | |
| CN114839550A (en) | UKF-AUKF-based lithium battery SOC joint estimation method | |
| Li et al. | Review of lithium-ion battery state of charge estimation | |
| CN111337832A (en) | Power battery multidimensional fusion SOC and SOH online joint estimation method | |
| CN112858928B (en) | Lithium battery SOC estimation method based on online parameter identification | |
| CN112989690B (en) | Multi-time-scale state-of-charge estimation method for lithium battery of hybrid electric vehicle | |
| CN112269133B (en) | SOC estimation method based on pre-charging circuit model parameter identification | |
| CN114114038A (en) | Lithium battery SOC and available capacity joint estimation method under full-life and full-temperature conditions | |
| CN112946481A (en) | Based on federation H∞Filtering sliding-mode observer lithium ion battery SOC estimation method and battery management system | |
| CN113625182A (en) | Method for on-line estimating battery state | |
| CN116047308A (en) | Lithium battery SOC estimation method based on PID control and DEKF | |
| Zhao et al. | Digital twin-driven estimation of state of charge for Li-ion battery | |
| CN115455842A (en) | Method for estimating state of charge of super capacitor based on variable-temperature model weighted fusion | |
| CN112946480B (en) | Lithium battery circuit model simplification method for improving SOC estimation real-time performance | |
| CN116718920B (en) | Lithium battery SOC estimation method based on RNN (RNN-based optimized extended Kalman filter) | |
| CN116298913A (en) | Lithium battery state of charge estimation method based on online and offline identification combination | |
| CN113420444A (en) | Lithium ion battery SOC estimation method based on parameter online identification | |
| CN116449219A (en) | Lithium battery model parameter identification method for improving VFFRLS | |
| CN114660463A (en) | Lithium battery SOC estimation method based on self-adaptive weighted volume particle filter | |
| CN113960482A (en) | Lithium battery state of charge intelligent prediction method based on improved wolf particle filtering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |