SOC estimation method, device, equipment and storage medium for energy storage system

Abstract

The embodiment of the present invention discloses a method and device, equipment and storage medium for estimating the SOC of an energy storage system. The method includes: based on the charge and discharge characteristics of the energy storage system and the collected current data, voltage data and temperature data of the energy storage system, determining the interaction model of the high-frequency equivalent circuit and the interaction model of the medium- and low-frequency equivalent circuit of the energy storage system, the charge and discharge characteristics at least include the hysteresis characteristics of the charge and discharge of the energy storage system; performing particle filtering on the interaction models respectively to determine the estimated SOC of each interaction model; performing SOC fusion processing using the estimated SOC of each interaction model and the preset interactive multi-model to determine the target SOC of the energy storage system. In the above manner, the interaction model of the high-frequency equivalent circuit and the interaction model of the medium- and low-frequency equivalent circuit of the energy storage system are respectively established, and the estimated SOC of each interaction model is obtained by particle filtering, and then the estimated SOC is fused through the interactive multi-model to make the obtained target SOC of the energy storage system more accurate.

Description

Technical Field

[0001] The present invention relates to the field of battery SOC technology, and in particular to a SOC estimation method and device, equipment and storage medium for an energy storage system. Background Art

[0002] High-frequency power electronic components that can be found everywhere in the power system will generate high-frequency noise and current ripple. However, the energy storage electrochemical batteries used in the power system will also be affected by high-frequency noise and current ripple, which will lead to inaccurate estimation of the state of charge (SOC) of the energy storage battery. Inaccurate battery SOC estimation will lead to errors in the control logic of the battery management system (BMS), which will seriously lead to instability of the power grid system, causing large-scale power outages and huge economic losses.

[0003] The existing energy storage battery SOC estimation method is mainly based on the second-order RC equivalent circuit model. When there is high-frequency noise and current ripple, the second-order RC model cannot well reflect the internal reaction state of the energy storage battery, resulting in inaccurate SOC estimation of the energy storage battery, causing serious consequences. The existing energy storage battery SOC estimation also has a separate equivalent circuit model for high frequency, but when the battery works at medium and low frequency excitation, its high-frequency equivalent circuit model cannot adapt to medium and low frequency excitation well, resulting in inaccurate SOC estimation of the energy storage battery, causing serious consequences.

[0004] Moreover, when estimating the battery SOC, the traditional Kalman filter has better adaptability to linear systems, but has poorer adaptability to nonlinear systems such as battery SOC estimation.

[0005] Therefore, the prior art still lacks a method for improving the accuracy of SOC estimation. Summary of the invention

[0006] The main purpose of the present invention is to provide a SOC estimation method and device, equipment and storage medium for an energy storage system, which can solve the problem of low SOC estimation accuracy in the prior art.

[0007] To achieve the above object, the present invention provides a first aspect of a method for estimating the SOC of an energy storage system, the method comprising:

[0008] Determine an interaction model of a high-frequency equivalent circuit and an interaction model of a medium- and low-frequency equivalent circuit of the energy storage system based on the charge and discharge characteristics of the energy storage system and the collected current data, voltage data, and temperature data of the energy storage system, wherein the charge and discharge characteristics at least include a hysteresis characteristic of charge and discharge of the energy storage system;

[0009] Performing particle filtering processing on the interaction models respectively to determine an estimated SOC of each interaction model;

[0010] The estimated SOC of each interactive model and the preset interactive multi-model are used to perform SOC fusion processing to determine the target SOC of the energy storage system.

[0011] In a feasible implementation, performing particle filtering on the interaction models respectively to determine the estimated SOC of each interaction model includes:

[0012] Using the current data, voltage data, temperature data and a preset forgetting factor recursive least squares algorithm to perform parameter identification on the interaction models respectively, and determine a parameter identification result of each interaction model, wherein the parameter identification result at least includes a model parameter of the interaction model;

[0013] The interaction models are respectively subjected to particle filtering processing using the model parameters of the interaction models to determine an estimated SOC of each interaction model.

[0014] In a feasible implementation, using the model parameters of the interaction model to perform particle filtering on the interaction model respectively to determine the estimated SOC of each interaction model includes:

[0015] Determining the state equation and observation equation of the energy storage system using the model parameters of the interaction model;

[0016] For each interaction model, at the initialization time k=0, particle sampling is performed based on the prior probability of the state equation of the energy storage system to determine the N state particles at the initialization time, and the weight of each of the N state particles at the initialization time is 1 / N; let k=k+1, and enter the iterative process of particle filtering;

[0017] The system state transition probability density at time k is used as the importance probability density at time k, and at time k, the importance probability density is used to perform particle sampling at time k to obtain N state particles at time k;

[0018] Determine the update weights of the N state particles at time k by using the observation values ​​of the observation equations of the N state particles at time k and a preset weight update algorithm;

[0019] Resampling at time k is performed using the update weight and a preset resampling threshold to determine the state particle at time k after filtering;

[0020] According to the filtered state particles at time k and the preset state update algorithm, the state estimation value updated by the interaction model at time k is determined, wherein the state estimation value updated at time k includes at least the estimated SOC of the interaction model at time k, and k=k+1 is set, and the step of using the system state transition probability density at time k as the importance probability density at time k and sampling the particles at time k using the importance probability density at time k to obtain N state particles at time k is returned to execute, until the particle filtering at all times is completed.

[0021] In a feasible implementation, the estimated SOC state of each interactive model and the preset interactive multi-model are used to perform SOC fusion processing to determine the target SOC state of the energy storage system, including:

[0022] Inputting the state estimation value and covariance estimation value updated by each interactive model at time k into the preset interactive multi-model to determine the state estimation value and covariance estimation value of the energy storage system at time k;

[0023] Using the current data, voltage data, temperature data of the energy storage system, the state estimation value and covariance estimation value of the energy storage system at time k, particle filtering is performed on each interaction model to estimate the state estimation value, covariance estimation value, residual estimation value and residual covariance estimation value of each interaction model at time k+1;

[0024] Determine the likelihood function of each interaction model at time k+1 by using the residual estimate, the residual covariance estimate and a preset density function that obeys a Gaussian distribution;

[0025] Using the likelihood function at time k+1 and the preset model probability algorithm, the model probability of each interaction model at time k+1 is determined;

[0026] Perform fusion processing based on the model probability and the state estimation value and covariance estimation value at time k+1 to determine the target state estimation value of the energy storage system at time k+1;

[0027] The target state estimation value and a preset SOC extraction algorithm are used to determine the target SOC of the energy storage system.

[0028] In a feasible implementation, the parameter identification result includes the following mathematical expression:

[0029]

[0030] In the formula, λ is the forgetting factor, θ k+1 is the parameter identification result at the current time k+1, θ k is the input for parameter identification, is the observed value of the k+1 energy storage system at the current moment, y(k+1) is the true feedback value of the k+1 energy storage system at the current moment, K(k+1) is the gain of the k+1 energy storage system at the current moment, P(k+1) is the covariance matrix of the k+1 energy storage system at the current moment, and E is the unit matrix.

[0031] In a feasible implementation, the state update algorithm includes the following mathematical expression:

[0032]

[0033] In the formula, is the state estimate updated at time k, is the covariance estimate updated at time k; N is the total number of state particles; is the i-th state particle at time k; is the weight of the i-th state particle at time k.

[0034] In a feasible implementation, the target state estimation value and the target covariance estimation value of the energy storage system at time k+1 include the following mathematical expressions:

[0035]

[0036]

[0037] In the formula, is the estimated target state of the energy storage system at time k+1; P n (k+1|k+1) is the target covariance estimate of the energy storage system at time k+1; m is the total number of interaction models, is the model probability of interaction model j at time k+1; n is the number of particles in particle filtering; is the estimated value of the model j state at time k+1;

[0038] The SOC extraction algorithm includes the following mathematical expressions:

[0039]

[0040] In the formula, SOC ksum is the target SOC of the energy storage system, is the estimated value of the target state of the energy storage system at time k+1.

[0041] To achieve the above object, the second aspect of the present invention provides a SOC estimation device for an energy storage system, the device comprising:

[0042] Model determination module: used to determine the interaction model of the high-frequency equivalent circuit and the interaction model of the medium- and low-frequency equivalent circuit of the energy storage system based on the charge and discharge characteristics of the energy storage system and the collected current data, voltage data and temperature data of the energy storage system, wherein the charge and discharge characteristics at least include the hysteresis characteristics of the charge and discharge of the energy storage system;

[0043] SOC estimation module: used to perform particle filtering on the interaction models respectively to determine the estimated SOC of each interaction model;

[0044] SOC fusion module: used to perform SOC fusion processing using the estimated SOC of each interactive model and the preset interactive multi-model to determine the target SOC of the energy storage system.

[0045] To achieve the above-mentioned purpose, the third aspect of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps shown in the first aspect or any feasible implementation method.

[0046] To achieve the above-mentioned purpose, the fourth aspect of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps shown in the first aspect or any feasible implementation method.

[0047] The embodiments of the present invention have the following beneficial effects:

[0048] The present invention provides a method for estimating the SOC of an energy storage system, the method comprising: determining an interactive model of a high-frequency equivalent circuit and an interactive model of a medium- and low-frequency equivalent circuit of the energy storage system based on the charging and discharging characteristics of the energy storage system and collected current data, voltage data, and temperature data of the energy storage system, wherein the charging and discharging characteristics at least include hysteresis characteristics of the charging and discharging of the energy storage system; performing particle filtering processing on the interactive models respectively to determine an estimated SOC of each interactive model; and performing SOC fusion processing using the estimated SOC of each interactive model and a preset interactive multi-model to determine a target SOC of the energy storage system.

[0049] Through the above method, a method for estimating the SOC of an energy storage battery based on an interactive multi-model particle filter is provided. First, a high-frequency equivalent circuit and a medium-low frequency equivalent circuit are modeled according to the working frequency of the energy storage battery, so that the established model conforms to the actual charging and discharging characteristics of the battery, and the SOC calculation error caused by the frequency band change of the high frequency and the medium-low frequency is reduced. Then, a particle filter is used to estimate the SOC of the high-frequency equivalent circuit and the medium-low frequency equivalent circuit, and then the fused SOC is calculated by an interactive multi-model algorithm. Since the particle filter adopts a Monte Carlo simulation method and uses a group of random particles to represent the posterior probability density of random variables in the state space, an approximate optimal numerical solution based on a physical model is obtained. It has strong applicability to strongly nonlinear and non-Gaussian noise systems. Compared with traditional Kalman filtering, it can effectively improve the adaptability to nonlinear systems such as battery SOC estimation and improve the accuracy of SOC estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] in:

[0052] Figure 1 This is a flow chart of a method for estimating the SOC of an energy storage system according to an embodiment of the present invention;

[0053] Figure 2 This is a waveform diagram of an experimental test current and open circuit voltage signal in an embodiment of the present invention;

[0054] FIG3( a ) is a particle filter diagram of a high frequency equivalent circuit model according to an embodiment of the present invention;

[0055] FIG3( b ) is a particle filter diagram of a low-frequency equivalent circuit model according to an embodiment of the present invention;

[0056] Figure 4 A diagram showing hysteresis characteristics of charging and discharging of an energy storage battery in an embodiment of the present invention;

[0057] Figure 5 Another flow chart of a method for estimating the SOC of an energy storage system according to an embodiment of the present invention;

[0058] FIG6( a ) is a parameter identification result of a low-frequency equivalent circuit interaction model in an embodiment of the present invention;

[0059] FIG6( b ) is a parameter identification result of a high-frequency equivalent circuit interaction model in an embodiment of the present invention;

[0060] FIG. 7( a ) is a SOC estimation diagram and an error diagram using a particle filter (PF) for a low-frequency equivalent circuit interaction model according to an embodiment of the present invention;

[0061] FIG7( b ) is a SOC estimation diagram and error diagram using a particle filter (PF) in a high-frequency equivalent circuit interaction model according to an embodiment of the present invention;

[0062] Figure 8 A block diagram of an interactive multi-model particle filtering system in an embodiment of the present invention;

[0063] Fig. 9 A battery SOC diagram and error diagram using interactive multi-model particle filtering in an embodiment of the present invention;

[0064] Fig.10 is a structural block diagram of a SOC estimation device for an energy storage system in an embodiment of the present invention;

[0065] Fig.11 4 is a structural block diagram of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0066] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] See also Figure 1 , Figure 1 is a flow chart of a method for estimating the SOC of an energy storage system in an embodiment of the present invention. Figure 1 The method shown includes the following steps:

[0068] 101. Determine an interaction model of a high-frequency equivalent circuit and an interaction model of a medium- and low-frequency equivalent circuit of the energy storage system based on the charge and discharge characteristics of the energy storage system and the collected current data, voltage data, and temperature data of the energy storage system, wherein the charge and discharge characteristics at least include a hysteresis characteristic of charge and discharge of the energy storage system;

[0069] It should be noted that the SOC estimation method of the energy storage system shown in the present application is to perform real-time SOC calculation of the energy storage system so as to dynamically grasp the SOC state of the energy storage system. First of all, in order to calculate the SOC of the energy storage system, it is necessary to establish an equivalent model of the energy storage system. The present application establishes two interaction models of equivalent circuits for the energy storage system according to different frequency bands to improve the accuracy of modeling. Specifically, based on the charging and discharging characteristics of the energy storage system and the collected current data, voltage data and temperature data of the energy storage system, the interaction model of the high-frequency equivalent circuit of the energy storage system and the interaction model of the medium- and low-frequency equivalent circuit are determined. The above-mentioned charging and discharging characteristics at least include the hysteresis characteristics of the charging and discharging of the energy storage system, wherein the energy storage system can be an energy storage battery, etc.

[0070] Taking the energy storage system as an energy storage battery as an example, step 101 may include the following steps Step 1.1 and Step 1.2:

[0071] Step 1.1: Collect the voltage and current of the energy storage battery through the voltage and current acquisition equipment, and collect the pole temperature of the energy storage battery through the temperature sensor.

[0072] See also Figure 2 , Figure 2 This is an experimental test current and open circuit voltage signal waveform diagram in an embodiment of the present invention. This embodiment uses a certain model of lead-carbon battery with a rated capacity of 660Ah as the experimental object. The rated voltage of the battery is 2V, the discharge cut-off voltage is 1.8V, and the maximum charging current is 300A. 0.1C discharge is used, the discharge current is 66A, and each discharge is 5% SOC and then left to stand for 10 minutes. The blue battery test system is used to test the battery and collect charge and discharge voltage and current data. The DS18B20 adhesive temperature sensor is used for temperature measurement. The collected current and open circuit voltage signal waveforms are shown in the figure below. Figure 2 shown.

[0073] Step 1.2: Based on the collected energy storage battery voltage, current fluctuation frequency and the hysteresis characteristics of the battery charging and discharging, a high-frequency equivalent circuit interaction model of the energy storage battery considering the hysteresis characteristics and a medium- and low-frequency equivalent circuit interaction model considering the hysteresis characteristics are established respectively.

[0074] Based on electrochemical impedance spectroscopy, the high-frequency equivalent circuit interaction model considering the hysteresis characteristics is composed of the battery equilibrium potential EMF and the hysteresis voltage U H The voltage source is used to replace the open circuit voltage U of the battery. OCV(SOC,H,T), the ohmic internal resistance of the energy storage battery is replaced by the resistor R0, the high-frequency part of the equivalent circuit is replaced by the inductor L, and the other parts of the equivalent circuit are replaced by the resistor R1 and the capacitor C1 in parallel. For details, please refer to Figure 3(a), which is a high-frequency equivalent circuit model in an embodiment of the present invention. In the high-frequency equivalent circuit interaction model considering the hysteresis characteristic, its open circuit voltage can be expressed as:

[0075] U OCV(SOC,H,T) =IR0+U L +U1+U0;

[0076] Among them, the open circuit voltage U OCV(SOC,H,T) It can be expressed as:

[0077] U OCV(SOC,H,T) =EMF+U H ;

[0078] Its hysteresis voltage U H The equilibrium potential EMF calculation formula is as follows:

[0079]

[0080] Among them, η is 0.5~1, U charge is the equilibrium terminal voltage of the battery during charging, U discharge It is the equilibrium terminal voltage of the battery during discharge.

[0081] Among them, see Figure 4 The charge and discharge hysteresis voltage curve, Figure 4 This is a diagram showing hysteresis characteristics of charging and discharging of an energy storage battery in an embodiment of the present invention.

[0082] Furthermore, based on electrochemical impedance spectroscopy, the low-frequency equivalent circuit interaction model considering the hysteresis characteristics is composed of the battery equilibrium potential EMF and the hysteresis voltage U H The voltage source is used to replace the open circuit voltage U of the battery. OCV(SOC,H,T) , the ohmic internal resistance of the energy storage battery is replaced by the resistor R0, and the medium and low frequency part of the equivalent circuit is replaced by the second-order RC circuit. For details, please refer to Figure 3(b), which is a low-frequency equivalent circuit model in an embodiment of the present invention. In the medium and low frequency equivalent circuit interaction model considering the hysteresis characteristics, its open circuit voltage can be expressed as:

[0083] U OCV(SOC,H,T) =IR0+U1+U2+U0;

[0084] Among them, the open circuit voltage U OCV(SOC,H,T) It can be expressed as:

[0085] U OCV(SOC,H,T) =EMF+U H ;

[0086] Based on the low-frequency and medium-frequency equivalent circuit interaction model considering the hysteresis characteristics, the discrete state space equation can be established as follows:

[0087]

[0088] Among them, τ1=R1C1, τ2=R2C2,

[0089] 102. Perform particle filtering on the interaction models respectively to determine an estimated SOC of each interaction model;

[0090] It should be noted that after obtaining the above interaction model, the particle filter (PF) can be used to perform particle filtering on the above interaction model to determine the estimated SOC of each interaction model. Specifically, the collected voltage, current, and temperature data and the established high-frequency equivalent circuit interaction model considering hysteresis characteristics and the medium- and low-frequency equivalent circuit interaction model considering hysteresis characteristics are used to respectively estimate the SOC using particle filtering (PF).

[0091] The so-called particle filter refers to the process of finding a group of random samples propagating in the state space to approximate the probability density function, replacing the integral operation with the sample mean, and then obtaining the minimum variance estimate of the system state. These samples are figuratively called "particles", so it is called particle filtering.

[0092] Compared with Kalman Filter, the idea of ​​particle filtering is based on Monte Carlo methods, which uses particle sets to represent probabilities and can be used in any form of state space model. Its core idea is to express its distribution through random state particles extracted from the posterior probability, which is a sequential importance sampling method. In simple terms, particle filtering refers to the process of approximating the probability density function by finding a set of random samples propagating in the state space, replacing the integral operation with the sample mean, and thus obtaining the state minimum variance distribution. The sample here refers to the particle, and when the number of samples N→∝, it can approximate any form of probability density distribution.

[0093] Although the probability distribution in the algorithm is only an approximation of the real distribution, due to its non-parametric characteristics, it gets rid of the constraint that the random quantity must satisfy the Gaussian distribution when solving nonlinear filtering problems. It can express a wider distribution than the Gaussian model and has a stronger modeling ability for the nonlinear characteristics of variable parameters. Therefore, the particle filter can more accurately express the posterior probability distribution based on the observed and controlled quantities, and can be used to solve SLAM problems.

[0094] 103. Perform SOC fusion processing using the estimated SOC of each interactive model and the preset interactive multi-model to determine the target SOC of the energy storage system.

[0095] After obtaining the estimated SOC of each interactive model, the estimated SOC of each interactive model and the preset interactive multiple model (IMM) can be used for SOC fusion processing to determine the target SOC of the energy storage system, that is, the SOC state of the energy storage system. In this way, the estimated SOC of the two equivalent models can be fused in real time to achieve real-time calculation of the SOC of the energy storage system.

[0096] Among them, the main idea of ​​the interactive multi-model control algorithm is the automatic identification and switching between models based on Bayesian theory: at any tracking moment, real-time maneuver model detection is performed by setting model filters corresponding to the number of possible target models, setting weight coefficients and model update probabilities for each filter, and finally weighted calculation to obtain the current optimal estimation state, thereby achieving the purpose of model adaptive tracking.

[0097] Therefore, the real-time SOC of the energy storage battery is calculated by fusing the two interactive models through the interactive multi-model (IMM) mentioned above, which has higher calculation accuracy, better extends the battery life and better maintains the safe and stable operation of the power grid.

[0098] The present invention provides a method for estimating the SOC of an energy storage system, the method comprising: determining an interactive model of a high-frequency equivalent circuit and an interactive model of a medium- and low-frequency equivalent circuit of the energy storage system based on the charging and discharging characteristics of the energy storage system and collected current data, voltage data, and temperature data of the energy storage system, wherein the charging and discharging characteristics at least include hysteresis characteristics of the charging and discharging of the energy storage system; performing particle filtering processing on the interactive models respectively to determine an estimated SOC of each interactive model; and performing SOC fusion processing using the estimated SOC of each interactive model and a preset interactive multi-model to determine a target SOC of the energy storage system. Through the above method, a method for estimating the SOC of an energy storage battery based on an interactive multi-model particle filter is provided. First, a high-frequency equivalent circuit and a medium-low frequency equivalent circuit are modeled according to the working frequency of the energy storage battery, so that the established model conforms to the actual charging and discharging characteristics of the battery, and the SOC calculation error caused by the frequency band change of the high frequency and the medium-low frequency is reduced. Then, a particle filter is used to estimate the SOC of the high-frequency equivalent circuit and the medium-low frequency equivalent circuit, and then the fused SOC is calculated by an interactive multi-model algorithm. Since the particle filter adopts a Monte Carlo simulation method and uses a group of random particles to represent the posterior probability density of random variables in the state space, an approximate optimal numerical solution based on a physical model is obtained. It has strong applicability to strongly nonlinear and non-Gaussian noise systems. Compared with traditional Kalman filtering, it can effectively improve the adaptability to nonlinear systems such as battery SOC estimation and improve the accuracy of SOC estimation.

[0099] See also Figure 5 , Figure 5 FIG. 4 is another flow chart of a method for estimating the SOC of an energy storage system according to an embodiment of the present invention. Figure 5 The method shown includes the following steps:

[0100] 501. Determine an interaction model of a high-frequency equivalent circuit and an interaction model of a medium- and low-frequency equivalent circuit of the energy storage system based on the charge and discharge characteristics of the energy storage system and the collected current data, voltage data, and temperature data of the energy storage system, wherein the charge and discharge characteristics at least include a hysteresis characteristic of charge and discharge of the energy storage system;

[0101] It should be noted that step 501 and Figure 1 The content of step 101 is similar and will not be described here to avoid repetition. For details, please refer to the content of the aforementioned step 101.

[0102] 502. Perform parameter identification on the interaction models respectively using the current data, voltage data, temperature data, and a preset forgetting factor recursive least squares algorithm to determine a parameter identification result of each interaction model, wherein the parameter identification result at least includes a model parameter of the interaction model;

[0103] It should be noted that when using particle filtering to determine the estimated SOC of each interaction model, the forgetting factor recursive least squares method (FFRLS) is first used to identify the parameters of the interaction model, and then the SOC estimation is performed using the identification results. Therefore, for each interaction model, parameter identification must be performed through step 502, that is, the collected voltage, current, and temperature data are used to perform real-time parameter identification on the established high-frequency equivalent circuit interaction model considering hysteresis characteristics and the medium- and low-frequency equivalent circuit interaction model considering hysteresis characteristics using the forgetting factor recursive least squares (FFRLS) to obtain the model parameters of each model, so as to facilitate the subsequent SOC estimation based on the determined model parameters.

[0104] Exemplarily, the calculation formula for the above parameter identification can be derived as the following mathematical expression:

[0105]

[0106] In the formula, λ is the forgetting factor, which is generally 0.95-1 in energy storage batteries, and θ k+1 is the parameter identification result at the current time k+1, θ k is the input for parameter identification, is the observed value of the k+1 energy storage system at the current moment, y(k+1) is the true feedback value of the k+1 energy storage system at the current moment, K(k+1) is the gain of the k+1 energy storage system at the current moment, P(k+1) is the covariance matrix of the k+1 energy storage system at the current moment, and E is the unit matrix.

[0107] For example, please refer to FIG. 6(a), which is a parameter identification result of a low-frequency equivalent circuit interaction model in an embodiment of the present invention; wherein, the resistance R0 fluctuates between 0.05-0.055Ω, the resistance R1 fluctuates between 0.055-0.065Ω, the resistance R2 fluctuates between 0.11-0.12Ω, the capacitance C1 fluctuates between 1.8-2.0F, and the capacitance C2 fluctuates between 4.5-5F. Further, FIG. 6(b) is a parameter identification result of a high-frequency equivalent circuit interaction model in an embodiment of the present invention. As shown in FIG. 6(b), the resistance R0 fluctuates between 0.01-0.015Ω, the resistance R1 fluctuates between 0.001-0.0012Ω, the inductance L fluctuates between 0.01-0.015H, and the capacitance C1 fluctuates between 40-45F. It can be understood that this embodiment only shows an embodiment in which more than 4,000 data are collected to illustrate the beneficial effects of the present application. In practice, the amount of data collected will increase accordingly as the real-time collection time increases. This embodiment is only used as an example without any specific limitation.

[0108] 503. Perform particle filtering on the interaction models respectively using the model parameters of the interaction models to determine an estimated SOC of each interaction model;

[0109] Furthermore, after obtaining the real-time parameter identification results of the above interaction model, the interaction model can be example filtered using the real-time model parameters to obtain an estimated SOC of the interaction model.

[0110] In a feasible implementation, the state equation of the energy storage system needs to be filtered and updated. Therefore, step 503 includes A1-A6:

[0111] A1. Determine the state equation and observation equation of the energy storage system using the model parameters of the interaction model;

[0112] First, the state equation and observation equation of the energy storage system can be determined by the above model parameters. The state equation and observation equation of the system are:

[0113]

[0114] In the formula, A k-1 is the system state transfer matrix at time k-1, B k-1 is the system control matrix at time k-1, W k-1 is the system process noise at time k-1, C k is the observation matrix of the system at time k, D k is the direct connection matrix at time k, V k is the observation noise at time k, x k is the state variable at time k, u k Enter the variables for time k.

[0115] A2. For each interaction model, at the initialization time k=0, particle sampling is performed based on the prior probability of the state equation of the energy storage system to determine the N state particles at the initialization time, and the weight of each of the N state particles at the initialization time is 1 / N; let k=k+1 and enter the iterative process of particle filtering;

[0116] It should be noted that for each interaction model, at the initialization time k = 0, there is only one data point. Therefore, in order to sample particles at the initialization time, particle sampling is performed based on the prior probability of the state equation at the initialization time to obtain N state particles at the initialization time. And will The weight of each particle in is set to 1 / N to update the state equation x0 at time k = 0, and the updated Afterwards, let k=k+1 and enter the iterative process of particle filtering.

[0117] in,

[0118] In the formula, the function E[·] is used to find the expectation, P0 i This is the covariance at this time.

[0119] A3. The system state transition probability density at time k is used as the importance probability density at time k, and at time k, the importance probability density is used to perform particle sampling at time k to obtain N state particles at time k;

[0120] Furthermore, when k = 1, 2, 3, ..., N, the state particles at time k are updated by importance sampling. After initialization, the system state transition probability density at time k is used as the importance probability density at time k, and at time k, the importance probability density is used to perform particle sampling at time k to obtain N state particles x at time k. k i , i=1,2,……,N 。

[0121] For example, importance sampling takes the system state transition probability density as the importance probability density, and samples the importance probability density at time k to obtain x k i , i=1,2,…,N.

[0122] q(x k ∣x 0:k-1 (i),z 1:k )=p(x k ∣x k-1 (i));

[0123] In the formula, p(x k ∣x k-1 (i)) is the system state transition probability density, q(x k ∣x 0:k-1 (i),z 1:k ) is the importance probability density.

[0124] A4. Determine the update weights of the N state particles at time k by using the observation values ​​of the observation equations of the N state particles at time k and the preset weight update algorithm;

[0125] Furthermore, step A4 is used to update and normalize the weights, update the weights at time k, and then perform normalization.

[0126]

[0127] In the formula, w k i is the updated x at time k k i Weight, is the normalized x at time k k i Weight.

[0128] A5. Resampling at time k is performed using the updated weight and a preset resampling threshold to determine the state particle at time k after filtering;

[0129] Furthermore, the updated weights and the preset resampling threshold are used to perform resampling at time k to determine the state particle at time k after filtering. Specifically, for resampling, a resampling threshold is set, and the particle weight is compared with the threshold. If the particle weight is less than the threshold, the particle is eliminated, and the original particle is resampled to a particle with an equal weight, otherwise the amplitude becomes a new particle.

[0130] A6. According to the filtered state particles at time k and the preset state update algorithm, determine the state estimate value updated by the interaction model at time k, wherein the state estimate value updated at time k includes at least the estimated SOC of the interaction model at time k, set k=k+1, return to execute the step of using the system state transition probability density at time k as the importance probability density at time k, and at time k, use the importance probability density to perform particle sampling at time k to obtain N state particles at time k, until particle filtering at all times is completed.

[0131] Finally, the state value at time k is updated by using the filtered particles in step A6. According to the filtered state particles at time k and the preset state update algorithm, the state estimate value updated by the interaction model at time k is determined, and the battery state is updated and the estimated covariance is updated. Exemplarily, the state update algorithm includes the following mathematical expression:

[0132]

[0133] In the formula, is the state estimate updated at time k, is the covariance estimate updated at time k; N is the total number of state particles; is the i-th state particle at time k; is the weight of the i-th state particle at time k.

[0134] It should be noted that the medium and low frequency equivalent circuit interaction model uses the identified parameters and the voltage and current obtained by the test, and uses particle filtering to estimate the SOC and the waveform diagram and SOC estimation error are shown in Figure 7(a). Figure 7(a) is a low frequency equivalent circuit interaction model in an embodiment of the present invention using a particle filter (PF) SOC estimation diagram and error diagram. It can be seen from the error diagram that the error of the sampled medium and low frequency equivalent circuit interaction model is smaller than the error of the ampere-hour integration method.

[0135] Among them, the high-frequency equivalent circuit interaction model uses the identified parameters and the voltage and current obtained by the test, and uses particle filtering to estimate the SOC. The waveform diagram and SOC estimation error are shown in Figure 7(b). Figure 7(b) is a high-frequency equivalent circuit interaction model in an embodiment of the present invention using a particle filter (PF) SOC estimation diagram and error diagram. It can be seen from the figure that the SOC estimated by the high-frequency equivalent circuit interaction model is more accurate than the SOC estimated by the ampere-hour integration method. It can be seen from the error diagram that the estimation error of the ampere-hour integration method is increasing.

[0136] 504. Perform SOC fusion processing using the estimated SOC of each interactive model and the preset interactive multi-model to determine the target SOC of the energy storage system.

[0137] It should be noted that step 504 and Figure 1 The contents of step 103 are similar to those shown in the figure, so they are not described here in detail to avoid repetition. Figure 1 The content of step 103 is shown.

[0138] In a feasible implementation, the SOC estimated by each model is fused by using an interactive multi-model (IMM) in step 504 to obtain a fused SOC, which may include the following steps B1-B6:

[0139] B1. Inputting the state estimation value and covariance estimation value updated by each interactive model at time k into the preset interactive multi-model to determine the state estimation value and covariance estimation value of the energy storage system at time k;

[0140] It should be noted that the state estimation values ​​of the two models will be obtained at each moment. In order to obtain the SOC state of the energy storage battery, it is necessary to fuse the data of the two models at the same moment. Specifically, the state estimation value and covariance estimation value updated by each interactive model at moment k are input into the preset interactive multi-model to determine the state estimation value and covariance estimation value of the energy storage system at moment k.

[0141] See also Figure 8 , Figure 8 This is a block diagram of an interactive multi-model particle filter system in an embodiment of the present invention, that is, the first step is performed through step B1: IMM input interaction, that is, the model conditions are initialized or reinitialized to obtain the state vector and covariance matrix of each particle filter input of the model at the current moment. For model j, at time k, the state estimation value and covariance estimation value of the energy storage system at time k are as follows:

[0142]

[0143]

[0144]

[0145]

[0146] In the above formula: is the estimated value of the state of the nth particle in model j at the kth moment; π lj is the probability of switching from interaction model j to interaction model l; is the value of the covariance matrix at time k; is the probability of model j at time k, m is the total number of interactive models, n is the number of particles in the particle filter, and N is the total number of particles at each time.

[0147] B2. Use the current data, voltage data, temperature data of the energy storage system, the state estimation value and covariance estimation value of the energy storage system at time k to perform particle filtering on each interaction model, and estimate the state estimation value, covariance estimation value, residual estimation value and residual covariance estimation value of each interaction model at time k+1;

[0148] Furthermore, after obtaining the state particles updated by the model probability, another particle filter is performed. Specifically, in the second step, B2 uses the current data, voltage data, temperature data, and state estimation value of the energy storage system at time k to obtain the state particles updated by the model probability. and the covariance estimates Perform particle filtering on each interaction model separately to estimate the state estimation value of each interaction model at time k+1 Covariance estimate Residual estimates and residual covariance estimates For example, for the interaction model j=1, ... N, the voltage, current, pole temperature, as well as As input for particle filtering, using period k and Get the state and covariance estimate of the next period k+1 and The residuals and their covariance are

[0149] Furthermore, the model probability is updated by a likelihood function, and the specific details are as follows.

[0150] B3, using the residual estimate, the residual covariance estimate and the preset density function obeying the Gaussian distribution, determine the likelihood function of each interaction model at time k+1;

[0151] Exemplarily, the likelihood function of the j-th model can be expressed as:

[0152]

[0153] Where N[·] represents the density function that obeys the Gaussian distribution, is the likelihood function of model j at time k+1.

[0154] B4. Determine the model probability of each interaction model at time k+1 by using the likelihood function at time k+1 and the preset model probability algorithm;

[0155] Exemplarily, the model probability algorithm formula at the model k+1 time is as follows:

[0156]

[0157]

[0158] In the formula, is the model probability of interaction model j at time k+1, is the likelihood function of the interaction model j at time k+1, is the model probability of interaction model j at time k, and n is the number of particles in the particle filter.

[0159] B5. Perform fusion processing based on the model probability and the state estimation value and covariance estimation value at time k+1 to determine the target state estimation value of the energy storage system at time k+1;

[0160] B6. Determine the target SOC of the energy storage system using the target state estimation value and a preset SOC extraction algorithm.

[0161] Finally, the IMM output interaction is performed through step B5 to obtain the target state estimation value. The fused SOC is output through B6. Exemplarily, the target state estimation value and the target covariance estimation value of the energy storage system at time k+1 include the following mathematical expressions:

[0162]

[0163]

[0164] In the formula, is the estimated target state of the energy storage system at time k+1; P n (k+1k+1) is the target covariance estimate of the energy storage system at time k+1; m is the total number of interaction models, is the model probability of interaction model j at time k+1; n is the number of particles in particle filtering; is the estimated value of the model j state at time k+1;

[0165] The SOC extraction algorithm includes the following mathematical expressions:

[0166]

[0167] In the formula, SOC ksum is the target SOC of the energy storage system, is the estimated target state of the energy storage system at time k+1

[0168] It should be noted that is a state matrix, which includes the SOC state. Therefore, the target SOC of the energy storage battery can be extracted through the above algorithm.

[0169] In this embodiment, by using IMM to interact the covariance matrix, model probability and state equation of the medium-low frequency equivalent circuit interaction model and the high-frequency equivalent circuit interaction model, a more accurate battery SOC estimation value and its error are finally calculated, such as Fig. 9 As shown, Fig. 9 This is a battery SOC diagram and error diagram using interactive multi-model particle filtering in an embodiment of the present invention. From the IMM filtering error diagram, it can be seen that the battery SOC estimation error using IMM filtering is the smallest, tending to be within 0.001, and the estimation effect is much better than a single filtering estimation model.

[0170] The present invention provides a SOC estimation method for an energy storage system, which is an energy storage battery SOC estimation method based on interactive multi-model particle filtering. The method firstly models a high-frequency equivalent circuit and a medium-low frequency equivalent circuit according to the working frequency of the energy storage battery, and then uses particle filtering to estimate the SOC of the high-frequency equivalent circuit and the medium-low frequency equivalent circuit after inputting voltage, current and temperature data. Then, the fused SOC is calculated by an interactive multi-model algorithm. As a result, the SOC estimation is more accurate, the battery service life is better extended, and the safe and stable operation of the power grid is better maintained.

[0171] See also Fig.10 , Fig.10 FIG. 1 is a structural block diagram of a SOC estimation device for an energy storage system according to an embodiment of the present invention. Fig.10 The device shown comprises:

[0172] Model determination module 1001: used to determine the interaction model of the high-frequency equivalent circuit and the interaction model of the medium- and low-frequency equivalent circuit of the energy storage system based on the charge and discharge characteristics of the energy storage system and the collected current data, voltage data and temperature data of the energy storage system, wherein the charge and discharge characteristics at least include the hysteresis characteristics of the charge and discharge of the energy storage system;

[0173] SOC estimation module 1002: used to perform particle filtering processing on the interaction models respectively to determine the estimated SOC of each interaction model;

[0174] SOC fusion module 1003: used to perform SOC fusion processing using the estimated SOC of each interactive model and the preset interactive multi-model to determine the target SOC of the energy storage system.

[0175] It should be noted that Fig.10 The functions of each module in the device shown are Figure 1 The contents of each step in the method shown are similar, so they are not described here to avoid repetition. For details, please refer to Figure 1 The content of each step in the method shown.

[0176] The present invention provides an SOC estimation device for an energy storage system, the device comprising: a model determination module: used for determining an interaction model of a high-frequency equivalent circuit and an interaction model of a medium- and low-frequency equivalent circuit of an energy storage system based on the charge and discharge characteristics of the energy storage system and collected current data, voltage data and temperature data of the energy storage system, wherein the charge and discharge characteristics at least include the hysteresis characteristics of the charge and discharge of the energy storage system; an SOC estimation module: used for performing particle filter processing on the interaction models respectively to determine the estimated SOC of each interaction model; and an SOC fusion module: used for performing SOC fusion processing using the estimated SOC of each interaction model and a preset interactive multi-model to determine the target SOC of the energy storage system. Through the above method, a method for estimating the SOC of an energy storage battery based on an interactive multi-model particle filter is provided. First, a high-frequency equivalent circuit and a medium-low frequency equivalent circuit are modeled according to the working frequency of the energy storage battery, so that the established model conforms to the actual charging and discharging characteristics of the battery, and the SOC calculation error caused by the frequency band change of the high frequency and the medium-low frequency is reduced. Then, a particle filter is used to estimate the SOC of the high-frequency equivalent circuit and the medium-low frequency equivalent circuit, and then the fused SOC is calculated by an interactive multi-model algorithm. Since the particle filter adopts a Monte Carlo simulation method and uses a group of random particles to represent the posterior probability density of random variables in the state space, an approximate optimal numerical solution based on a physical model is obtained. It has strong applicability to strongly nonlinear and non-Gaussian noise systems. Compared with traditional Kalman filtering, it can effectively improve the adaptability to nonlinear systems such as battery SOC estimation and improve the accuracy of SOC estimation.

[0177] Fig.11 FIG. 1 shows an internal structure diagram of a computer device in an embodiment. The computer device may be a terminal or a server. Fig.11As shown, the computer device includes a processor, a memory and a network interface connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor can implement the above method. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor can implement the above method. Those skilled in the art can understand that Fig.11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0178] In one embodiment, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the following steps: Figure 1 or Figure 2 The steps of the method are shown.

[0179] In one embodiment, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the processor performs the following steps: Figure 1 or Figure 2 The steps of the method are shown.

[0180] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0181] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0182] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for estimating the SOC of an energy storage system, characterized in that: The method comprises: Determine an interaction model of a high-frequency equivalent circuit and an interaction model of a medium- and low-frequency equivalent circuit of the energy storage system based on the charge and discharge characteristics of the energy storage system and the collected current data, voltage data, and temperature data of the energy storage system, wherein the charge and discharge characteristics at least include a hysteresis characteristic of charge and discharge of the energy storage system; Performing particle filtering processing on the interaction models respectively to determine an estimated SOC of each interaction model; Performing SOC fusion processing using the estimated SOC of each interactive model and the preset interactive multi-model to determine the target SOC of the energy storage system; The performing particle filtering on the interaction models respectively to determine the estimated SOC of each interaction model includes: Using the current data, voltage data, temperature data and a preset forgetting factor recursive least squares algorithm to perform parameter identification on the interaction models respectively, and determine a parameter identification result of each interaction model, wherein the parameter identification result at least includes a model parameter of the interaction model; Using the model parameters of the interaction models, respectively, the interaction models are subjected to particle filtering processing to determine an estimated SOC of each interaction model; The method of using the model parameters of the interaction model to perform particle filtering on the interaction model respectively to determine the estimated SOC of each interaction model includes: Determining the state equation and observation equation of the energy storage system using the model parameters of the interaction model; For each interaction model, at the initialization time k =0, perform particle sampling based on the prior probability of the state equation of the energy storage system to determine the N state particles at the initialization time, and the weight of each of the N state particles at the initialization time is 1 / N; k = k +1, enter the iterative process of particle filtering; Will k The system state transition probability density at time k The probability density of the importance of the moment, and k At this moment, use the importance probability density to k The particle sampling at time k N state particles at the moment; use k The observation values ​​of the observation equations of the N state particles at the moment and the preset weight update algorithm determine k Update weights of N state particles at the moment; Using the update weight and the preset resampling threshold k Resampling at the moment to determine the filtered k Moment state particle; According to the filtered k The state particles at the moment and the preset state update algorithm determine the interaction model at k The state estimate value updated at all times, k The state estimate updated at any time includes at least the interaction model k The estimated SOC at the moment, let k = k+ 1, return to execute the k The system state transition probability density at time k The probability density of the importance of the moment, and k At this moment, use the importance probability density to k The particle sampling at time k The steps of N state particles at each moment are repeated until the particle filtering at all moments is completed.

2. The method according to claim 1, characterized in that: The method of using the estimated SOC state of each interactive model and the preset interactive multi-model to perform SOC fusion processing to determine the target SOC state of the energy storage system includes: Each interaction model is k The updated state estimates and covariance estimates are input into the preset interactive multi-model to determine k The state estimation value and covariance estimation value of the energy storage system at the moment; Using the current data, voltage data, temperature data, k The state estimation value and covariance estimation value of the energy storage system at the time are particle filtered for each interaction model respectively, and it is estimated that k +1 time for each interaction model's state estimate, covariance estimate, residual estimate, and residual covariance estimate; The residual estimation value, the residual covariance estimation value and the preset density function obeying the Gaussian distribution are used to determine the interaction model in k Likelihood function at time +1; use k +1 time likelihood function and the preset model probability algorithm determine k +1 model probability of each interaction model; According to the model probability and k The state estimation value and covariance estimation value at time +1 are fused to determine the energy storage system k +1 time point target state estimate; The target state estimation value and a preset SOC extraction algorithm are used to determine the target SOC of the energy storage system.

3. The method according to claim 1, characterized in that: The parameter identification result includes the following mathematical expression: ； In the formula, λ is the forgetting factor, is the parameter identification result at the current time k+1, is the input for parameter identification, is the observed value of the k+1 energy storage system at the current moment, y ( k +1) is the current time k +1 The actual feedback value of the energy storage system, K ( k +1) is the current time k +1 Energy Storage System Gain, P ( k +1) is the current time k +1 energy storage system covariance matrix, E is the identity matrix.

4. The method according to claim 1, characterized in that: The state update algorithm includes the following mathematical expression: In the formula, for k The state estimate is updated all the time. for k The covariance estimate is updated all the time; N is the total number of particles in the state; for k The moment i state particles; for k The moment i The weight of the particle in each state.

5. The method according to claim 2, characterized in that: The energy storage system k The target state estimate and target covariance estimate at time +1 include the following mathematical expressions: ； ； In the formula, For energy storage systems k +1 time point target state estimate; For energy storage systems k +1 time point target covariance estimate; m is the total number of interaction models, for k +1 moment interaction model j The model probability of n is the number of particles in the particle filter; for k +1 Moment Model j State estimates; The SOC extraction algorithm includes the following mathematical expressions: ； In the formula, SOC ksum is the target SOC of the energy storage system, For energy storage systems k The estimated value of the target state at time +1.

6. A SOC estimation device for an energy storage system, characterized in that: The device comprises: Model determination module: used to determine the interaction model of the high-frequency equivalent circuit and the interaction model of the medium- and low-frequency equivalent circuit of the energy storage system based on the charge and discharge characteristics of the energy storage system and the collected current data, voltage data and temperature data of the energy storage system, wherein the charge and discharge characteristics at least include the hysteresis characteristics of the charge and discharge of the energy storage system; SOC estimation module: used to perform particle filtering on the interaction models respectively to determine the estimated SOC of each interaction model; SOC fusion module: used to perform SOC fusion processing using the estimated SOC of each interactive model and the preset interactive multi-model to determine the target SOC of the energy storage system; The SOC estimation module is specifically used to use the current data, voltage data, temperature data and a preset forgetting factor recursive least squares algorithm to perform parameter identification on the interaction model respectively, and determine the parameter identification result of each interaction model, wherein the parameter identification result at least includes the model parameters of the interaction model; and use the model parameters of the interaction model to perform particle filtering on the interaction model respectively, and determine the estimated SOC of each interaction model; The method of using the model parameters of the interaction model to perform particle filtering on the interaction model to determine the estimated SOC of each interaction model includes: using the model parameters of the interaction model to determine the state equation and observation equation of the energy storage system; for each interaction model, at the initialization time k =0, perform particle sampling based on the prior probability of the state equation of the energy storage system to determine the N state particles at the initialization time, and the weight of each of the N state particles at the initialization time is 1 / N; k = k +1, enter the iterative process of particle filtering; k The system state transition probability density at time k The probability density of the importance of the moment, and k At this moment, use the importance probability density to k The particle sampling at time k N state particles at the moment; using k The observation values ​​of the observation equations of the N state particles at the moment and the preset weight update algorithm determine k The update weights of the N state particles at the time; using the update weights and the preset resampling threshold to perform k Resampling at the moment to determine the filtered k The state of particles at each moment; according to the filtered k The state particles at the moment and the preset state update algorithm determine the interaction model at k The state estimate value updated at all times, k The state estimate updated at any time includes at least the interaction model k The estimated SOC at the moment, let k = k+ 1, return to execute the k The system state transition probability density at time k The probability density of the importance of the moment, and k At this moment, use the importance probability density to k The particle sampling at time k The steps of N state particles at each moment are repeated until the particle filtering at all moments is completed.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 5.

8. A computer device comprising a memory and a processor, characterized in that: The memory stores a computer program, and when the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 5.

Images