CN115598535A - Battery state of charge estimation method and system considering state of health - Google Patents

Abstract

The invention discloses a method and system for estimating the battery state of charge considering the state of health, and belongs to the technical field of battery state of charge estimation. Set the battery model to estimate the open circuit voltage of the battery; use the estimated open circuit voltage and battery health state as the input of the neural network model, and obtain the SOC through preliminary estimation; combine the current information to obtain the final estimated value of SOC through the filtering algorithm. The invention combines a battery model method and a data-driven method to realize accurate estimation of the state of charge of the battery under different health states.

Description

A method and system for estimating battery state of charge considering state of health

technical field

The invention belongs to the technical field of battery state-of-charge estimation, and in particular relates to a battery state-of-charge estimation method and system considering the state of health.

Background technique

The state of charge (SOC) of the battery is an important parameter used to characterize the battery's short-term power changes. It is generally defined as the ratio of the current remaining power to the current maximum available capacity. SOC estimation is one of the core functions of the battery management system. Accurate SOC estimation can improve battery energy utilization, prevent battery overcharge and overdischarge, ensure battery safety and reliability during use, and prolong battery life. However, it is difficult to obtain high-precision SOC estimation results due to the nonlinear and time-varying characteristics of the battery body and the complex working environment.

Commonly used battery SOC estimation methods mainly include: ampere-hour integration method, open circuit voltage method, equivalent model method, and data-driven method. Among them, the calculation of the ampere-hour integration method is simple and efficient, but there is an obvious error accumulation effect; the open circuit voltage method requires the battery to stand still for more than half an hour, which is difficult to apply in practice; the equivalent model method usually uses a battery model plus a filtering algorithm to observe SOC estimation can be achieved by a device, which can achieve a good balance between estimation accuracy and computational complexity, but its applicability and generalization ability are limited; data-driven methods usually use current and voltage information as input to establish machine learning or deep learning models to directly estimate SOC, the estimation accuracy is high, but this method has the risk of open-loop divergence, the amount of calculation is large and it is difficult to run online, and the established model is a black box model, which lacks interpretability. At the same time, the interpretability of the data-driven method to estimate SOC has not been fully obtained. Research.

Most of the above estimation methods do not consider how to continue to ensure the SOC accuracy in the case of battery capacity decay, that is, as the battery health status changes, the SOC estimation error gradually increases.

Contents of the invention

In order to overcome the shortcomings of the above-mentioned prior art, the object of the present invention is to provide a method and system for estimating the battery state of charge considering the state of health, which can effectively solve the problem of increasing errors in SOC estimation as the state of health changes, and data-driven methods. Open-loop divergence risks and technical issues with poor explainability.

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

The invention discloses a battery state of charge (SOC) estimation method considering the state of health (SOH), comprising the following steps:

Step 1: Use public data or charge and discharge equipment to obtain battery charge and discharge data and their corresponding health status, and build a data set;

Step 2: Establish the corresponding battery model;

Step 3: Estimate the open circuit voltage of the battery based on the data set in step 1 and the battery model in step 2;

Step 4: Construct a neural network model with open circuit voltage (OCV) and state of health as input and battery state of charge as output;

Step 5: Combining the current information in the data set in step 1 and the estimated initial value of the battery state of charge output by the neural network model in step 3, perform fusion through a filtering algorithm to obtain the final estimated value of the battery state of charge.

Preferably, in step 1, the battery charging and discharging data and corresponding health status are obtained by using public data or through charging and discharging equipment.

Preferably, the charging and discharging device is a battery detection system or a charging pile.

Preferably, the battery is a lead-acid battery, a nickel-metal hydride battery, a nickel-chromium battery or a lithium-ion battery.

Preferably, in step 2, building a battery model includes the following steps:

Step 21, establish a state space equation based on the first-order RC equivalent circuit model of the battery:

Among them, E ₀ is the open circuit voltage, R ₁ is the ohmic internal resistance, R ₂ is the polarization internal resistance, C ₂ is the polarization capacitance, U ₂ is the polarization voltage, I is the current, and U ₀ is the battery terminal voltage;

Step 22. Calculate the transfer function of the state space equation, and add a zero-order retainer link to calculate the transfer function G(s):

Among them, T is the sampling interval, H(s) is the transfer function of the zero-order holder, s is a variable symbol; e is an irrational number;

Step 23, perform an inverse Laplace transform on the transfer function G(s), and perform discretization processing:

Find the difference equation at time k:

Step 24, combined with formula (1), calculate the open circuit voltage at time k:

in,

Step 25. Set V _k =E _0(k) -U _0(k) to establish a controlled autoregressive moving average model of the battery, namely the CARMA model:

Wherein, θ ₁ =a, θ ₂ =[R ₁ +R ₂ (1/a-1)], θ ₃ =aR ₀ , q ^-1 is a one-step translation operator, and ξ _k is system noise.

Preferably, in step 3, the battery CARMA model parameters are estimated online using the recursive least squares method with a forgetting factor, including the following steps:

Step 31. Since the open circuit voltage changes very little in a very short time, it is assumed that the open circuit voltage does not change in a sampling interval, that is, E _0(k) =E _0(k-1) , and according to the battery CARMA model, it is obtained:

U _0(k) ＝[U _0(k-1) I _k -I _k-1 1][θ _1k θ _2k θ _3k θ _4k ] ^T (26)

Among them, θ _4k =(1-a)E _0(k) , thus obtain the estimated value of the open circuit voltage E _0(k) :

Step 32. Use the recursive least squares method with forgetting factor to estimate the open circuit voltage online, and initialize the estimated value:

Among them, x is the parameter to be estimated, x ₀ is the initial value of the parameter to be estimated, P ₀ is the initial value of the covariance matrix of the error, and E is the identity matrix;

Step 33, perform an iterative solution, and estimate the open circuit voltage according to formula (9), the iterative process is as follows:

e _k =y _k -H _k x _k-1 (29)

x _k ＝x _k-1 +K _k e _k (31)

Among them, k refers to time k, y _k is the battery terminal voltage, e _k is the expected error of y _k , H _k = [U _0(k-1) I _k -I _k-1 1], x _k = [ θ _1k θ _2k θ _3k θ _4k ] ^T is the parameter array to be estimated, K _k is the gain coefficient, P _k is the covariance moment, and λ is the forgetting factor.

Preferably, in step 4, the concrete method of constructing neural network model includes:

Construct a BP neural network model including an input layer, a hidden layer and an output layer. The input feature sequence is x=[OCVSOH] ^T , which is the open circuit voltage and state of health of the battery, and the output sequence is y=SOC, which is the charge of the battery state, the number of neurons in the hidden layer is set to 64, and the activation function is the ReLU function. The output of the entire network for any sample is:

Among them, β _i is the connection weight between the i-th neuron in the hidden layer and the output layer, g(x) is the activation function, a _i is the i-th row of the connection weight matrix between the input layer and the hidden layer, x _j is the input matrix of the jth sample, b _i is the ith row of the bias matrix connecting the input layer and the hidden layer, b _i ⁱ is the output layer bias value, and P is the number of samples.

Preferably, in step 5, fusion is performed through a filtering algorithm, specifically including:

Step 51, calculate the relationship between current and battery state of charge:

Among them, Q is the current maximum available capacity of the battery, T is the sampling interval, and I is the current;

Step 52, in conjunction with the SOC result estimated by the BP neural network and formula (16), obtain the recursive formula of the state space equation:

Wherein, SOC _k is the SOC estimated value of k moment, SOC _BP,k is the SOC estimated value of k moment obtained by BP neural network model, I _k is the electric current of k moment;

Step 53. Simplify and obtain the state equation and observation equation of the system, and use the Kalman filter algorithm to perform time update and measurement update to obtain the final estimated value of the battery state of charge:

Among them, x _k is the system state variable, that is, SOC, y _k is the output, that is, the SOC estimated by the BP neural network model, u _k is the input, A is the state matrix, B is the input matrix, C is the output matrix, ω _k is the process noise, and v _k is the observation noise.

The present invention also discloses a battery state of charge estimation system considering the state of health, including:

The data set building block is used to obtain battery charge and discharge data and corresponding health status;

Battery model building blocks for building battery models;

The open circuit voltage estimation module is used to estimate the open circuit voltage of the battery according to the constructed data set and battery model;

The neural network model building block is used to construct a neural network model that takes the open circuit voltage and the state of health of the charge and discharge data as input, and takes the state of charge of the battery as the output;

The battery state of charge estimation module is used to combine the current information in the data set and the estimated initial value of the battery state of charge output by the neural network model, and fuse them through a filtering algorithm to obtain the final estimated value of the battery state of charge.

Compared with the prior art, the present invention has the following beneficial effects:

The estimation method proposed by the present invention estimates the open circuit voltage of the battery through voltage, current data and battery model, and uses it as the input of the neural network to endow the neural network with domain knowledge. Compared with the traditional voltage and current input, this method have a certain degree of interpretability. The invention adopts a filter algorithm to fuse estimation results, which can avoid the open-loop divergence risk brought by the data-driven method, so that the method has higher robustness. The invention considers the health state of the battery, and uses the health state as the input of the neural network to realize accurate estimation of the battery SOC under different health states, which is of great significance for improving the performance of the battery management system.

Description of drawings

Fig. 1 is a flow chart of the method proposed by the present invention.

FIG. 2 is a first-order RC equivalent circuit model in an embodiment of the present invention.

Fig. 3 is a structural diagram of the BP neural network constructed by the embodiment of the present invention.

Fig. 4 is the estimation result of the state of charge of the battery when the SOH of the battery is 97.5% in the embodiment of the present invention.

Fig. 5 is the estimation result of the state of charge of the battery when the SOH of the battery is 73.2% in the embodiment of the present invention.

detailed description

In order to enable those skilled in the art to better understand the solutions of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only It is an embodiment of a part of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

It should be noted that the terms "first" and "second" in the description and claims of the present invention and the above drawings are used to distinguish similar objects, but not necessarily used to describe a specific sequence or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances such that the embodiments of the invention described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having", as well as any variations thereof, are intended to cover a non-exclusive inclusion, for example, a process, method, system, product or device comprising a sequence of steps or elements is not necessarily limited to the expressly listed Those steps or elements may instead include other steps or elements not explicitly listed or inherent to the process, method, product or apparatus.

The present invention is described in further detail below in conjunction with accompanying drawing:

Referring to Fig. 1, a method for estimating the state of charge of a battery considering the state of health provided by the present invention includes the following steps:

The first step is to use public data or charge and discharge equipment (battery detection system, charging pile, etc.) to obtain battery charge and discharge data and its corresponding state of health (SOH), and build a data set;

The second step is to establish the corresponding battery model. The specific establishment process includes the following steps:

Step 21, establish a state space equation based on the first-order RC equivalent circuit model of the battery:

Among them, E ₀ is the open circuit voltage, R ₁ is the ohmic internal resistance, R ₂ is the polarization internal resistance, C ₂ is the polarization capacitance, U ₂ is the polarization voltage, I is the current, and U ₀ is the battery terminal voltage;

Step 22. Calculate the transfer function of the state space equation, and add a zero-order retainer link to calculate the transfer function G(s):

Among them, T is the sampling interval, H(s) is the transfer function of the zero-order holder, s is a variable symbol; e is an irrational number;

Step 23, perform an inverse Laplace transform on the transfer function G(s), and perform discretization processing:

Find the difference equation at time k:

Step 24, combined with formula (1), calculate the open circuit voltage at time k:

in,

Step 25. Set V _k =E _0(k) -U _0(k) to establish a controlled autoregressive moving average model of the battery, namely the CARMA model:

Wherein, θ ₁ =a, θ ₂ =[R ₁ +R ₂ (1/a-1)], θ ₃ =aR ₀ , q ^-1 is a one-step translation operator, and ξ _k is system noise.

In the third step, based on the dataset and the battery model, the open circuit voltage (OCV) of the battery is estimated. Specifically, the recursive least squares method with forgetting factor is used to estimate the parameters of the battery CARMA model online, including the following steps:

Step 31. Since the open circuit voltage changes very little in a very short time, it is assumed that the open circuit voltage does not change in a sampling interval, that is, E _0(k) =E _0(k-1) , and according to the battery CARMA model, it is obtained:

U _0(k) ＝[U _0(k-1) I _k -I _k-1 1][θ _1k θ _2k θ _3k θ _4k ] ^T (44)

Among them, θ _4k =(1-a)E _0(k) , thus obtain the estimated value of the open circuit voltage E _0(k) :

Step 32. Use the recursive least squares method with forgetting factor to estimate the open circuit voltage online, and initialize the estimated value:

Among them, x is the parameter to be estimated, x ₀ is the initial value of the parameter to be estimated, P ₀ is the initial value of the covariance matrix of the error, and E is the identity matrix;

Step 33, perform an iterative solution, and estimate the open circuit voltage according to formula (9), the iterative process is as follows:

e _k =y _k -H _k x _k-1 (47)

x _k ＝x _k-1 +K _k e _k (49)

Among them, k refers to time k, y _k is the battery terminal voltage, e _k is the expected error of y _k , H _k = [U _0(k-1) I _k -I _k-1 1], x _k = [ θ _1k θ _2k θ _3k θ _4k ] ^T is the parameter array to be estimated, K _k is the gain coefficient, P _k is the covariance moment, and λ is the forgetting factor.

The fourth step is to build a neural network model with open circuit voltage and health status as input and SOC as output. The specific method of constructing the neural network model includes:

Construct a BP neural network model including an input layer, a hidden layer and an output layer. The input feature sequence is x=[OCVSOH] ^T , which is the open circuit voltage and state of health of the battery, and the output sequence is y=SOC, which is the charge of the battery state, the number of neurons in the hidden layer is set to 64, and the activation function is the ReLU function. The output of the entire network for any sample is:

Among them, β _i is the connection weight between the i-th neuron in the hidden layer and the output layer, g(x) is the activation function, a _i is the i-th row of the connection weight matrix between the input layer and the hidden layer, x _j is the input matrix of the jth sample, b _i is the ith row of the bias matrix connecting the input layer and the hidden layer, b _i ⁱ is the output layer bias value, and P is the number of samples.

The fifth step is to combine the current information in the data set with the estimated initial value of SOC output by the neural network model, and fuse it through a filtering algorithm to obtain the final and accurate estimated value of SOC.

Fusion is performed through filtering algorithms, including:

Step 51, calculate the relationship between current and battery state of charge:

Among them, Q is the current maximum available capacity of the battery, T is the sampling interval, and I is the current;

Step 52, in conjunction with the SOC result estimated by the BP neural network and formula (16), obtain the recursive formula of the state space equation:

Wherein, SOC _k is the SOC estimated value of k moment, SOC _BP,k is the SOC estimated value of k moment obtained by BP neural network model, I _k is the electric current of k moment;

Step 53. Simplify and obtain the state equation and observation equation of the system, and use the Kalman filter algorithm to perform time update and measurement update to obtain the final estimated value of the battery state of charge:

Among them, x _k is the system state variable, that is, SOC, y _k is the output, that is, the SOC estimated by the BP neural network model, u _k is the input, A is the state matrix, B is the input matrix, C is the output matrix, ω _k is the process noise, and v _k is the observation noise.

Specific application examples:

The test object in this embodiment is a polyfluorine multi-ternary soft-pack lithium battery (DFDPSP1265132-10Ah) with a nominal capacity of 10Ah, and it is tested for pulse characteristics and urban road conditions (UDDS) and new European conditions under different health states. Driving cycle (NEDC) discharge test, obtain the voltage, current, time and their corresponding SOH during the test, and build a data set. Among them, the discharge data of UDDS operating conditions under different aging degrees are used for modeling, and the discharge data of NEDC operating conditions under different aging degrees are used for verification.

Based on the first-order RC equivalent circuit model of the battery, _a differential equation mathematical model is established, see Figure ₂ , where E0 is the open circuit voltage, R1 is the ohmic internal resistance, R2 is the polarization internal resistance, _C2 is the _polarization capacitance, U ₂ is the polarization voltage, I is the current, and U ₀ is the battery terminal voltage. By performing discrete and differential processing on the mathematical model, a controlled autoregressive moving average model (CARMA model) of the battery is established, as shown in the following formula:

θ₂为[R₁+R₂(1/a-1)]；θ₃为aR₀；q^-1为一步平移算子；ξ_k为系统噪声。In the formula: θ ₁ is

θ ₂ is [R ₁ +R ₂ (1/a-1)]; θ ₃ is aR ₀ ; q ^-1 is a one-step translation operator; ξ _k is system noise.

In order to avoid the "data saturation" problem, the forgetting factor is introduced into the recursive least squares method, taking into account the tracking performance of the program and the error requirements after stabilization. The parameters of the battery CARMA model are estimated online by the recursive least squares method with forgetting factor (FFRLS method), so as to obtain the estimated value of the open circuit voltage. The forgetting factor in the recursive least squares method is set to 0.9, and the initial value P ₀ of the covariance matrix is the identity matrix.

The BP neural network model is used as the main body to realize the SOC estimation based on OCV and SOH, that is, the SOH of the battery and the OCV obtained in the third step are used as the input characteristics of the BP neural network, and the battery SOC is used as the output. The BP neural network model is divided into a three-layer structure of input layer, hidden layer and output layer, and its structure is shown in Figure 3. The number of neurons in the hidden layer of the BP neural network model is set to 64, and the activation function is the ReLU function.

The fourth step can preliminarily realize the SOC estimation, but the estimation accuracy is relatively low and fluctuates greatly. At the same time, considering the open-loop divergence risk of the data-driven method, the fifth step combines the current information with the Kalman filter algorithm to further improve the estimation accuracy. . The maximum absolute error (MaxAE) and the root mean square error (RMSE) are used as the quantitative indicators of the estimated error.

In this embodiment, accurate SOC estimation under different SOHs is realized, and compared with the double-adaptive attenuation Kalman filter (DAFEKF) based on the first-order RC model: the method proposed by the present invention can improve the battery SOH from 97.7% to 73.2% In the range between, the MaxAE is not more than 1.5%, and the RMSE is about 0.5%, while the DAFEKF method as a comparison, with the decrease of SOH, the SOC estimation error gradually increases, indicating the superiority of the method proposed by the present invention.

Figure 4 and Figure 5 show in detail the SOC estimation results and comparisons under two different SOHs. When the SOH is 97.5%, the error of the SOC estimated by the method proposed by the present invention is close to the SOC obtained by the DAFEKF method, and the error is no more than 2%, and the SOC value can be estimated more accurately. However, when the battery SOH reaches 73.2%, the estimation error of the DAFEKF method has exceeded 6%, while the SOC estimation method proposed by the present invention can still ensure that the maximum absolute error does not exceed 1.5%, and can overcome the initial value dependence in common filtering methods problem, there is no obvious convergence process at the initial stage of calculation. This proves that the estimation accuracy of the proposed SOC estimation method can be guaranteed under different battery health states.

The above content is only to illustrate the technical ideas of the present invention, and cannot limit the protection scope of the present invention. Any changes made on the basis of the technical solutions according to the technical ideas proposed in the present invention shall fall within the scope of the claims of the present invention. within the scope of protection.

Claims (9)

1. A battery state of charge estimation method considering the state of health, characterized in that, comprising the following steps: Step 1: Obtain battery charge and discharge data and corresponding health status, and build a data set; Step 2: Build the battery model; Step 3: Estimate the open circuit voltage of the battery based on the data set in step 1 and the battery model in step 2; Step 4: Construct a neural network model that takes the open circuit voltage and the state of health of charge and discharge data as input, and takes the state of charge of the battery as output; Step 5: Combining the current information in the data set in step 1 and the initial battery state of charge estimate output by the neural network model in step 3, fuse them through a filtering algorithm to obtain the final battery state of charge estimate. 2. The method for estimating the state of charge of a battery considering the state of health according to claim 1, wherein in step 1, the charging and discharging data of the battery and the corresponding state of health are acquired by using public data or through charging and discharging equipment. 3. The method for estimating the state of charge of a battery considering the state of health according to claim 2, wherein the charging and discharging device is a battery detection system or a charging pile. 4. The method for estimating the state of charge of a battery considering the state of health according to claim 1, wherein the battery is a lead-acid battery, a nickel-metal hydride battery, a nickel-chromium battery or a lithium-ion battery. 5. The method for estimating the state of charge of a battery considering the state of health according to claim 1, wherein, in step 2, building a battery model includes the following steps: Step 21, establish a state space equation based on the first-order RC equivalent circuit model of the battery:

Among them, E ₀ is the open circuit voltage, R ₁ is the ohmic internal resistance, R ₂ is the polarization internal resistance, C ₂ is the polarization capacitance, U ₂ is the polarization voltage, I is the current, and U ₀ is the battery terminal voltage; Step 22. Calculate the transfer function of the state space equation, and add a zero-order retainer link to calculate the transfer function G(s):

Among them, T is the sampling interval, and H(s) is the transfer function of the zero-order holder; Step 23, perform an inverse Laplace transform on the transfer function G(s), and perform discretization processing:

Find the difference equation at time k:

Step 24, combined with formula (1), calculate the open circuit voltage at time k:

in,

Step 25. Set V _k =E _0(k) -U _0(k) to establish a controlled autoregressive moving average model of the battery, namely the CARMA model:

Wherein, θ ₁ =a, θ ₂ =[R ₁ +R ₂ (1/a-1)], θ ₃ =aR ₀ , q ^-1 is a one-step translation operator, and ξ _k is system noise. 6. The method for estimating the state of charge of a battery considering the state of health according to claim 5, wherein in step 3, the online estimation of battery CARMA model parameters is carried out using the recursive least squares method with a forgetting factor, comprising the following steps : Step 31. Since the open circuit voltage changes very little in a very short time, it is assumed that the open circuit voltage does not change in a sampling interval, that is, E _0(k) =E _0(k-1) , and according to the battery CARMA model, it is obtained: U _0(k) ＝[U _0(k-1) I _k -I _k-1 1][θ _1k θ _2k θ _3k θ _4k ] ^T (8) Among them, θ _4k =(1-a)E _0(k) , thus obtain the estimated value of the open circuit voltage E _0(k) :

Step 32. Use the recursive least squares method with forgetting factor to estimate the open circuit voltage online, and initialize the estimated value:

Among them, x is the parameter to be estimated, x ₀ is the initial value of the parameter to be estimated, P ₀ is the initial value of the covariance matrix of the error, and E is the identity matrix; Step 33, perform an iterative solution, and estimate the open circuit voltage according to formula (9), the iterative process is as follows: e _k =y _k -H _k x _k-1 (11)

x _k ＝x _k-1 +K _k e _k (13)

Among them, k refers to time k, y _k is the battery terminal voltage, e _k is the expected error of y _k , H _k = [U _0(k-1) I _k -I _k-1 1], x _k = [ θ _1k θ _2k θ _3k θ _4k ] ^T is the parameter array to be estimated, K _k is the gain coefficient, P _k is the covariance moment, and λ is the forgetting factor. 7. The battery state of charge estimation method considering the state of health according to claim 1, characterized in that, in step 4, the specific method of constructing a neural network model includes: Construct a BP neural network model including an input layer, a hidden layer and an output layer. The input feature sequence is x=[OCVSOH] ^T , which is the open circuit voltage and state of health of the battery, and the output sequence is y=SOC, which is the charge of the battery state, the number of neurons in the hidden layer is set to 64, and the activation function is the ReLU function. The output of the entire network for any sample is:

Among them, β _i is the connection weight between the i-th neuron in the hidden layer and the output layer, g(x) is the activation function, a _i is the i-th row of the connection weight matrix between the input layer and the hidden layer, x _j is the input matrix of the jth sample, b _i is the ith row of the bias matrix connecting the input layer and the hidden layer, b _i ⁱ is the output layer bias value, and P is the number of samples. 8. The method for estimating the state of charge of a battery considering the state of health according to claim 1, wherein in step 5, fusion is performed through a filtering algorithm, specifically comprising: Step 51, calculate the relationship between current and battery state of charge:

Among them, Q is the current maximum available capacity of the battery, T is the sampling interval, and I is the current; Step 52, in conjunction with the SOC result estimated by the BP neural network and formula (16), obtain the recursive formula of the state space equation:

Wherein, SOC _k is the SOC estimated value of k moment, SOC _BP,k is the SOC estimated value of k moment obtained by BP neural network model, I _k is the electric current of k moment; Step 53. Simplify and obtain the state equation and observation equation of the system, and use the Kalman filter algorithm to perform time update and measurement update to obtain the final estimated value of the battery state of charge:

Among them, x _k is the system state variable, that is, SOC, y _k is the output, that is, the SOC estimated by the BP neural network model, u _k is the input, A is the state matrix, B is the input matrix, C is the output matrix, ω _k is the process noise, and v _k is the observation noise. 9. A battery state of charge estimation system considering the state of health, characterized in that it comprises: The data set building block is used to obtain battery charge and discharge data and corresponding health status; Battery model building blocks for building battery models; The open circuit voltage estimation module is used to estimate the open circuit voltage of the battery according to the constructed data set and battery model; The neural network model building block is used to construct a neural network model that takes the open circuit voltage and the state of health of the charge and discharge data as input, and takes the state of charge of the battery as the output; The battery state of charge estimation module is used to combine the current information in the data set and the estimated initial value of the battery state of charge output by the neural network model, and perform fusion through a filtering algorithm to obtain the final estimated value of the battery state of charge.

Images