Method and apparatus for estimating state of charge of a battery

Abstract

The application provides a battery state of charge estimation method and device, the method comprises the following steps: establishing an equivalent circuit model of the battery; determining a first state of charge parameter under a current state based on the equivalent circuit model; correcting the first state of charge parameter based on a fuzzy unscented Kalman particle filtering algorithm to obtain a second state of charge parameter. Through the application, the battery state of charge estimation combining particle filtering, unscented Kalman filtering and fuzzy control is realized, and the estimation accuracy is improved.

Description

Technical Field

[0001] This invention relates to the field of batteries, and in particular to a method and apparatus for estimating the state of charge of a battery. Background Technology

[0002] Currently, due to their advantages in energy density, cycle life, and power density, lithium-ion batteries are widely used in electric vehicles. However, improper management of lithium-ion batteries may cause accidents such as fires and explosions, affecting the safety of electric vehicles.

[0003] In the management of power lithium batteries, the state of charge (SOC) plays a very important role. However, in the existing technology, the estimated SOC is not accurate enough, which affects the management of power lithium batteries. Summary of the Invention

[0004] In view of the above problems, a method and apparatus for estimating the state of charge of a battery are proposed to overcome or at least partially solve the above problems, which may include:

[0005] A method for estimating the state of charge (SOC) of a battery, the method may include:

[0006] Establish an equivalent circuit model for the battery;

[0007] Based on the equivalent circuit model, determine the first charge state parameters in the current state;

[0008] Based on the fuzzy unscented Kalman particle filter algorithm, the first charged state parameters are corrected to obtain the second charged state parameters.

[0009] Optionally, the step of correcting the first charged state parameters based on the fuzzy unscented Kalman particle filter algorithm to obtain the second charged state parameters includes:

[0010] By combining particle filtering, unscented Kalman filtering, and fuzzy control, the first charged state parameters are corrected to obtain the second charged state parameters.

[0011] Optionally, determining the first charge state parameters in the current state based on the equivalent circuit model may include:

[0012] Based on the equivalent circuit model, the state and observation variable model for capacitive and resistive parameters is determined.

[0013] By combining extended Kalman filtering, parameter identification is performed on the state and observed variable model for capacitive and resistive parameters, and the parameter identification results are obtained.

[0014] Based on the parameter identification results, the first charge state parameter in the current state is determined.

[0015] Optionally, the parameter identification result is the target capacitive resistance state variable. Combined with extended Kalman filtering, parameter identification is performed on the state and observed variable model for the capacitive resistance parameter to obtain the parameter identification result, which may include:

[0016] Based on the state and observation variable model for the capacitive and resistive parameters, the initial state variables are determined, and the target capacitive and resistive state variables are obtained by performing extended Kalman filtering based on the initial state variables.

[0017] Optionally, determining the first state-of-charge parameter in the current state based on the parameter identification result may include:

[0018] Based on the target capacitive and resistive state variables, determine the target capacitive and resistive parameters of the equivalent circuit model;

[0019] Based on the target capacitive and resistive parameters, determine the first state-of-charge parameters in the current state.

[0020] Optionally, determining the first state-of-charge parameters in the current state based on the target capacitive-resistive parameters may include:

[0021] Based on the target capacitive and resistive parameters, determine the current open-circuit voltage of the equivalent circuit model;

[0022] Determine the first state-of-charge parameters in the current state based on the current open-circuit voltage.

[0023] Optionally, by combining particle filtering, unscented Kalman filtering, and fuzzy control, the first charge state parameters are corrected to obtain the second charge state parameters, which may include:

[0024] Based on the equivalent circuit model, determine the state and observation variable model for the state of charge;

[0025] Based on the state and observation variable model for the state of charge, the initial state of charge variable corresponding to the first state of charge parameter is determined, and particle filtering and unscented Kalman filtering are performed on the initial state of charge variable to obtain the target state of charge variable.

[0026] Determine the second charge state parameters based on the target charge state variable.

[0027] Optionally, the target charged state variable can be obtained by performing particle filtering and unscented Kalman filtering based on the initial charged state variable, which may include:

[0028] Based on the initial charged state variables, sample multiple charged state particles;

[0029] For each charged state particle, perform an unscented Kalman filter to obtain updated charged state particles;

[0030] Determine the number of valid charged state particles in the updated charged state particles;

[0031] When the number of effective charged state particles is greater than or equal to the preset number of charged state particles, the charged state variable of the updated charged state particles is determined as the target charged state variable.

[0032] Optionally, it may also include:

[0033] When the number of effective charged state particles is less than the preset number of charged state particles, resampling is performed, and for each of the resampled charged state particles, an unscented Kalman filter is applied to obtain updated charged state particles.

[0034] Optionally, performing an unscented Kalman filter on each charged-state particle to obtain updated charged-state particles may include:

[0035] During the unscented Kalman filtering of each charged particle, the difference between the theoretical covariance and the actual covariance of the noise is determined, and the adjustment amount is determined based on the difference.

[0036] Based on the adjustment amount, update the noise covariance matrix, and based on the updated Kalman gain matrix based on the noise covariance matrix, determine the updated charged state particles.

[0037] Optionally, the equivalent circuit model may include at least three equivalent sub-circuits, which may be parallel circuits of capacitors and resistors;

[0038] The equivalent sub-circuit may include:

[0039] Parallel circuits of ohmic polarized resistors and ohmic polarized capacitors, parallel circuits of electrochemical polarized resistors and electrochemical polarized capacitors, and parallel circuits of concentration difference polarized resistors and concentration difference polarized capacitors.

[0040] A battery state-of-charge estimation device, the device may include:

[0041] The equivalent circuit model can be used to establish the equivalent circuit model of a battery.

[0042] The circuit for determining the first state of charge parameters can be used to determine the first state of charge parameters in the current state based on an equivalent circuit model.

[0043] The circuit for obtaining the second charged state parameter can be used to correct the first charged state parameter based on the fuzzy unscented Kalman particle filter algorithm to obtain the second charged state parameter.

[0044] The embodiments of the present invention have the following advantages:

[0045] In this embodiment of the invention, by establishing an equivalent circuit model of the battery, the first state-of-charge parameter in the current state is determined based on the equivalent circuit model. The first state-of-charge parameter is then corrected based on the fuzzy unscented Kalman particle filter algorithm to obtain the second state-of-charge parameter. This achieves battery state-of-charge estimation by combining particle filtering, unscented Kalman filtering, and fuzzy control, thereby improving the accuracy of the estimation. Attached Figure Description

[0046] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description of the present invention will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1a This is a schematic diagram of a third-order resistor-capacitor equivalent circuit provided in an embodiment of the present invention;

[0048] Figure 1b This is a schematic diagram of an example of state-of-charge estimation based on a third-order resistor-capacitor equivalent circuit provided in an embodiment of the present invention.

[0049] Figure 2 This is a flowchart of the steps of a battery state-of-charge estimation method provided in an embodiment of the present invention;

[0050] Figure 3a This is a flowchart of another method for estimating the state of charge of a battery according to an embodiment of the present invention;

[0051] Figure 3b This is a schematic diagram of parameter identification based on the extended Kalman filter algorithm according to an embodiment of the present invention;

[0052] Figure 4a This is a flowchart of another method for estimating the state of charge of a battery according to an embodiment of the present invention;

[0053] Figure 4b This is a schematic diagram of a method for estimating charge state parameters based on a fuzzy unscented Kalman particle filter algorithm according to an embodiment of the present invention;

[0054] Figure 5 This is a flowchart of another method for estimating the state of charge of a battery according to an embodiment of the present invention;

[0055] Figure 6 This is a structural block diagram of a battery state-of-charge estimation device provided in an embodiment of the present invention. Detailed Implementation

[0056] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0057] In this embodiment of the invention, the state of charge of a battery can be estimated based on a third-order resistor-capacitor (RC) equivalent circuit model, combined with an unscented Kalman filter algorithm, a particle filter algorithm, and a fuzzy control algorithm, thereby improving the accuracy of the state of charge estimation.

[0058] Among them, such as Figure 1a The circuit diagram is a third-order resistor-capacitor equivalent circuit. R1 is the ohmic polarization resistor, R2 is the electrochemical polarization internal resistance, R3 is the concentration difference polarization internal resistance, C1 is the ohmic polarization capacitor, C2 is the electrochemical polarization capacitor, and C3 is the concentration difference polarization capacitor.

[0059] In one example, the third-order resistor-capacitor equivalent circuit can also include the fixed ohmic internal resistance of the battery, such as... Figure 1a R0 can be connected in series with the negative terminal of the battery.

[0060] Specifically, such as Figure 1b By establishing a third-order capacitor-resistance equivalent circuit model, parameter identification is performed based on the extended Kalman filter algorithm to obtain the capacitive and resistive parameters of the equivalent circuit model (resistance values ​​R0, R1, R2, R3, and capacitance values ​​C1, C2, C3). Then, the open circuit voltage (OCV) is calculated based on the capacitive and resistive parameters. Furthermore, based on the relationship between the open circuit voltage and the state of charge, the state of charge parameter (SOC0) is obtained as input. This SOC0 is then input into the fuzzy unscented Kalman particle filter algorithm to obtain an estimated value of the state of charge parameter.

[0061] It can mainly include the following:

[0062] 1. Collect parameters such as voltage and current of the battery under the current state, and establish a third-order resistive-capacitive equivalent circuit model of the battery based on the collected parameters. Based on the third-order resistive-capacitive equivalent circuit model, establish the corresponding state and observation variable model.

[0063] 2. Based on the state and observation variable model for the capacitive and resistive parameters, the extended Kalman filter algorithm is used to identify the parameters and obtain the estimated capacitive and resistive parameters.

[0064] 3. Based on the system state-observation model for the state of charge, a fuzzy unscented Kalman particle filter algorithm, which combines the unscented Kalman filter algorithm, the particle filter algorithm, and the fuzzy control algorithm, is used to estimate the state of charge of lithium-ion batteries.

[0065] The beneficial effects achieved through the embodiments of the present invention include the following:

[0066] 1. Establishment of the equivalent circuit model: By adopting a third-order resistor-capacitor equivalent circuit model, the influence of the battery's polarization impedance on performance is fully considered, and the battery's state information can be reflected in a timely and accurate manner. Moreover, the equivalent circuit model has high accuracy and is easy to implement.

[0067] 2. Identification of capacitive and resistive parameters: By using the extended Kalman particle filter algorithm to identify capacitive and resistive parameters, the accuracy of the capacitive and resistive parameters is improved, which in turn improves the accuracy of subsequent estimation.

[0068] 3. For the estimation of the state of charge: By adopting the fuzzy unscented Kalman particle filter algorithm, the combination of fuzzy control, unscented Kalman filtering, and particle filtering algorithms is used. This combines the advantages of unscented Kalman filtering and particle filtering algorithms, realizes adaptive adjustment of the filter, reduces the cumulative noise error and iteration error of the unscented Kalman filtering algorithm, and improves the accuracy of the algorithm estimation.

[0069] The following is a detailed explanation:

[0070] (It should be noted that the parameters of the formulas and expressions in the embodiments of this invention are all based on...) Figure 1a (Third-order resistor-capacitor equivalent circuit, with relevant parameters cross-referenced)

[0071] Reference Figure 2 The diagram illustrates a flowchart of a method for estimating the state of charge of a battery according to an embodiment of the present invention, which may specifically include the following steps:

[0072] Step 201: Establish the equivalent circuit model of the battery.

[0073] Among them, the battery can be a lithium iron phosphate battery, which can be used as a power lithium battery in electric vehicles.

[0074] While first-order and second-order equivalent circuit models of resistors and capacitors are easy to establish and can improve the accuracy of model prediction by combining with algorithms, they still cannot fundamentally solve the bottleneck of relatively large model errors. In the embodiments of this invention, by establishing third-order (or higher) equivalent circuits of resistors and capacitors, the influence of the polarization impedance of the battery on performance can be fully considered, and the state information of the battery can be reflected in a timely and accurate manner. The model has high accuracy, is easy to implement, and can take into account the voltage rebound characteristics and hysteresis voltage characteristics of lithium-ion batteries.

[0075] In one embodiment of the present invention, the equivalent circuit model may include at least three equivalent sub-circuits composed of resistors and capacitors, such as... Figure 1a In the third-order resistive-capacitive equivalent circuit model, there are three equivalent sub-circuits connected in series with the positive terminal of the battery, and each equivalent sub-circuit consists of a resistor and a capacitor connected in parallel.

[0076] In one embodiment of the present invention, the equivalent sub-circuit may include: a parallel circuit of an ohmic polarization resistor and an ohmic polarization capacitor, a parallel circuit of an electrochemical polarization resistor and an electrochemical polarization capacitor, and a parallel circuit of a concentration difference polarization resistor and a concentration difference polarization capacitor.

[0077] Step 202: Based on the equivalent circuit model, determine the first charge state parameters in the current state.

[0078] After establishing the equivalent circuit model, the first state-of-charge parameters of the battery in the current state can be determined based on the capacitive and resistive parameters of the equivalent circuit model.

[0079] Step 203: Based on the fuzzy unscented Kalman particle filter algorithm, the first charged state parameters are corrected to obtain the second charged state parameters.

[0080] In one embodiment of the present invention, step 203 may include:

[0081] By combining particle filtering, unscented Kalman filtering, and fuzzy control, the first charged state parameters are corrected to obtain the second charged state parameters.

[0082] Since the unscented Kalman filter is a highly adaptable filtering algorithm, but suffers from the drawback of accumulated iteration errors, while the particle filter is suitable for solving filtering simulation problems of nonlinear complex systems, it also has shortcomings such as poor adaptability and particle degeneration. Therefore, the particle filter algorithm and the unscented Kalman filter algorithm are integrated to form the unscented Kalman particle filter algorithm, which retains the ability of the particle filter algorithm to handle nonlinear and non-Gaussian systems, while taking into account the advantages of the unscented Kalman filter algorithm such as good correction ability and strong robustness.

[0083] Furthermore, since the statistical characteristics of the observed noise are constantly changing during battery use, the measurement noise is also constantly changing. Therefore, the covariance matrix can be dynamically adjusted through fuzzy control to achieve adaptive adjustment of the filter, forming a fuzzy unscented Kalman particle filter algorithm. This reduces the cumulative noise error of the unscented Kalman particle filter algorithm and improves the accuracy of the algorithm estimation.

[0084] Based on this, the first charged state parameter can be used as the input quantity in the equivalent circuit model. The fuzzy unscented Kalman particle filter algorithm is used to perform particle filtering, unscented Kalman filtering, and fuzzy control on the first charged state parameter to obtain the second charged state parameter after correcting the first charged state parameter.

[0085] In this embodiment of the invention, by establishing an equivalent circuit model of the battery, the first state-of-charge parameter in the current state is determined based on the equivalent circuit model. The first state-of-charge parameter is then corrected based on the fuzzy unscented Kalman particle filter algorithm to obtain the second state-of-charge parameter. This achieves battery state-of-charge estimation by combining particle filtering, unscented Kalman filtering, and fuzzy control, thereby improving the accuracy of the estimation.

[0086] Reference Figure 3a The diagram illustrates a flowchart of another method for estimating the state of charge of a battery according to an embodiment of the present invention, which may specifically include the following steps:

[0087] Step 301: Establish the equivalent circuit model of the battery.

[0088] Step 302: Based on the equivalent circuit model, determine the state and observation variable model for the capacitive and resistive parameters.

[0089] Based on the equivalent circuit model, expressions for the voltage and current of each part can be obtained, such as... Figure 1a In the case of a third-order resistor-capacitor equivalent circuit model, the expression is as follows:

[0090]

[0091] Discretizing the expressions for current and voltage yields the following:

[0092]

[0093] Where s is the Laplace operator, and τ1=R1C1, τ2=R2C2, τ3=R3C3 are the RC time constants in the circuit model.

[0094] After discretization, the equations are transformed to obtain the state and observation variable equations for the capacitive and resistive parameters, i.e., the state and observation variable model, as follows:

[0095]

[0096] Among them, y k For the capacitive and resistive observation variable, x k H is the capacitive-resistive state variable. k Let k be the data matrix, and k represent the sampling time.

[0097] Step 303: Combine extended Kalman filtering to perform parameter identification on the state and observation variable model for capacitive and resistive parameters, and obtain the parameter identification results.

[0098] After determining the state and observation variable model for the capacitive and resistive parameters, the extended Kalman filter algorithm can be used to identify the parameters of the state and observation variable model for the capacitive and resistive parameters, and obtain the parameter identification results. The parameter identification results can include the target capacitive and resistive state variables, that is, the input capacitive and resistive state variables.

[0099] In one embodiment of the present invention, step 303 may include:

[0100] Sub-step 11: Based on the state and observation variable model for the capacitive and resistive parameters, determine the initial state variables, and perform extended Kalman filtering based on the initial state variables to obtain the target capacitive and resistive state variables.

[0101] Based on the state and observation variable model for the capacitive and resistive parameters, the expression for the capacitive and resistive state variables can be determined, which can be composed of the resistance and capacitance of each equivalent sub-circuit.

[0102] Based on the expression for the capacitive and resistive state variables, initialization can be performed to obtain the initial state variables, as follows:

[0103]

[0104] After obtaining the initial state variables, parameter identification can be performed based on the extended Kalman filter to obtain the target capacitive-resistive state variables.

[0105] The following combination Figure 3b Explanation of sub-step 12:

[0106] 1. Parameter initialization:

[0107]

[0108] 2. State prior estimation:

[0109]

[0110] 3. Calculate the error covariance state estimate:

[0111]

[0112] 4. Calculate the Kalman gain:

[0113]

[0114] 5. State estimate update

[0115]

[0116] 6. Update the error covariance matrix:

[0117]

[0118] Among them, u k Let A be the input value of the system at time k. k For the transfer matrix, B k Given the input matrix, C k For the output matrix, D k Let Q be the feedforward matrix. k The system noise ω k covariance, R k For observing noise ν k The covariance of E is the identity matrix.

[0119] 7. Repeat steps 2 through 7 until the preset step size is reached, and output the parameter identification results.

[0120] Step 304: Determine the first charge state parameter in the current state based on the parameter identification result.

[0121] After determining the parameter identification results, the capacitive and resistive parameters can be determined based on the parameter identification results, and then the first state of charge parameters in the current state can be determined based on the capacitive and resistive parameters.

[0122] In one embodiment of the present invention, step 304 may include:

[0123] Sub-step 21: Determine the target capacitive and resistive parameters of the equivalent circuit model based on the target capacitive and resistive state variables.

[0124] After obtaining the target capacitive resistance state variables, the target capacitive resistance parameters, i.e., the parameter identification results, can be obtained based on the target capacitive resistance state variables and the preset coefficient matrix, as follows:

[0125]

[0126] Where σ is the coefficient matrix, σ = [σ1 σ2 σ3 σ4 σ5 σ6], and the capacitive / resistive parameter matrix is...

[0127] Sub-step 22: Determine the first state-of-charge parameters in the current state based on the target capacitive and resistive parameters.

[0128] In one embodiment of the present invention, sub-step 22 may include:

[0129] Based on the target capacitive and resistive parameters, determine the current open-circuit voltage of the equivalent circuit model; based on the current open-circuit voltage, determine the first state-of-charge parameters in the current state.

[0130] After obtaining the target capacitive and resistant parameters, the current open-circuit voltage can be determined based on the target capacitive and resistant parameters. Then, the first state-of-charge parameter can be determined based on the relationship between the open-circuit voltage and the state-of-charge parameter.

[0131] Step 305: Based on the fuzzy unscented Kalman particle filter algorithm, the first charged state parameters are corrected to obtain the second charged state parameters.

[0132] In this embodiment of the invention, an equivalent circuit model of the battery is established. Based on the equivalent circuit model, the state and observation variable model for the capacitive and resistive parameters is determined. Combined with extended Kalman filtering, parameter identification is performed on the state and observation variable model for the capacitive and resistive parameters to obtain parameter identification results. Based on the parameter identification results, the first state-of-charge parameter in the current state is determined. Based on the fuzzy unscented Kalman particle filter algorithm, the first state-of-charge parameter is corrected to obtain the second state-of-charge parameter. This realizes the use of extended Kalman filtering for capacitive and resistive parameter identification, thereby improving the accuracy of capacitive and resistive parameters.

[0133] Reference Figure 4a The diagram illustrates a flowchart of another method for estimating the state of charge of a battery according to an embodiment of the present invention, which may specifically include the following steps:

[0134] Step 401: Establish the equivalent circuit model of the battery.

[0135] Step 402: Based on the equivalent circuit model, determine the first charge state parameters in the current state.

[0136] Step 403: Based on the equivalent circuit model, determine the state and observation variable model for the charged state.

[0137] Based on the equivalent circuit model, state and observation variable equations can be established for the charged state, i.e., the state and observation variable model, as shown below:

[0138]

[0139] Where, x k Let y be a state variable. k For the observed variable, ω k For the process noise of the system, νk To observe the noise, its noise covariance is Q. k and R k Q n Let η be the lithium battery capacity, η be the lithium battery coulombic efficiency, k be the sampling time, and T be the sampling period.

[0140] Where U1, U2, and U3 are the voltages across capacitors C1, C2, and C3, respectively, and I is the current.

[0141] Step 404: Based on the state and observation variable model for the state of charge, determine the initial state of charge variables corresponding to the first state of charge parameters, and perform particle filtering, unscented Kalman filtering, and fuzzy control based on the initial state of charge variables to obtain the target state of charge variables.

[0142] Based on the observations of the state of charge and the model of the state of charge variables, the expression for the state of charge variables can be determined, which can be composed of the input first state of charge parameters and the voltages of each equivalent sub-circuit, as shown below:

[0143] x k =[U 1,k U 2,k U 3,k SOC k ] T

[0144] Among them, SOC k For input charge state parameters.

[0145] Based on the expression for the charged state variable, initialization can be performed by setting k = 0, that is, at time 0, the voltage is initialized to 0, and the initial charged state variable can be obtained.

[0146]

[0147] After obtaining the initial charged state variables, particle filtering, unscented Kalman filtering, and fuzzy control can be performed on the basis of the initial charged state variables to obtain the target charged state variables.

[0148] The following combination Figure 4b Explanation of step 404:

[0149] In one embodiment of the present invention, obtaining the target charged state variable by performing particle filtering and unscented Kalman filtering based on the initial charged state variable may include:

[0150] Sub-step 31: Based on the initial charged state variables, sample multiple charged state particles.

[0151] Based on the initial charged state variables, multiple charged state particles can be sampled using a particle filtering algorithm, and their charged state variables are the initial charged state variables.

[0152] For example, drawing N particles from a prior distribution can be expressed as:

[0153] In one example, the covariance and particle weights can be initialized as follows:

[0154]

[0155]

[0156] Where P is the covariance and ω is the particle weight.

[0157] Sub-step 32: Perform unscented Kalman filtering on each charged state particle to obtain updated charged state particles.

[0158] For each charged state particle, its charged state variables and covariance can be updated using the unscented Kalman filter algorithm to obtain the updated charged state particle. For example, the unupdated charged state particle would be... Its updated charged state variable is The updated covariance is

[0159] In one embodiment of the present invention, sub-step 32 may include:

[0160] During the unscented Kalman filtering process for each charged state particle, the difference between the theoretical covariance and the actual covariance of the noise is determined, and the adjustment amount is determined based on the difference; the noise covariance matrix is ​​updated based on the adjustment amount, and the updated charged state particle is determined based on the updated Kalman gain matrix based on the noise covariance matrix.

[0161] The process of performing unscented Kalman filtering may include the following:

[0162] 1. First, consider the charged state variables. Perform an unscented transformation to obtain the sigma point set (sample point set). And the weight coefficient W of the sampling points can be calculated, as shown below:

[0163]

[0164]

[0165] Where α is a non-negative factor (e.g., 0.01), β is a non-negative weighting coefficient, and λ is a scaling factor.

[0166] 2. Nonlinearly transform the 2n+1 sigma point set to obtain the predicted state variables. As shown below:

[0167]

[0168] Then, the first-order predicted values ​​of the state variables and the covariance matrix can be calculated, as shown below:

[0169]

[0170]

[0171] 3. Perform an unscented transformation on the predicted values ​​to obtain an updated set of sampling points.

[0172] 4. Apply the observation equation (i.e., the observation state model) to the sample point set. Perform a nonlinear transformation to obtain the predicted observed variables of the new sampling point set. As shown below:

[0173]

[0174] The first predicted value of the weighted observed variable And the covariance matrix, as shown below:

[0175]

[0176]

[0177]

[0178] 5. Calculate the theoretical covariance S of the system process noise. k and actual covariance C k As shown below:

[0179]

[0180]

[0181] in,

[0182] 6. Since the statistical characteristics of observed noise are constantly changing during battery use, the measurement noise is also constantly changing. Therefore, fuzzy control can be used to dynamically adjust the covariance matrix, achieving adaptive adjustment of the filter, reducing the accumulated noise error of the unscented Kalman particle filter algorithm, and improving the accuracy of the algorithm's estimation. Based on this, the difference between the theoretical covariance and the actual covariance of the noise can be determined during the unscented Kalman filtering process for each charged particle, as shown below:

[0183] e k =S k -C k

[0184] Among them, S k For theoretical covariance, C k Let e ​​be the actual covariance matrix. k The difference is the sum of the two values.

[0185] After obtaining the difference, the difference is used as the input to fuzzy control to obtain the adjustment amount of the output. The relationship between the difference and the adjustment amount can be shown as follows:

[0186] If the difference is e k If ≈0, then the adjustment amount r k If the difference e remains unchanged k If the value is greater than 0, then the adjustment amount r k Decrease, if the difference e k If the value is less than 0, then the adjustment amount r k Increase.

[0187] 7. After obtaining the adjustment amount, update the noise covariance matrix according to the adjustment amount, as shown below:

[0188] R k =R k-1 +r k

[0189] Among them, R k For the updated variance matrix, R k-1 This is the variance matrix before the update.

[0190] 8. After updating the covariance matrix, the Kalman gain matrix after updating the noise covariance matrix can be calculated, as shown below:

[0191]

[0192] Among them, K k For Kalman gain.

[0193] 9. After obtaining the Kalman gain, the charged state particles can be updated, such as updating the charged state particles to... The updated state variables and covariance of the charged particles are shown below:

[0194]

[0195]

[0196] Sub-step 33: Determine the number of valid charged state particles among the updated charged state particles.

[0197] After updating the charged state particles, the number of valid charged state particles among the updated charged state particles can be determined. Specifically, the particle weights of the updated charged state particles can be calculated first, and then the particle weights can be normalized to obtain normalized weights. Based on the normalized weights, the number of valid charged state particles can be determined.

[0198] For example, updated charged state particles The particle weights are updated as follows:

[0199]

[0200] After normalization, the normalized weights are shown below:

[0201]

[0202] After obtaining the normalized weights, the number of effective charged state particles can be calculated as follows:

[0203]

[0204] Sub-step 34: When the number of effective charged state particles is greater than or equal to the preset number of charged state particles, determine the charged state variable of the updated charged state particles as the target charged state variable.

[0205] The number of particles in an effective state of charge is greater than or equal to the preset number of particles in a state of charge (N) th When ), the updated charged state variable of the charged state particle can be determined as the target charged state variable.

[0206] In one embodiment of the present invention, it may further include:

[0207] When the number of effective charged state particles is less than the preset number of charged state particles, resampling is performed, and unscented Kalman filtering is performed on each of the resampled charged state particles to obtain updated charged state particles.

[0208] When the number of effective charged state particles is less than the preset number of charged state particles, the charged state particles can be resampled. For example, if the number of resampled charged state particles is... Accordingly, its particle weights are updated as follows: Then sub-steps 32 and 33 can be re-executed until the number of valid charged state particles is greater than or equal to the preset number of charged state particles, thus obtaining the target charged state variable.

[0209] Step 405: Determine the second charge state parameter based on the target charge state variable.

[0210] After determining the target charged state variable, the second charged state parameter can be obtained by combining the target charged state variable and the particle weights of the charged state particles, as shown below:

[0211]

[0212] In one example, it can be determined whether the current state is the end of discharge. If so, the second state of charge parameter is output; otherwise, the step of determining the second state of charge parameter is repeated.

[0213] In this embodiment of the invention, by establishing an equivalent circuit model of the battery, a first state-of-charge parameter in the current state is determined based on the equivalent circuit model. A state and observation variable model for the state of charge is then determined based on the equivalent circuit model. Based on this model, an initial state-of-charge variable corresponding to the first state-of-charge parameter is determined. Particle filtering, unscented Kalman filtering, and fuzzy control are then applied to the initial state-of-charge variable to obtain a target state-of-charge variable. Based on the target state-of-charge variable, a second state-of-charge parameter is determined. This process establishes a state and observation variable model for the state of charge, updates the state-of-charge variable using particle filtering, unscented Kalman filtering, and fuzzy control, and then estimates the battery's state of charge based on the state-of-charge variable, thus improving the accuracy of the estimation.

[0214] Reference Figure 5 The diagram illustrates a flowchart of another method for estimating the state of charge of a battery according to an embodiment of the present invention, which may specifically include the following steps:

[0215] Step 501: Establish the equivalent circuit model of the battery.

[0216] Step 502: Based on the equivalent circuit model, determine the state and observation variable model for the capacitive and resistive parameters.

[0217] Step 503: Combine extended Kalman filtering to perform parameter identification on the state and observation variable model for capacitive and resistive parameters, and obtain the parameter identification results.

[0218] Step 504: Determine the first charge state parameter in the current state based on the parameter identification result.

[0219] Step 505: Based on the equivalent circuit model, determine the state and observation variable model for the charged state.

[0220] Step 506: Based on the state and observation variable model for the state of charge, determine the initial state of charge variables corresponding to the first state of charge parameters, and perform particle filtering, unscented Kalman filtering, and fuzzy control according to the initial state of charge variables to obtain the target state of charge variables.

[0221] Step 507: Determine the second charge state parameter based on the target charge state variable.

[0222] In this embodiment of the invention, an equivalent circuit model of the battery is established. Based on the equivalent circuit model, a state and observation variable model for capacitive and resistive parameters is determined. Combined with extended Kalman filtering, parameter identification is performed on the state and observation variable model for capacitive and resistive parameters to obtain parameter identification results. Based on the parameter identification results, the first state-of-charge parameter in the current state is determined. Based on the equivalent circuit model, a state and observation variable model for the state of charge is determined. Based on the state and observation variable model for the state of charge, the initial state-of-charge variable corresponding to the first state-of-charge parameter is determined. Based on the initial state-of-charge variable, particle filtering, unscented Kalman filtering, and fuzzy control are performed to obtain the target state-of-charge variable. Based on the target state-of-charge variable, the second state-of-charge parameter is determined. This achieves the use of extended Kalman filtering for capacitive and resistive parameter identification, improving the accuracy of capacitive and resistive parameters. Furthermore, by combining particle filtering, unscented Kalman filtering, and fuzzy control, the state of charge of the battery is estimated, improving the accuracy of the estimation.

[0223] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.

[0224] Reference Figure 6 The diagram shows a structural schematic of a battery state-of-charge estimation device according to an embodiment of the present invention, which may specifically include the following modules:

[0225] The equivalent circuit model establishment circuit 601 can be used to establish the equivalent circuit model of a battery.

[0226] The first charge state parameter determination circuit 602 can be used to determine the first charge state parameters in the current state based on the equivalent circuit model.

[0227] The circuit 603 for obtaining the second charge state parameter can be used to correct the first charge state parameter based on the fuzzy unscented Kalman particle filter algorithm to obtain the second charge state parameter.

[0228] In one embodiment of the present invention, the first charge state parameter is corrected based on the fuzzy unscented Kalman particle filter algorithm to obtain the second charge state parameter, which may include:

[0229] By combining particle filtering, unscented Kalman filtering, and fuzzy control, the first charged state parameters are corrected to obtain the second charged state parameters.

[0230] In one embodiment of the present invention, determining the first state-of-charge parameter in the current state based on an equivalent circuit model may include:

[0231] Based on the equivalent circuit model, the state and observation variable model for capacitive and resistive parameters is determined.

[0232] By combining extended Kalman filtering, parameter identification is performed on the state and observed variable model for capacitive and resistive parameters, and the parameter identification results are obtained.

[0233] Based on the parameter identification results, the first charge state parameter in the current state is determined.

[0234] In one embodiment of the present invention, the parameter identification result is the target capacitive resistance state variable. Combined with extended Kalman filtering, parameter identification is performed on the state and observed variable model targeting the capacitive resistance parameter to obtain the parameter identification result, which may include:

[0235] Based on the state and observation variable model for the capacitive and resistive parameters, the initial state variables are determined, and the target capacitive and resistive state variables are obtained by performing extended Kalman filtering based on the initial state variables.

[0236] In one embodiment of the present invention, determining the first state-of-charge parameter in the current state based on the parameter identification result may include:

[0237] Based on the target capacitive and resistive state variables, determine the target capacitive and resistive parameters of the equivalent circuit model;

[0238] Based on the target capacitive and resistive parameters, determine the first state-of-charge parameters in the current state.

[0239] In one embodiment of the present invention, determining the first state-of-charge parameter in the current state based on the target capacitive resistance parameter may include:

[0240] Based on the target capacitive and resistive parameters, determine the current open-circuit voltage of the equivalent circuit model;

[0241] Determine the first state-of-charge parameters in the current state based on the current open-circuit voltage.

[0242] In one embodiment of the present invention, the first charge state parameter is corrected by combining particle filtering, unscented Kalman filtering, and fuzzy control to obtain the second charge state parameter, which may include:

[0243] Based on the equivalent circuit model, determine the state and observation variable model for the state of charge;

[0244] Based on the state and observation variable model for the state of charge, the initial state of charge variable corresponding to the first state of charge parameter is determined, and particle filtering, unscented Kalman filtering, and fuzzy control are performed on the initial state of charge variable to obtain the target state of charge variable.

[0245] Determine the second charge state parameters based on the target charge state variable.

[0246] In one embodiment of the present invention, obtaining the target charged state variable by performing particle filtering, unscented Kalman filtering, and fuzzy control based on the initial charged state variable may include:

[0247] Based on the initial charged state variables, sample multiple charged state particles;

[0248] For each charged state particle, perform an unscented Kalman filter to obtain updated charged state particles;

[0249] Determine the number of valid charged state particles in the updated charged state particles;

[0250] When the number of effective charged state particles is greater than or equal to the preset number of charged state particles, the charged state variable of the updated charged state particles is determined as the target charged state variable.

[0251] In one embodiment of the present invention, it can also be used for:

[0252] When the number of effective charged state particles is less than the preset number of charged state particles, resampling is performed, and for each of the resampled charged state particles, an unscented Kalman filter is applied to obtain updated charged state particles.

[0253] In one embodiment of the present invention, performing unscented Kalman filtering on each charged-state particle to obtain updated charged-state particles may include:

[0254] During the unscented Kalman filtering of each charged particle, the difference between the theoretical covariance and the actual covariance of the noise is determined, and the adjustment amount is determined based on the difference.

[0255] Based on the adjustment amount, update the noise covariance matrix, and based on the updated Kalman gain matrix based on the noise covariance matrix, determine the updated charged state particles.

[0256] In one embodiment of the present invention, the equivalent circuit model may include at least three equivalent sub-circuits, which may be a parallel circuit of capacitors and resistors.

[0257] The equivalent sub-circuit may include:

[0258] Parallel circuits of ohmic polarized resistors and ohmic polarized capacitors, parallel circuits of electrochemical polarized resistors and electrochemical polarized capacitors, and parallel circuits of concentration difference polarized resistors and concentration difference polarized capacitors.

[0259] In this embodiment of the invention, by establishing an equivalent circuit model of the battery, the first state-of-charge parameter in the current state is determined based on the equivalent circuit model. The first state-of-charge parameter is then corrected based on the fuzzy unscented Kalman particle filter algorithm to obtain the second state-of-charge parameter. This achieves the estimation of the battery's state of charge by combining particle filtering, unscented Kalman filtering, and fuzzy control, thereby improving the accuracy of the estimation.

[0260] An embodiment of the present invention also provides an electronic device, which may include a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, it implements the above-mentioned method for estimating the state of charge of a battery.

[0261] An embodiment of the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for estimating the state of charge of a battery.

[0262] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0263] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0264] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0265] Embodiments of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0266] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0267] These computer program instructions may also be loaded onto a computer or other programmable data processing terminal equipment to cause a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable terminal equipment, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0268] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0269] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0270] The above provides a detailed description of a method and apparatus for estimating the state of charge of a battery. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for estimating the state of charge of a battery, characterized in that, The method includes: Establish an equivalent circuit model for the battery; Based on the equivalent circuit model, determine the state and observation variable model for the capacitive and resistive parameters; By combining extended Kalman filtering, parameter identification is performed on the state and observed variable model for capacitive and resistive parameters to obtain parameter identification results; Based on the parameter identification results, determine the first state of charge parameter in the current state; Based on the equivalent circuit model, determine the state and observation variable model for the state of charge; Based on the state and observation variable model for the state of charge, the initial state of charge variable corresponding to the first state of charge parameter is determined. Based on the initial charged state variables, multiple charged state particles are sampled using a particle filtering algorithm; During the unscented Kalman filtering of each charged particle, the difference between the theoretical covariance and the actual covariance of the noise is determined, and the adjustment amount is determined based on the difference. Based on the adjustment amount, update the noise covariance matrix, and based on the updated Kalman gain matrix based on the noise covariance matrix, determine the updated charged state particles. Determine the number of valid charged state particles in the updated charged state particles; When the number of effective charged state particles is greater than or equal to the preset number of charged state particles, the charged state variable of the updated charged state particles is determined as the target charged state variable. The second charge state parameter is determined based on the target charge state variable.

2. The method according to claim 1, characterized in that, The parameter identification result is the target capacitive resistance state variable. The parameter identification is performed on the state and observed variable model for the capacitive resistance parameter using extended Kalman filtering, yielding the parameter identification result, including: Based on the state and observation variable model for the capacitive and resistive parameters, the initial state variables are determined, and the target capacitive and resistive state variables are obtained by performing extended Kalman filtering based on the initial state variables.

3. The method according to claim 2, characterized in that, The step of determining the first state-of-charge parameter in the current state based on the parameter identification result includes: Based on the target capacitive and resistive state variables, determine the target capacitive and resistive parameters of the equivalent circuit model; Based on the target capacitive and resistive parameters, determine the first state-of-charge parameters in the current state.

4. The method according to claim 3, characterized in that, The step of determining the first state-of-charge parameter in the current state based on the target capacitive resistance parameter includes: Based on the target capacitive and resistive parameters, determine the current open-circuit voltage of the equivalent circuit model; Based on the current open-circuit voltage, determine the first state-of-charge parameter in the current state.

5. The method according to claim 1, characterized in that, Also includes: When the number of effective charged state particles is less than the preset number of charged state particles, resampling is performed, and for the resampled charged state particles, the unscented Kalman filtering is performed on each charged state particle to obtain updated charged state particles.

6. The method according to claim 1, characterized in that, The equivalent circuit model includes at least three equivalent sub-circuits, which are parallel circuits of capacitors and resistors. The equivalent sub-circuit includes: Parallel circuits of ohmic polarized resistors and ohmic polarized capacitors, parallel circuits of electrochemical polarized resistors and electrochemical polarized capacitors, and parallel circuits of concentration difference polarized resistors and concentration difference polarized capacitors.

7. A battery state-of-charge estimation device, characterized in that, The device includes: Equivalent circuit model establishment circuit, used to establish the equivalent circuit model of the battery; The first state-of-charge parameter determination circuit is used to determine the state and observation variable model for the capacitive and resistive parameters based on the equivalent circuit model; combine extended Kalman filtering to perform parameter identification on the state and observation variable model for the capacitive and resistive parameters to obtain parameter identification results; and determine the first state-of-charge parameter in the current state based on the parameter identification results. The circuit for obtaining the second state-of-charge parameter is used to determine the state and observation variable model for the charged state based on the equivalent circuit model; determine the initial state-of-charge variable corresponding to the first state-of-charge parameter based on the state and observation variable model for the charged state; sample multiple charged state particles based on the initial state-of-charge variable and a particle filtering algorithm; determine the difference between the theoretical covariance and the actual covariance of the noise during the unscented Kalman filtering process for each charged state particle, and determine the adjustment amount based on the difference; update the noise covariance matrix based on the adjustment amount, and determine the updated charged state particles based on the updated Kalman gain matrix based on the noise covariance matrix; determine the number of effective charged state particles in the updated charged state particles; when the number of effective charged state particles is greater than or equal to the preset number of charged state particles, determine the charged state variable of the updated charged state particles as the target charged state variable; and determine the second state-of-charge parameter based on the target charged state variable.

Images