CN113219342B - Method and system for estimating battery SOC (state of charge) by using dimension reduction gas-liquid dynamic battery model - Google Patents
Method and system for estimating battery SOC (state of charge) by using dimension reduction gas-liquid dynamic battery model Download PDFInfo
- Publication number
- CN113219342B CN113219342B CN202110582487.9A CN202110582487A CN113219342B CN 113219342 B CN113219342 B CN 113219342B CN 202110582487 A CN202110582487 A CN 202110582487A CN 113219342 B CN113219342 B CN 113219342B
- Authority
- CN
- China
- Prior art keywords
- battery
- temperature
- soc
- liquid dynamic
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—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]
- G01R31/378—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC] specially adapted for the type of battery or accumulator
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/382—Arrangements for monitoring battery or accumulator variables, e.g. SoC
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/382—Arrangements for monitoring battery or accumulator variables, e.g. SoC
- G01R31/3842—Arrangements for monitoring battery or accumulator variables, e.g. SoC combining voltage and current measurements
Abstract
The invention provides a method and a system for estimating the SOC of a battery by using a dimension-reducing gas-liquid dynamic battery model, which comprises the following steps: estimating the average temperature inside the battery to ensure that the definition of the average temperature inside the battery is consistent with that in the gas-liquid dynamic battery model; reducing the dimension of the gas-liquid dynamic battery model, and reducing the two-dimensional gas-liquid dynamic battery model to one dimension; the one-dimensional gas-liquid dynamic battery model is coupled with the EKF algorithm, so that the calculated amount is reduced, and the robustness is improved; and verifying the SOC estimation precision of the battery. The jacobian matrix is not easy to generate ill-condition during one-dimensional data operation, and the robustness of the algorithm is obviously improved, so that the variance Q of the state equation and the equation R of the observation equation can be set to be larger values during algorithm initialization so as to improve the anti-interference capability and convergence speed of the algorithm, and the method has obvious precision advantage compared with the same type of technology.
Description
Technical Field
The invention relates to the field of battery management systems, in particular to a method and a system for estimating the SOC of a battery by using a dimension reduction gas-liquid dynamic battery model, and particularly relates to an SOC online estimation technology of a lithium ion power battery of an electric vehicle.
Background
In the twenty-first century, human beings have made unprecedented breakthroughs in the fields of science and technology, but face the problems of energy shortage and increasingly serious environmental pollution. The transportation industry accounts for a large proportion of the whole fossil energy consumption, the lithium ion battery technology is continuously broken through, lithium ion battery products are mature day by day, electric energy can be generated by solar energy, wind energy, hydraulic power and other clean modes, the transmission cost is low, the lithium ion battery has wide application in the aspects of social production and life, most countries in the world successively make policies for developing new energy automobiles, and the new energy automobile industry is greatly supported. At present, electric automobiles account for more than eight times of the whole new energy automobiles, and the industry generally considers that electromotion is one of the final development forms of the automobiles. A Battery Management System (BMS) is an indispensable component of an electric vehicle and is also a core problem and a difficult problem in the current research field of electric vehicles. The main functions of the BMS include battery data acquisition and detection, battery state estimation, online fault diagnosis and early warning, battery thermal management, etc., where real-time estimation of state of charge (SOC) is one of the core issues of the BMS.
The SOC of the battery cell reflects the remaining capacity of the battery, and the SOC of the entire battery pack reflects the remaining mileage of the electric vehicle. The accurate SOC can obviously improve the management efficiency and the use safety of the battery, reduce the power consumption of hundreds of kilometers, prolong the service life of the battery and further reduce the comprehensive use cost of the electric automobile. At present, an observer method with a feedback mechanism coupled with a battery mechanism model is a research hotspot of a battery SOC estimation technology, the battery mechanism model usually selected comprises an electrochemical model, an equivalent circuit model and a pneumatic and hydraulic mechanical battery model, and an observer algorithm for designing the feedback mechanism comprises an Extended Kalman Filter (EKF) algorithm, a particle filter algorithm, a synovial membrane algorithm and the like. However, in the research process, the battery model is usually multidimensional (larger than one dimension) and the established SOC estimation method is also multidimensional, so that the estimation method is very complex and difficult to apply in a vehicle-mounted embedded system, and finally, a large number of SOC estimation methods are stopped from theoretical research.
Disclosure of Invention
Aiming at the technical problem, the invention provides an SOC estimation method and system based on a corrected gas-liquid dynamic battery model. The method estimates the average temperature inside the battery through the surface temperature of the battery, and improves the modeling precision of the gas-liquid dynamic battery model; the method for estimating the SOC through the EKF coupling gas-liquid dynamic battery model with the one-dimensional Extended Kalman Filter (EKF) is designed, the one-dimensional EKF coupling gas-liquid dynamic battery model with the feedback mechanism is simplified into the one-dimensional state model, the jacobian matrix is not prone to ill conditions during one-dimensional data operation, and the robustness of the algorithm is remarkably improved.
The invention is realized by the following technical scheme: a method for estimating the SOC of a battery by using a dimension-reduced gas-liquid dynamic battery model comprises the following steps:
step S1, estimating the average temperature inside the battery: estimating the average temperature inside the battery through the surface temperature of the battery, so that the average temperature inside the battery is consistent with the average temperature inside the gas-liquid dynamic battery model;
step S2: dimension reduction gas-liquid dynamic battery model: reducing the two-dimensional gas-liquid dynamic battery model to a one-dimensional gas-liquid dynamic battery model;
and step S3: the one-dimensional gas-liquid dynamic battery model coupling EKF algorithm: coupling the one-dimensional gas-liquid dynamic battery model with a one-dimensional EKF, and inputting the average temperature in the battery obtained in the step S1 into a one-dimensional gas-liquid dynamic battery model coupling EKF algorithm to obtain a battery SOC estimated value;
and step S4: and verifying the SOC estimation precision of the battery.
In the above solution, the step S1 of estimating the average temperature inside the battery includes the following steps:
the method comprises the steps of establishing a battery temperature model, and estimating the average temperature of the battery by adopting an assumed battery temperature model, wherein the first method adopts a temperature gradient calculation method when two or more temperature acquisition points are installed on the surface of the battery, and the second method adopts a Bernadi heat generation model calculation method when only one temperature acquisition point is installed on the surface of the battery.
In the above solution, the first method for calculating a temperature gradient when two or more temperature collection points are installed on a surface of a battery includes the following steps:
a temperature sensor is arranged at the center of the surface formed by the length and the height of the battery, one or more temperature sensors are arranged at other positions of the surface, and the temperature sensor at the center of the battery collects the temperature as the highest center temperature T 0 The temperature collected by the temperature sensors at other positions is T 1 ,T 2 ,T 3 …T n ，n≥1,n∈N + Other position of the temperature sensor to the centerThe distance between the position sensors is d 1 ,d 2 ,d 3 …d n ，n≥1,n∈N + And the battery temperature gradient GradT is calculated by a formula I:
average temperature T of battery aver Calculated by formula two:
wherein r is the pole diameter, θ is the pole angle, V is the cell volume, and b is the cell thickness.
In the above solution, the second calculation method for the Bernadi heat generation model when only one temperature acquisition point is installed on the surface of the battery includes the following steps:
a temperature sensor is arranged at any position of the surface formed by the length and the height of the battery, and the temperature sensor acquires the temperature T 1 The distance from the temperature sensor to the center of the surface is d 1 The heat production rate of the battery is obtained by calculating according to a Bernadi formula III, the temperature gradient GradT is obtained by calculating according to a formula IV, and the highest temperature of the center of the battery is obtained by calculating according to a formula V:
wherein I is current, U L Terminal voltage, V battery volume, T 1 Collecting temperature for a temperature sensor, wherein OCV is open-circuit voltage of a battery;
wherein k is s For heat transfer coefficient, T b R is the temperature of the heat exchange fluid, r is the diameter of the pole, k l Is the thermal conductivity of the battery in the length direction.
T 0 ＝T 1 +d 1 XGradT formula five
Knowing the highest center temperature T of the cell 0 And temperature gradient GradT, average temperature T of the cell aver The formula II is calculated to obtain:
wherein b is the thickness of the battery, T 0 The highest center temperature, θ is the polar angle.
In the above scheme, the dimension reduction gas-liquid dynamic battery model comprises the following steps:
the charge state estimation equation of the gas-liquid dynamic battery model is a formula six, and the terminal voltage estimation equation is a formula seven:
wherein, SOC 1 To pre-sample time state of charge, SOC 2 For the state of charge, Q, at the present sampling moment T Is the rated capacity of the battery, t 1 For pre-sampling time points, t 2 The tau is an integral variable for the current sampling time point;
wherein, U L Terminal voltage, OCV pres Open circuit voltage, OCV, for the current sampling instant init For open circuit voltage, T, at the previous sampling instant aver Is the average temperature of the battery, I is the current, k 1 、k 2 、k 3 、k 4 Parameters of a gas-liquid dynamic battery model are obtained;
there are two state variables OCV in equation seven pres And OCV init The state vector is a two-dimensional column vector when coupled with the EKF, which correlates the OCV init Set as S, and the evaluation is finished by assignment at each timeThe OCV pres Assign S to the state variable OCV init As elements of the input vector, formula seven is rewritten as formula eight:
wherein S is an initial open circuit voltage.
In the above scheme, the EKF algorithm coupled to the one-dimensional gas-liquid dynamic battery model includes the following steps:
in the EKF algorithm, equation nine is a state equation, equation ten is an observation equation, and the discrete equation is:
wherein w is the system noise, and P (w) to N (0, Q), i.e. w obeys 0 expected normal distribution, and Q is the system noise variance; v is the measurement noise, and P (v) -N (0, R), i.e., v obeys 0 the expected normal distribution, R is the measurement noise variance; SOC k Is the kth state of charge value, SOC k-1 Is the k-1 th charge state value, Q T Rated capacity, I, of the battery k For the k-th current sample value, w, of the battery k-1 Is the k-1 th noise, U of the state equation k Estimating terminal voltage, OCV, for battery kth k Is the kth open circuit voltage, T, of the battery aver_k Is the k-th average temperature, S, of the battery k Is the k-th initial open circuit voltage, v k Measuring noise for the kth time of an observation equation, wherein t is sampling time and delta t is time interval;
written as standard EKF format, set:
z k ＝U k ,μ k ＝[I k ,T aver_k ,S k ],
then:
if two or more temperature sensors:
if one temperature sensor:
wherein x is k Is the k-th state vector, x k-1 Is the k-1 st state vector, z k For the k-th observation vector, OCV (x) k ) In order to obtain the k-th open-circuit voltage by looking up the OCV-SOC curve table,
for optimal estimation of the previous sampling instant, A k-1 Is a state equation in
A derivative;
for a priori estimation of the current sampling instant, H k For observing the equation
Derivative of (a), T 0 Maximum temperature at center, V cell volume, k l The thermal conductivity coefficient of the battery in the length direction, b the thickness of the battery and theta the polar angle;
coupling EKF of a one-dimensional gas-liquid dynamic battery model:
initialization:
k,SOC 0 ,S 1 ,Q T ,P 0 ,Q,R,k 1 ,k 2 ,k 3 ,k 4
and (3) cycle estimation:
P k ＝(E-K k H k )P k - equation fifteen
Where k is the sampling instant u k For the kth equation of state input matrix, μ k For the k-th observation equation input matrix, P k - Is the k-th prior covariance matrix, P k Is the K-th posterior covariance matrix, K k Is the k-th Kalman gain, E is the identity matrix with the same dimension as the operation matrix, P 0 As an initial covariance matrix, P k-1 Is the k-1 th a posteriori covariance matrix,
for the k-1 st state coefficient matrix transposition,
for the observation equation in
Transposing the derivative matrix of (A), H k For the observation equation in
The matrix of the derivatives of (a) and (b),
and performing optimal estimation on the state vector at the k time.
In the above scheme, in the estimating of the average temperature inside the battery, the battery is a square battery, a battery temperature model is established, the length of the battery is l, the height of the battery is h, and the thickness of the battery is b, the thickness of the square battery is smaller than the length and the height of the battery, and the heat transfer coefficient in the thickness direction is also smaller than the length and the height of the battery.
In the scheme, the SOC estimation accuracy of the battery is verified through the actual measurement working condition of FUDS.
A system for realizing the method for estimating the SOC of the battery by using the dimension-reduced gas-liquid dynamic battery model comprises a signal acquisition module, an average temperature estimation module, an SOC estimation module, a data storage module and a display module;
the signal acquisition module is used for acquiring the terminal voltage, the surface temperature and the current of the battery, is respectively connected with the average temperature estimation module and the SOC estimation module, and transmits the acquired current, surface temperature and terminal voltage signals to the average temperature estimation module and the SOC estimation module; the average temperature estimation module is used for estimating the average temperature inside the battery so that the average temperature inside the battery is consistent with the average temperature inside the gas-liquid dynamic battery model; the SOC estimation module is used for estimating the SOC of the battery by using a one-dimensional gas-liquid dynamic battery model coupled EKF algorithm according to the current and terminal voltage data acquired by the signal acquisition module and the average temperature estimated by the average temperature estimation module; the SOC estimation module is connected with the display module and sends the battery acquisition data, the estimated average temperature and the SOC estimation value to the display module.
In the above scheme, the signal acquisition module includes voltage sensor, temperature sensor and current sensor, and voltage sensor is used for gathering the terminal voltage of battery, and temperature sensor is used for gathering the surface temperature of battery, and current sensor is used for gathering the electric current of battery. Compared with the prior art, the invention has the beneficial effects that: according to the invention, the modeling precision of the gas-liquid dynamic battery model is improved by the fact that the calculated internal average temperature of the battery is consistent with the internal average temperature defined by the gas-liquid dynamic battery model. The gas-liquid dynamic battery model is reduced to the one-dimensional model, all data operations are one-dimensional without matrix operations when the gas-liquid dynamic battery model is coupled with the EKF algorithm, and the calculation amount is greatly reduced, so that the gas-liquid dynamic battery model can be directly applied to a vehicle-mounted embedded system, and the high industrial application value is shown. The jacobian matrix is not easy to generate ill-conditioned (irregular) condition during one-dimensional data operation, and the robustness of the algorithm is obviously improved, so that the variance Q of the state equation and the equation R of the observation equation can be set to be larger values during the initialization of the algorithm so as to improve the anti-interference capability and the convergence speed of the algorithm, and the method has obvious precision advantage compared with the same type of technology.
Drawings
FIG. 1: is an implementation flow diagram of one embodiment of the present invention;
FIG. 2: is a selected square lithium ion battery of an embodiment of the invention;
FIG. 3: the estimation result of the SOC under the FUDS working condition is obtained according to the embodiment of the invention;
FIG. 4: the local development map of the SOC estimation result of the FUDS working condition is an embodiment of the invention;
FIG. 5: is a system component of an embodiment of the present invention;
FIG. 6: an OCV-SOC curve according to an embodiment of the present invention;
FIG. 7: is an embodiment of the present invention
Curve line.
Detailed Description
The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.
Fig. 1 shows a technical route of the method for estimating the SOC of the battery by using the dimension-reduced gas-liquid dynamic battery model, and the method for estimating the SOC of the battery by using the dimension-reduced gas-liquid dynamic battery model comprises the following steps:
the method comprises the following steps: estimating the average temperature inside the battery to ensure that the definition of the average temperature inside the battery is consistent with that in the gas-liquid dynamic battery model, and improving the modeling precision of the gas-liquid dynamic battery model;
step two: reducing the dimension of the gas-liquid dynamic battery model, and reducing the two-dimensional gas-liquid dynamic battery model to one dimension;
step three: the one-dimensional gas-liquid dynamic battery model is coupled with the extended Kalman filter EKF algorithm, so that the calculated amount is reduced, and the robustness is improved;
step four: verifying the SOC estimation precision of the battery, and verifying the SOC estimation precision of the battery through the actual measurement working condition of FUDS;
the first step of estimating the average temperature inside the battery specifically comprises the following steps:
firstly, in order to realize application in a vehicle-mounted embedded system, an assumption needs to be made on a battery temperature model to achieve the purpose of reducing the calculation amount; according to this embodiment, preferably, taking a square packaged lithium ion battery adopting a laminated cell as an example, a battery temperature model is established, where the length of the battery is l, the height of the battery is h, and the thickness of the battery is b, the thickness of the square battery is usually much smaller than the length and the height, and the heat transfer coefficient in the thickness direction is also much smaller than the length and the height (this is the present embodiment)In the embodiment, the measured physical parameters of the selected square battery are that the thermal conductivity coefficients in the length direction, the height direction and the thickness direction are respectively k l ＝29.9W/(m·K)、k h ＝29.9W/(m·K)、k d = 1.03W/(m · K); length, height and thickness dimensions l =122mm, h =88mm, b =9mm, respectively), it can be assumed that the temperature distribution of the battery is uniform in the thickness direction; establishing a polar coordinate system of the battery by taking the center of the battery as a polar coordinate origin and taking a plane formed by the length direction and the height direction as a polar coordinate plane; the method comprises the following steps of adopting an assumed battery temperature model to estimate the average temperature of the battery, adopting a temperature gradient calculation method when two or more temperature acquisition points are installed on the surface of the battery in the first method, and adopting a Bernadi heat generation model calculation method when only one temperature acquisition point is installed on the surface of the battery in the second method;
the first method for calculating the temperature gradient when two or more temperature collection points are installed on the surface of a battery comprises the following steps:
a temperature sensor is arranged at the center of the surface formed by the length and the height of the battery, one or more temperature sensors are arranged at other positions of the surface, and the temperature sensor at the center of the battery collects the temperature as the highest center temperature T 0 The temperature collected by the temperature sensors at other positions is T 1 ,T 2 ,T 3 …T n ，n≥1,n∈N + The distances from the temperature sensors at other positions to the sensor at the central position are respectively d 1 ,d 2 ,d 3 …d n ，n≥1,n∈N + And the battery temperature gradient GradT is obtained by calculating the formula I:
average temperature T of battery aver Calculated by formula two:
wherein r is the pole diameter, θ is the pole angle, V is the cell volume, and b is the cell thickness.
The second calculation method for the Bernadi heat generation model when only one temperature acquisition point is installed on the surface of the battery comprises the following steps:
a temperature sensor is arranged at any position of the surface formed by the length and the height of the battery, and the temperature sensor acquires the temperature T 1 The distance from the temperature sensor to the center of the surface is d 1 The heat production rate of the battery is obtained by calculating according to a Bernadi formula III, the temperature gradient GradT is obtained by calculating according to a formula IV, and the highest temperature of the center of the battery is obtained by calculating according to a formula V:
wherein I is current, U L The voltage is terminal voltage, V is battery volume, T is temperature acquired by a temperature sensor, and OCV is battery open-circuit voltage;
wherein k is s For heat transfer coefficient, T b R is the temperature of the heat exchange fluid, r is the diameter of the pole, k l Is the thermal conductivity of the battery in the length direction.
T 0 ＝T 1 +d 1 XGradT formula five
Knowing the highest center temperature T of the cell 0 And temperature gradient GradT, average temperature T of the cell aver The formula II is calculated to obtain:
wherein b is the thickness of the battery, T 0 The center maximum temperature, θ is the polar angle.
The step two dimension reduction gas-liquid dynamic battery model comprises the following steps:
the charge state estimation equation (formula six) and the terminal voltage estimation equation (formula seven) of the gas-liquid dynamic battery model are as follows:
here, SOC 1 To the state of charge, SOC, at the previous sampling time 2 For the state of charge, Q, at the current sampling instant T Is the rated capacity of the battery, t 1 For pre-sampling time points, t 2 Tau is an integral variable for the current sampling time point;
here, U L Terminal voltage, OCV pres Open circuit voltage, OCV, for the current sampling instant init For open circuit voltage, T, at the previous sampling instant aver Is the average temperature of the battery, I is the current, k 1 、k 2 、k 3 、k 4 Parameters of a gas-liquid dynamic battery model are obtained;
there are two state variables OCV in equation seven pres And OCV init The state vector is a two-dimensional column vector when coupled with the EKF if the OCV is considered init Set as S, the OCV is assigned at the end of each estimation pres Assigned to S, the state variable OCV can be set init As an element of the input vector, the formula seven can be rewritten as the formula eight:
wherein S is an initial open circuit voltage.
The three-dimensional gas-liquid dynamic battery model coupling EKF algorithm comprises the following steps:
in the EKF algorithm, equation nine is a state equation, equation ten is an observation equation, and the discrete equation is:
here, w is the system noise, and P (w) to N (0, Q), i.e., w follows an expected normal distribution of 0, Q is the system noise variance; v is the measurement noise and P (v) to N (0, R), i.e. v obeys 0 the expected normal distribution, R is the measurement noise variance; the subscripts of all the letters contain k or k-1 to respectively represent data at the k-th sampling moment or data at the k-1 sampling moment; SOC k Is the kth state of charge value, SOC k-1 Is the k-1 th charge state value, Q T Rated capacity, I, for the battery k For the k-th current sample value, w, of the battery k-1 Is the k-1 th noise, U of the state equation k Estimating terminal voltage, OCV, for battery kth k Is the kth open circuit voltage, T, of the battery aver_k Is the k-th average temperature, S, of the battery k Is the kth initial open circuit voltage, v k Measuring noise for the kth time of an observation equation, wherein t is sampling time, and delta t is time interval;
written as standard EKF format, set:
z k ＝U k ,μ k ＝[I k ,T aver_k ,S k ],
then:
if two or more temperature sensors:
if there is one temperature sensor:
here, x k Is the kth state vector, x k-1 Is the k-1 th state vector, z k For the k-th observation vector, OCV (x) k ) In order to obtain the k-th open-circuit voltage by looking up the OCV-SOC curve table,
for optimal estimation of the previous sampling instant, A k-1 Is a state equation in
A derivative;
for a priori estimation of the current sampling instant, H k For observing the equation
Derivative of (a), T 0 The highest temperature at the center, V the cell volume, k l The thermal conductivity coefficient of the battery in the length direction, b the thickness of the battery and theta the polar angle;
the coupling EKF of the one-dimensional gas-liquid dynamic battery model comprises the following steps:
initialization:
k,SOC 0 ,S 1 ,Q T ,P 0 ,Q,R,k 1 ,k 2 ,k 3 ,k 4
and (3) cycle estimation:
P k ＝(E-K k H k )P k - equation fifteen
Where k is the sampling instant u k For the kth equation of state input matrix, μ k For the k-th observation equation input matrix, P k - Is the k-th prior covariance matrix, P k Is the K-th posterior covariance matrix, K k Is the k-th Kalman gain, E is the identity matrix with the same dimension as the operation matrix, P 0 As an initial covariance matrix, P k-1 Is the k-1 th a-posteriori covariance matrix,
for the k-1 st state coefficient matrix transposition,
for the observation equation in
Transposing of the derivative matrix of (H) k For the observation equation in
The matrix of the derivatives of (a) is,
and (4) optimally estimating the state vector at the k time.
Wherein, the OCV-SOC curve of the selected square battery is shown in FIG. 6,
the curves are shown in FIG. 7;
the step four of verifying the SOC estimation precision of the battery specifically comprises the following steps:
the accuracy of the SOC estimation method is verified by selecting the FUDS actual measurement working condition data of a square battery.
The specific embodiment is as follows:
method for estimating battery SOC by using dimension-reduced gas-liquid dynamic battery model, and parameter (k) of gas-liquid dynamic battery model 1 ＝0.4318,k 2 ＝0.0343,k 3 ＝0.00277,k 4 = 0.000014) identification method, refer to patent publication No. 201910137591.x.
Initialization: k =1,SOC 0 ＝100，S 1 ＝4.193，Q T ＝5.53，P 0 ＝1(P 0 ∈R + ，R + Positive and real), Q =20 (Q ∈ R) + ，R + Positive real number), R =1 (R ∈ R) + ，R + Positive real number), k 1 ＝0.4318,k 2 ＝0.0343,k 3 ＝0.00277,k 4 =0.000014; and (3) cycle estimation: reading the collected terminal voltage z as shown in Table 1 k ＝4.190，I k ＝-3.428，T k =297.45, estimated average temperature is 297.50; calculating out
Computing
Calculating K k =0.326, calculation
Calculating P k K is added by 1 for 20.893, and the next time is assignedAnd (4) repeating the steps in a circulating way to complete the online estimation of the SOC of the battery.
TABLE 1 sample data and estimation results
The accuracy of the SOC estimation method is verified through FUDS actual measurement working condition data, the estimation result is shown in fig. 3 and fig. 4, fig. 3 shows an SOC curve (black cross curve) obtained through experimental tests, an SOC curve (light gray curve) estimated by a model under the condition of no initial error, an SOC curve (gray curve) estimated by the model under the condition of 50% initial error and an SOC curve (dark gray curve) estimated by an ampere-hour integration method under the condition of 50% initial error, and the maximum estimation error of the SOC of the model under the condition of no initial error is 1.28%; FIG. 4 is an enlarged view of the first 60 second curve of FIG. 3, where the model substantially eliminates the initial error after about 25 seconds after 50% of the initial error is input, and the curve coincides with the estimated curve without the initial error, but the ampere-hour integration method is not always capable of eliminating the initial error. The jacobian matrix is not easy to generate ill-condition during one-dimensional data operation, and the robustness of the algorithm is obviously improved, so that the variance Q of the state equation and the equation R of the observation equation can be set to be larger values during the initialization of the algorithm, so that the anti-interference capability and the convergence speed of the algorithm are improved. The maximum SOC estimation error is 1.28% under the actual measurement FUDS working condition, and compared with the same type of technology, the method has the obvious precision advantage.
A system for realizing the method for estimating the SOC of the battery by using the dimension-reduced gas-liquid dynamic battery model comprises a signal acquisition module, an average temperature estimation module, an SOC estimation module, a data storage module and a display module, wherein the signal acquisition module, the average temperature estimation module, the SOC estimation module, the data storage module and the display module are shown in figure 5;
the signal acquisition module comprises a voltage sensor, a temperature sensor and a current sensor which are respectively used for acquiring the terminal voltage, the surface temperature and the current of the battery, and the signal acquisition module and the average temperature estimation module are connected with the SOC estimation module and transmit the acquired current, surface temperature and terminal voltage signals to the average temperature estimation module and the SOC estimation module; the average temperature estimation module estimates the internal average temperature of the battery by using the acquired battery surface temperature and a Bernadi heat generation rate formula, so that the internal average temperature of the battery is consistent with the internal average temperature of the gas-liquid dynamic battery model, and the modeling precision of the gas-liquid dynamic battery model is improved; the SOC estimation module is used for estimating the SOC of the battery by using a one-dimensional gas-liquid dynamic battery model coupled EKF algorithm according to the current and terminal voltage data acquired by the signal acquisition module and the average temperature estimated by the average temperature estimation module; the SOC estimation module is connected with the display module, and is used for sending the data collected by the battery, the estimated average temperature and the SOC estimation value to the display module for the reference of a user and storing the data in the data storage module.
The average temperature estimation module and the SOC estimation module are realized by a single chip microcomputer, and the single chip microcomputer is preferably a Shekal automobile grade single chip microcomputer. The method for estimating the SOC of the battery by using the dimension reduction gas-liquid dynamic battery model is realized on hardware, and codes written by C language on a singlechip can be realized on a CodeWarrior Development Studio Development platform.
According to this embodiment, preferably, the SOC estimation module is specifically:
firstly, loading a library function file of a singlechip, configuring a singlechip register by using the library function, and compiling a clock function, a timer function, a delay function, a storage function, a data verification function, an average temperature estimation function, a main function and the like;
(1) connecting a current sensor and a temperature sensor to a signal acquisition card, wherein the acquisition card can directly acquire the voltage of a single battery, and preferably, the voltage range of the single battery is within 0-5V;
(2) the acquisition card is connected with a serial port of the singlechip, RS-232 is selected in a communication mode, and signals of current, voltage and surface temperature of the battery are transmitted to the singlechip;
(3) a main function of the single chip microcomputer reads current, voltage and surface temperature signals of the battery, and an average temperature estimation function is called to calculate an average temperature value under current input; the master function estimates the SOC value at the current moment through the battery current, the terminal voltage, the average temperature value and the initialization data of the last sampling moment, and sends the battery current, the terminal voltage, the average temperature and the calculated SOC value to a display module of the upper computer for display;
(4) and (4) circulating the steps (1) to (3) in such a way, and finishing the real-time SOC estimation of the battery pack.
The upper computer is developed based on a Microsoft Visual Studio platform and is used for displaying the terminal voltage and the SOC of the battery pack, the SOC of all the series single batteries and the fitted lowest SOC of the batteries;
the singlechip comprises: 2 n A single chip machine, n =1,2,3, and arithmetic units of various ARM kernels;
the signal communication protocol used includes: RS-485, CAN, TCP, modbus, MPI, serial port communication and the like.
It should be understood that although the specification has been described in terms of various embodiments, not every embodiment includes every single embodiment, and such description is for clarity purposes only, and it will be appreciated by those skilled in the art that the specification as a whole can be combined as appropriate to form additional embodiments as will be apparent to those skilled in the art.
The above-listed detailed description is only a specific description of a possible embodiment of the present invention, and they are not intended to limit the scope of the present invention, and equivalent embodiments or modifications made without departing from the technical spirit of the present invention should be included in the scope of the present invention.
Claims (8)
1. A method for estimating the SOC of a battery by using a dimension-reduced gas-liquid dynamic battery model is characterized by comprising the following steps:
step S1, estimating the average temperature inside the battery: estimating the average temperature inside the battery through the surface temperature of the battery, so that the average temperature inside the battery is consistent with the average temperature inside the gas-liquid dynamic battery model;
step S2: dimension reduction gas-liquid dynamic battery model: reducing a two-dimensional gas-liquid dynamic battery model to a one-dimensional gas-liquid dynamic battery model, wherein the dimension reduction gas-liquid dynamic battery model comprises the following steps:
the charge state estimation equation of the gas-liquid dynamic battery model is a formula six, and the terminal voltage estimation equation is a formula seven:
wherein, SOC 1 To pre-sample time state of charge, SOC 2 For the state of charge, Q, at the current sampling instant T Rated capacity of the battery, t 1 For pre-sampling time points, t 2 Tau is an integral variable for the current sampling time point;
wherein, U L Terminal voltage, OCV pres Open circuit voltage, OCV, for the current sampling instant init For open circuit voltage, T, at the previous sampling instant aver Is the average temperature of the battery, I is the current, k 1 、k 2 、k 3 、k 4 Parameters of a gas-liquid dynamic battery model are obtained;
there are two state variables OCV in equation seven pres And OCV init The state vector is a two-dimensional column vector when coupled with the EKF, which correlates the OCV init Set as S, the OCV is assigned by means of value at the end of each estimation pres Assign S to the state variable OCV init As an element of the input vector, formula seven is rewritten as formula eight:
wherein S is an initial open circuit voltage;
and step S3: the one-dimensional gas-liquid dynamic battery model coupling EKF algorithm: coupling a one-dimensional gas-liquid dynamic battery model with a one-dimensional EKF, inputting the average temperature in the battery obtained in the step S1 into a one-dimensional gas-liquid dynamic battery model coupling EKF algorithm to obtain a battery SOC estimated value, wherein the one-dimensional gas-liquid dynamic battery model coupling EKF algorithm comprises the following steps:
in the EKF algorithm, equation nine is a state equation, equation ten is an observation equation, and the discrete equation is:
wherein w is the system noise, and P (w) to N (0, Q), i.e. w obeys 0 expected normal distribution, and Q is the system noise variance; v is the measurement noise and P (v) to N (0, R), i.e. v obeys 0 the expected normal distribution, R is the measurement noise variance; SOC k Is the kth state of charge value, SOC k-1 Is the k-1 th charge state value, Q T Rated capacity, I, for the battery k For the k-th current sample value, w, of the battery k-1 Is the k-1 th noise, U of the state equation k Estimating terminal voltage, OCV, for battery kth k Is the kth open circuit voltage, T, of the battery aver_k Is the k-th average temperature, S, of the battery k Is the k-th initial open circuit voltage, v k Measuring noise for the kth time of an observation equation, wherein t is sampling time, and delta t is time interval;
written as standard EKF format, set:
x k ＝SOC k ,u k ＝I k ,
z k ＝U k ,μ k ＝[I k ,T aver_k ,S k ],
then:
if two or more temperature sensors:
if there is one temperature sensor:
T 0 ＝T 1 +d 1 ×GradT
wherein x is k Is the k-th state vector, x k-1 Is the k-1 st state vector, z k For the k-th observation vector, OCV (x) k ) In order to obtain the k-th open-circuit voltage by looking up the OCV-SOC curve table,
for optimal estimation of the previous sampling instant, A k-1 Is a state equation in
A derivative;
for a priori estimation of the current sampling instant, H k For observing the equation
Derivative of (a), T 0 Maximum temperature at center, V cell volume, k l The thermal conductivity coefficient of the battery in the length direction, b the thickness of the battery and theta the polar angle; r is the diameter of the pole;
coupling EKF of a one-dimensional gas-liquid dynamic battery model:
initialization:
k,SOC 0 ,S 1 ,Q T ,P 0 ,Q,R,k 1 ,k 2 ,k 3 ,k 4
and (3) cycle estimation:
where k is the sampling instant u k For the kth State equation input matrix, μ k For the k-th observation equation input matrix, P k - Is the k-th prior covariance matrix, P k Is the K-th posterior covariance matrix, K k Is the k-th Kalman gain, E is the identity matrix with the same dimension as the operation matrix, P 0 As an initial covariance matrix, P k-1 Is the k-1 th a-posteriori covariance matrix,
for the k-1 st state coefficient matrix transposition,
for observing the equation
Transposing of the derivative matrix of (H) k For observing the equation
The matrix of the derivatives of (a) is,
performing optimal estimation on the k-th state vector;
and step S4: and verifying the SOC estimation precision of the battery.
2. The method for estimating the SOC of the battery using the dimension-reduced gas-liquid dynamic battery model according to claim 1, wherein the step S1 of estimating the average temperature inside the battery comprises the steps of:
the method comprises the steps of establishing a battery temperature model, and estimating the average temperature of the battery by adopting the battery temperature model based on hypothesis, wherein the first method adopts a temperature gradient calculation method when two or more temperature acquisition points are installed on the surface of the battery, and the second method adopts a Bernadi heat generation model calculation method when only one temperature acquisition point is installed on the surface of the battery.
3. The method for estimating the SOC of the battery by using the dimension-reduced gas-liquid dynamic battery model according to claim 2, wherein the first method for calculating the temperature gradient when two or more temperature collection points are installed on the surface of the battery comprises the following steps:
a temperature sensor is arranged at the center of the surface formed by the length and the height of the battery, one or more temperature sensors are arranged at other positions of the surface, and the temperature sensor at the center of the battery collects the temperature as the centerHigh temperature T 0 The temperature collected by the temperature sensors at other positions is T 1 ,T 2 ,T 3 …T n ，n≥1,n∈N + The distances from the temperature sensors at other positions to the sensor at the central position are respectively d 1 ,d 2 ,d 3 …d n ，n≥1,n∈N + And the battery temperature gradient GradT is obtained by calculating the formula I:
average temperature T of battery aver Calculated by formula two:
wherein r is the pole diameter, theta is the pole angle, V is the cell volume, and b is the cell thickness.
4. The method for estimating the SOC of the battery by using the dimension-reduced gas-liquid dynamic battery model as claimed in claim 2, wherein the second calculation method for the Bernadi heat generation model when only one temperature collection point is installed on the surface of the battery comprises the following steps:
a temperature sensor is arranged at any position of the surface formed by the length and the height of the battery, and the temperature sensor acquires the temperature T 1 The distance from the temperature sensor to the center of the surface is d 1 The heat production rate of the battery is obtained by calculating according to a Bernadi formula III, the temperature gradient GradT is obtained by calculating according to a formula IV, and the highest temperature of the center of the battery is obtained by calculating according to a formula V:
wherein I is current, U L Terminal voltage, V battery volume, T 1 Collecting temperature for a temperature sensor, wherein OCV is open-circuit voltage of a battery;
wherein k is s Is a heat transfer coefficient, T b R is the temperature of the heat transfer fluid, r is the diameter of the pole, k l The thermal conductivity coefficient of the battery in the length direction;
T 0 ＝T 1 +d 1 XGradT formula five
Knowing the center maximum temperature T of the cell 0 And temperature gradient GradT, average temperature T of the cell aver The formula II is calculated to obtain:
wherein b is the thickness of the battery, T 0 The center maximum temperature, θ is the polar angle.
5. The method according to claim 2, wherein in the step of estimating the average temperature inside the battery, the battery is a square battery, the battery temperature model is established, the battery length is l, the height is h, the thickness is b, the thickness of the square battery is smaller than the length and the height, the heat transfer coefficient in the thickness direction is smaller than the length and the height, and a battery polar coordinate system is established by taking the center of the battery as the origin of polar coordinates and the plane formed by the length direction and the height direction as the polar coordinate plane, assuming that the temperature of the battery in the thickness direction is uniformly distributed.
6. The method for estimating the SOC of the battery by using the dimension-reduced gas-liquid dynamic battery model according to claim 1, wherein the accuracy of estimating the SOC of the battery is verified by actually measuring the working condition of FUDS.
7. A system for realizing the method for estimating the SOC of the battery by using the dimension-reduced gas-liquid dynamic battery model according to any one of claims 1-6 is characterized by comprising a signal acquisition module, an average temperature estimation module, an SOC estimation module, a data storage module and a display module;
the signal acquisition module is used for acquiring the terminal voltage, the surface temperature and the current of the battery, is respectively connected with the average temperature estimation module and the SOC estimation module, and transmits the acquired current, surface temperature and terminal voltage signals to the average temperature estimation module and the SOC estimation module; the average temperature estimation module is used for estimating the average temperature inside the battery so that the average temperature inside the battery is consistent with the average temperature inside the gas-liquid dynamic battery model; the SOC estimation module is used for estimating the SOC of the battery by using a one-dimensional gas-liquid dynamic battery model coupled EKF algorithm according to the current and terminal voltage data acquired by the signal acquisition module and the average temperature estimated by the average temperature estimation module; the SOC estimation module is connected with the display module and sends the battery acquisition data, the estimated average temperature and the SOC estimation value to the display module.
8. The system for implementing the method for estimating the SOC of the battery using the dimension-reduced gas-liquid dynamic battery model as claimed in claim 7, wherein the signal acquisition module comprises a voltage sensor, a temperature sensor and a current sensor, the voltage sensor is used for acquiring the terminal voltage of the battery, the temperature sensor is used for acquiring the surface temperature of the battery, and the current sensor is used for acquiring the current of the battery.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110582487.9A CN113219342B (en) | 2021-05-27 | 2021-05-27 | Method and system for estimating battery SOC (state of charge) by using dimension reduction gas-liquid dynamic battery model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110582487.9A CN113219342B (en) | 2021-05-27 | 2021-05-27 | Method and system for estimating battery SOC (state of charge) by using dimension reduction gas-liquid dynamic battery model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113219342A CN113219342A (en) | 2021-08-06 |
| CN113219342B true CN113219342B (en) | 2023-01-17 |
Family
ID=77098766
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110582487.9A Active CN113219342B (en) | 2021-05-27 | 2021-05-27 | Method and system for estimating battery SOC (state of charge) by using dimension reduction gas-liquid dynamic battery model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113219342B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023144623A1 (en) * | 2022-01-27 | 2023-08-03 | 日産自動車株式会社 | Battery module and charge/discharge control device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111693877A (en) * | 2020-05-15 | 2020-09-22 | 江苏大学 | SOC (System on chip) online estimation method and system of lithium ion battery |
| CN112733466A (en) * | 2020-12-18 | 2021-04-30 | 江苏大学 | SOC estimation method and system based on corrected gas-liquid dynamic battery model |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108169682B (en) * | 2017-12-14 | 2020-02-21 | 江苏大学 | Lithium ion battery SOC accurate estimation method based on gas-liquid dynamic model |
| CN109884528B (en) * | 2019-02-25 | 2020-03-31 | 江苏大学 | Lithium ion battery open-circuit voltage estimation method and system with temperature input |
-
2021
- 2021-05-27 CN CN202110582487.9A patent/CN113219342B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111693877A (en) * | 2020-05-15 | 2020-09-22 | 江苏大学 | SOC (System on chip) online estimation method and system of lithium ion battery |
| CN112733466A (en) * | 2020-12-18 | 2021-04-30 | 江苏大学 | SOC estimation method and system based on corrected gas-liquid dynamic battery model |
Non-Patent Citations (4)
| Title |
|---|
| A new gas–liquid dynamics model towards robust state of charge estimation of lithium-ion batteries;Biao Chen 等;《Journal of Energy Storage》;20200402;1-10页 * |
| Online State-of-Charge Estimation Based on the Gas–Liquid Dynamics Model for Li(NiMnCo)O2 Battery;Haobin Jiang 等;《energies》;20210108;1-19页 * |
| State-of-charge estimation of lithium-ion batteries from a gas-liquid dynamics model including the direct temperature input;Haobin Jiang 等;《Journal of Energy Storage》;20210508;1-11页 * |
| 基于气液动力学模型的锂离子电池SOC估算研究;陈彪;《中国优秀硕士论文全文数据库 工程科技II辑》;20190215(第02期);1-62页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113219342A (en) | 2021-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100492751C (en) | Method for estimating nickel-hydrogen power battery charged state based on standard battery model | |
| CN108333528A (en) | SOC and SOT united state methods of estimation based on power battery electric-thermal coupling model | |
| CN108919137B (en) | A kind of battery aging status estimation method considering different battery status | |
| CN101629992B (en) | Method for estimating residual capacity of iron-lithium phosphate power cell | |
| CN107390127A (en) | A kind of SOC estimation method | |
| CN105425154B (en) | A kind of method of the state-of-charge for the power battery pack for estimating electric automobile | |
| CN107589379A (en) | A kind of On-line Estimation lithium battery SOC and the method for impedance | |
| CN105717460A (en) | Power battery SOC estimation method and system based on nonlinear observer | |
| CN102169168B (en) | Battery dump energy estimation method based on particle filtering | |
| CN103744026A (en) | Storage battery state of charge estimation method based on self-adaptive unscented Kalman filtering | |
| CN105572596B (en) | Lithium battery SOC estimation method and system | |
| CN112331941A (en) | Cloud auxiliary battery management system and method | |
| CN109856556A (en) | A kind of power battery SOC estimation method | |
| CN111948546A (en) | Lithium battery health degree evaluation method and system | |
| CN110554321B (en) | Method for detecting SOC (state of charge) of retired power battery in real time | |
| CN101625397A (en) | Mixed rapid estimation method for residual energy of battery | |
| CN106772104A (en) | A kind of electrokinetic cell SOC value evaluation method | |
| CN105631163A (en) | Electric vehicle power battery energy consumption hardware online simulation method and device | |
| CN106998086A (en) | MW class energy-accumulating power station battery management method and its system | |
| CN110456279A (en) | A kind of power battery cloud management system based on data-driven model | |
| CN111806446A (en) | Driving range evaluation method and system for fuel cell hybrid electric vehicle | |
| CN110146823A (en) | A kind of series-connected cell group SOC estimation method based on information fusion | |
| CN106249158A (en) | Ferric phosphate lithium cell actual active volume detection method, system and electric automobile | |
| CN113219342B (en) | Method and system for estimating battery SOC (state of charge) by using dimension reduction gas-liquid dynamic battery model | |
| CN115616428A (en) | Charging-detecting integrated electric vehicle battery state detection and evaluation method |
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 |