A joint estimation method for battery SOC, SOH, and SOE

Abstract

The present invention discloses a method for jointly estimating the state of charge (SOC), state of exhaustion (SOH), and state of exhaustion (SOE) of a battery, comprising the following steps: step S1) reading an initial value of a battery state and an initial value of a capacity; step S2) establishing a second-order RC equivalent circuit model of the battery according to battery characteristics; step S3) collecting data through high-performance computer program processing (HPPC) and identifying the data using a least squares method to obtain model parameters; step S4) establishing a multi-scale calculation formula for jointly estimating the battery state of charge (SOC), state of exhaustion (SOH), and state of exhaustion (SOE) using an extended kinematic factor (EKF) algorithm based on the initial value of the state, the initial value of the capacity, and the model parameters; and step S5) calculating the state of charge (SOC) and state of exhaustion (SOE) according to the real-time voltage, the real-time current, and the initial calculated value of the capacity. When time T is reached, the state of exhaustion (SOH) is calculated and then synchronously updated to the calculation of the state of charge and the state of exhaustion (SOC). This solution can ensure the real-time performance and accuracy of the calculation while reducing the calculation complexity.

Description

Technical Field

[0001] The present invention relates to the technical field of battery management, and in particular to a method for jointly estimating battery SOC, SOH, and SOE. Background Art

[0002] As the market share of new energy vehicles increases year by year and the scale of energy storage expands, the market has increasingly higher requirements for battery management systems. SOC (State of Charge), SOH (State of Health), and SOE (State of Energy) are key technologies in battery management systems. Accurate SOC, SOH, and SOE estimation can effectively avoid adverse phenomena such as battery overcharging, over-discharging, and rapid decay of cycle life. At present, the main SOC and SOE estimation methods are similar, namely the ampere-hour integral method, Kalman filter method, and neural network method; SOH estimation methods mainly include the cycle capacity method, model method, Kalman filter algorithm, etc. Due to the high correlation between SOC and SOP and SOE, high real-time performance is required, and the SOC estimation cycle is generally within 100ms. SOH is a long-term parameter with low real-time requirements, and the estimation cycle can be carried out on an hourly or daily basis. The ampere-hour integration method is simple but is significantly affected by the accuracy of the current integrator and prone to cumulative errors. The Kalman filter method is suitable for linear systems. Since battery SOC and SOH estimation are nonlinear systems, the Extended Kalman Filter (EKF) and Unscended Kalman Filter (UKF) are used based on the Kalman filter principle for nonlinear system estimation. The neural network method requires a large number of samples for training, is relatively complex, and is significantly affected by battery aging. A dual Kalman filter can be used for joint SOC and SOH estimation, but the dual Kalman filter calculation consumes a large amount of computer memory and has certain hardware requirements. Summary of the Invention

[0003] The present invention is mainly to solve the problems of low calculation accuracy and complex methods of battery SOC, SOH and SOE in the existing technology, and provides a battery SOC, SOH and SOE joint estimation method, which adopts the EKF algorithm for SOC, SOH and SOE joint estimation, which not only ensures the accuracy and real-time nature of the calculation, but also reduces the calculation complexity.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] A method for jointly estimating the state of charge (SOC), state of exhaust (SOH), and state of exhaust gas (SOE) of a battery comprises the following steps: step S1) reading an initial state value and an initial capacity value of the battery; step S2) establishing a second-order RC equivalent circuit model of the battery according to battery characteristics; step S3) collecting data through a hybrid pulse power performance test (HPPC) and identifying the data using a least squares method to obtain model parameters; step S4) establishing a multi-scale calculation formula for jointly estimating the state of charge (SOC), state of exhaust gas (SOH), and state of exhaust gas (SOE) of the battery according to the initial state value, the initial capacity value, and the model parameters through an extended Kalman filter (EKF); and step S5) first running an SOH calculation module to obtain an initial capacity calculation value, collecting real-time voltage and real-time current, and inputting the real-time voltage, real-time current, and initial capacity calculation value data into an SOC and SOE calculation module to calculate and obtain an SOC value and an SOE value. Simultaneously, an observation value updated in the SOC calculation is input into the SOH calculation module for calculation to obtain an SOH value. In order to improve the accuracy of SOC, SOH, and SOE calculations in battery management and reduce the complexity of calculations, the present invention provides a battery SOC, SOH, and SOE joint estimation method. First, the initial value of the battery state and the initial value of the capacity are read, and a battery second-order RC equivalent circuit model is established according to the battery characteristics. The OCV-SOC and OCV-SOE relationship, that is, the battery SOC and SOE estimation observation equations, are determined through testing. The model parameters, including battery internal resistance, battery polarization resistance, battery polarization capacitance, battery concentration difference resistance, and battery concentration difference capacitance, are obtained through online or offline identification. Then, according to the estimated state Equation and estimated observation equation, through the extended Kalman filter (EKF) to establish the battery SOC, SOH, SOE joint estimation multi-scale calculation formula; then first run the SOH calculation module to obtain the initial calculated value of the capacity, collect real-time voltage and real-time current, the real-time voltage, real-time current and initial calculated value of the capacity data are input into the SOC and SOE calculation modules, the SOC value and SOE value are calculated, when the SOH calculation is performed, the updated observation value in the SOC calculation is passed to the SOH calculation module to participate in the operation, and then the SOH value is calculated, the SOE operation cycle is set to 0.1s, and the SOH operation cycle is set to 3600s. The present invention adopts the EKF algorithm for SOC, SOH, and SOE joint estimation, which not only ensures the real-time and accuracy of SOC and SOE estimation and the accuracy of SOH estimation, but also reduces the computational complexity.

[0006] Preferably, the specific process of step S4 includes the following steps: step S41) obtaining a battery SOC estimation equation matrix by extending the Kalman filter based on the battery SOC estimation state equation and the battery SOC estimation observation equation; step S42) obtaining a battery SOH estimation equation matrix by extending the Kalman filter based on the battery SOH estimation state equation; step S43) obtaining a battery SOE estimation equation matrix by extending the Kalman filter based on the battery SOE estimation state equation and the battery SOE estimation observation equation. The present invention establishes a multi-scale calculation formula for joint estimation of battery SOC, SOH, and SOE through an extended Kalman filter (EKF) based on the estimated state equation and the estimated observation equation.

[0007] Preferably, the battery second-order RC equivalent circuit model includes a battery open-circuit voltage, a battery internal resistance, a battery polarization resistance, a battery polarization capacitance, a battery concentration difference resistance, and a battery concentration difference capacitance. The positive pole of the battery open-circuit voltage is connected to one end of the battery internal resistance, the other end of the battery internal resistance is connected to one end of the battery polarization resistance, the other end of the battery polarization resistance is connected to one end of the battery concentration difference resistance, the battery polarization capacitance is connected in parallel with the battery polarization resistance, the battery concentration difference capacitance is connected in parallel with the battery concentration difference resistance, and the other end of the battery concentration difference resistance and the negative pole of the battery open-circuit voltage serve as the output end of the equivalent circuit. The present invention establishes a battery second-order RC equivalent circuit model based on battery characteristics.

[0008] Preferably, the model parameters include battery internal resistance, battery polarization resistance, battery polarization capacitance, battery concentration difference resistance, and battery concentration difference capacitance. The present invention collects data through hybrid pulse power performance testing (HPPC) and uses the least squares method to identify the data to obtain model parameters.

[0009] Preferably, in step S41, the battery SOC estimation state equation is:

[0010]

[0011] The battery SOC estimation observation equation is:

[0012] y k,l =Uocv(SOC k,l )-U k,l *R2-U k,l *R1-U k,l *R0+ν k,l =g(x k,l ,U k,l )+ν k,l

[0013] According to the above battery SOC estimation state equation and battery SOC estimation observation equation, the battery SOC estimation equation matrix is ​​obtained by extended Kalman filtering. The battery SOC estimation equation matrix is:

[0014]

[0015] in,

[0016] Preferably, in step S42, the battery SOH estimation state equation is:

[0017] Q k+1 =Q k +r k

[0018] According to the battery SOH estimation state equation, the battery SOH estimation equation matrix is ​​obtained by extended Kalman filtering. The battery SOH estimation equation matrix is:

[0019]

[0020] Preferably, in step S41, the battery SOE estimation state equation is:

[0021]

[0022] The battery SOE estimation observation equation is:

[0023] e k,l =E ocv (SOE K,L )-U k,l *R2-U k,l *R1-U k,l *R0+v k.l =f(x k.l ,U k.l )+v k.l

[0024] According to the battery SOE estimation state equation and the battery SOE estimation observation equation, the battery SOE estimation equation matrix is ​​obtained by extended Kalman filtering. The battery SOE estimation equation matrix is:

[0025]

[0026] in, The present invention establishes a multi-scale calculation formula for joint estimation of battery SOC, SOH, and SOE based on the estimated state equation and the estimated observation equation through the extended Kalman filter (EKF), which not only ensures the real-time and accuracy of SOC and SOE estimation and the accuracy of SOH estimation, but also reduces the computational complexity.

[0027] As a preference, the battery SOC observation matrix value is:

[0028]

[0029] The present invention obtains the battery SOC observation matrix value according to the battery SOC estimation state equation and the battery SOC estimation observation equation.

[0030] As a preference, the battery SOH observation matrix value is:

[0031]

[0032] The present invention obtains the above-mentioned battery SOH observation matrix value based on the battery SOH estimation state equation.

[0033] As a preference, the battery SOE observation matrix value is:

[0034]

[0035] The present invention obtains the battery SOE observation matrix value according to the battery SOE estimation state equation and the battery SOE estimation observation equation.

[0036] Therefore, the advantages of the present invention are:

[0037] (1) The present invention adopts the EKF algorithm to jointly estimate SOC, SOH, and SOE, which not only ensures the real-time and accuracy of SOC and SOE estimation and the accuracy of SOH estimation, but also reduces the computational complexity;

[0038] (2) It has a wide range of applications and can be applied to various types of batteries. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 4 is a circuit diagram of a second-order RC equivalent circuit model of a battery in an embodiment of the present invention.

[0040] Figure 2 It is a structural diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0042] A method for jointly estimating the state of charge (SOC), state of exhaust (SOH), and state of exhaust gas (SOE) of a battery comprises the following steps: step S1) reading an initial state value and an initial capacity value of the battery; step S2) establishing a second-order RC equivalent circuit model of the battery according to battery characteristics; step S3) collecting data through a hybrid pulse power performance test (HPPC) and identifying the data using a least squares method to obtain model parameters; step S4) establishing a multi-scale calculation formula for jointly estimating the state of charge (SOC), state of exhaust gas (SOH), and state of exhaust gas (SOE) of the battery according to the initial state value, the initial capacity value, and the model parameters through an extended Kalman filter (EKF); and step S5) first running an SOH calculation module to obtain an initial capacity calculation value, collecting real-time voltage and real-time current, and inputting the real-time voltage, real-time current, and initial capacity calculation value data into an SOC and SOE calculation module to calculate and obtain an SOC value and an SOE value. Simultaneously, an observation value updated in the SOC calculation is input into the SOH calculation module for calculation to obtain an SOH value. In order to improve the accuracy of SOC, SOH, and SOE calculations and reduce the calculation complexity in battery management, the present invention provides a battery SOC, SOH, and SOE joint estimation method, which specifically includes the following steps:

[0043] 1) Read the initial value of battery status and capacity;

[0044] 2) Establish a second-order RC equivalent circuit model of the battery based on the battery characteristics, such as Figure 1 As shown, the second-order RC equivalent circuit model of the battery includes a battery open circuit voltage, a battery internal resistance, a battery polarization resistance, a battery polarization capacitance, a battery concentration difference resistance, and a battery concentration difference capacitance. The positive electrode of the battery open circuit voltage is connected to one end of the battery internal resistance, the other end of the battery internal resistance is connected to one end of the battery polarization resistance, the other end of the battery polarization resistance is connected to one end of the battery concentration difference resistance, the battery polarization capacitance is connected in parallel with the battery polarization resistance, the battery concentration difference capacitance is connected in parallel with the battery concentration difference resistance, and the other end of the battery concentration difference resistance and the negative electrode of the battery open circuit voltage serve as the output end of the equivalent circuit. The OCV-SOC and OCV-SOE relationship formulas, i.e., the battery SOC and SOE estimation observation equations, are determined through testing.

[0045] 3) Collecting data from a hybrid pulse power performance test (HPPC) and identifying the model parameters using the least squares method, including the battery internal resistance, battery polarization resistance, battery polarization capacitance, battery concentration difference resistance, and battery concentration difference capacitance;

[0046] 4) Based on the estimated state equation and the estimated observation equation, the extended Kalman filter (EKF) is used to establish a multi-scale calculation formula for the joint estimation of battery SOC, SOH, and SOE, specifically:

[0047] 41) Estimate the state equation based on the battery SOC and the battery SOC estimation observation equation y k,l =Uocv(SOC k,l )-U k,l *R2-U k,l *R1-U k,l *R0+ν k,l =g(x k,l ,U k,l )+ν k,l , the battery SOC estimation equation matrix is ​​obtained by extended Kalman filtering:

[0048]

[0049] in,

[0050] 42) Estimate the state equation Q based on the battery SOH k+1 =Q k +r k , the battery SOH estimation equation matrix is ​​obtained by extended Kalman filtering:

[0051]

[0052] 43) Estimate the state equation based on the battery SOE and the battery SOC estimation observation equation e k,l =E ocv (SOE K,L )-U k,l *R2-U k,l *R1-U k,l *R0+v k.l =f(x k.l ,U k.l )+v k.l , the battery SOE estimation equation matrix is ​​obtained by extended Kalman filtering:

[0053]

[0054] in, According to the above estimated state equation and estimated observation equation, the observation matrix values ​​of battery SOC, SOH, and SOE are obtained as follows:

[0055]

[0056]

[0057]

[0058] Among them, EKF x Indicates SOC estimation related, EKF c Indicates SOH estimation related, EKF e Indicates the SOE estimation correlation; k and l are time coefficients, is the state estimate, A represents the system matrix, B, B e represents the observation matrix, Q k is the capacity, P - k-1,l is the error covariance estimation matrix, var(X 0,0 ) is the initial value of the error covariance, P k,l is the filtering error covariance matrix, r, ζ are system noise matrices, Γ is the interference matrix, σ, ν, ψ are the observation noise matrices, K is the Kalman filter gain coefficient, I is the identity matrix, C k Represents the observation matrix value, y k 、e k is the observed value, U k is the control vector (measured current), U ocv 、E ocv To estimate the cell terminal voltage, R0 is the battery internal resistance, R1 is the battery polarization resistance, C1 is the battery polarization capacitance, R2 is the battery concentration difference resistance, C2 is the battery concentration difference capacitance, t is the system operation cycle, V k is the real-time voltage, E k is the rated capacity, t is the SOC and SOE operating cycle, and Q0 is the factory rated capacity of the battery cell;

[0059] 5) First run the SOH calculation module to obtain the initial calculated value of capacity, collect real-time voltage and real-time current, input the real-time voltage, real-time current and initial calculated value of capacity data into the SOC and SOE calculation modules, calculate and obtain the SOC value and SOE value. When calculating SOH, the updated observation value in the SOC calculation is passed to the SOH calculation module to participate in the operation, and then calculate and obtain the SOH value. The SOE operation cycle is set to 0.1s and the SOH operation cycle is set to 3600s. The specific process is as follows: Figure 2 As shown:

[0060] In fact, SOC and SOH, and SOH and SOE are estimated simultaneously. For the convenience of description, the joint estimation of SOC and SOH and SOH and SOE is described separately.

[0061] 51) Joint estimation of SOC and SOH:

[0062] 511) First calculate SOH, according tok-1,l-1 Established at all times Update the time and calculate the capacity value This is used as the input value of SOC after parameter establishment, and the SOC calculation module performs time update;

[0063] 512) After the SOC time is updated, t is obtained k-1,l State output value at the moment

[0064] 513) performing measurement update according to the updated data of the SOC time to obtain the SOC value;

[0065] 514) Determine whether the time l satisfies l≥T. If so, proceed to step 515); if not, l=l+1, and return to the initial step of SOC calculation at t k-1,l-1 Establish parameters at all times, and then continue to calculate according to steps 511)-514) until l≥T is established;

[0066] 515) SOC calculation module performs data transmission and output To the SOH calculation module, at the same time k=k+1, return to the initial step of SOC calculation at t k-1,l-1 Establish parameters at all times, and then continue to calculate according to steps 511)-514);

[0067] 516) The SOH calculation module calculates the state error based on the state values ​​(i.e., SOC, U1, U2) at different times input by the SOC calculation module in step 512) and step 515);

[0068] 517) The SOH calculation module performs measurement updates after state error calculation and outputs the capacity value According to the capacity value Calculate t k The SOH value at the moment and the capacity value The initial capacity value at the next moment is input into the parameter establishment step of the SOH calculation module, and is jointly estimated with the SOC according to steps 511)-517).

[0069] 52) Joint estimation of SOH and SOE:

[0070] 521) First calculate SOH, according to k-1,l-1 Established at all times Update the time and calculate the capacity value This is used as the input value after the SOE parameters are established, and the SOE calculation module performs time updates;

[0071] 522) the SOE calculation module performs measurement updates based on the updated data;

[0072] 523) The SOE calculation module obtains the SOE value after the measurement update, and then uses the calculated data as the parameter establishment value to perform SOE calculation at the next moment, and continues steps 521)-523).

Claims

1. A battery SOC, SOH, SOE joint estimation method, characterized in that: The following steps are involved: Step S1: Read the initial value of the battery state and the initial value of the battery capacity; Step S2: Establishing a second-order RC equivalent circuit model of the battery according to the battery characteristics; Step S3: collecting data through hybrid pulse power performance test, and using the least squares method to identify the data to obtain model parameters; Step S4: Based on the initial state value, initial capacity value and model parameters, a multi-scale calculation formula for joint estimation of battery SOC, SOH and SOE is established through extended Kalman filtering, including: According to the battery SOC estimation state equation and the estimated observation equation, according to the battery SOH estimation state equation, according to the battery SOE estimation state equation and the estimated observation equation, the battery SOC, SOH, SOE estimation equation matrix is ​​obtained respectively through the extended Kalman filter; Step S5: First run the SOH calculation module to obtain the initial calculated value of capacity, collect real-time voltage and real-time current, input the real-time voltage, real-time current and initial calculated value of capacity data into the SOC and SOE calculation modules, calculate and obtain the SOC value and SOE value, and at the same time input the updated observation value in the SOC calculation into the SOH calculation module to participate in the calculation and obtain the SOH value, including:

511. First perform SOH calculation, according to t k-1,l-1 Established at all times Update time and calculate capacity value As the SOC input value, the SOC calculation module performs time updates; 512. Get status output value 513. Perform measurement update to obtain SOC value; 514. Is l ≥ T satisfied? If so, proceed to 515. If not, l = l + 1 and repeat 511-514 until l ≥ T.

515. Output Go to the SOH calculation module, k=k+1, and repeat steps 511-514; 516. The SOH calculation module calculates the state error based on the state values ​​at different times input by 512 and 515.

517. Measurement update, output capacity value Calculate t k The SOH value at this moment will be As the initial value of the capacity at the next moment, it is passed to the SOH module and estimated jointly with the SOC according to 511-517; Among them, k and l are time coefficients.

2. A battery SOC, SOH, SOE joint estimation method according to claim 1, characterized in that: The second-order RC equivalent circuit model of the battery includes a battery open circuit voltage, a battery internal resistance, a battery polarization resistance, a battery polarization capacitance, a battery concentration difference resistance and a battery concentration difference capacitance. The positive electrode of the battery open circuit voltage is connected to one end of the battery internal resistance, the other end of the battery internal resistance is connected to one end of the battery polarization resistance, the other end of the battery polarization resistance is connected to one end of the battery concentration difference resistance, the battery polarization capacitance is connected in parallel with the battery polarization resistance, the battery concentration difference capacitance is connected in parallel with the battery concentration difference resistance, and the other end of the battery concentration difference resistance and the negative electrode of the battery open circuit voltage serve as the output end of the equivalent circuit.

3. The method for jointly estimating battery SOC, SOH, and SOE according to claim 2, characterized in that: The model parameters include battery internal resistance, battery polarization resistance, battery polarization capacitance, battery concentration difference resistance, and battery concentration difference capacitance.

Images