Power battery adaptive linear modeling and state-of-charge estimation method considering internal and external information of battery

Abstract

The invention relates to a power battery adaptive linear modeling and state of charge estimation method considering internal and external information of a battery, and belongs to the field of automobile power lithium battery modeling and SOC estimation. The method comprises the following steps: collecting dynamic data of a power battery, preprocessing the data, and preliminarily calculating the SOC of the power battery; establishing a linear relation between the open-circuit voltage Voc and the SOC by adopting a piecewise linear processing mode; establishing a power battery linear model considering internal and external information of the battery and a discrete state space equation thereof; a self-adaptive forgetting factor least square algorithm is adopted to identify battery parameters on line; and estimating the SOC according to the online identified battery parameters through a noise adaptive Kalman algorithm. The method can solve the problems that an existing lithium battery electrochemical model cannot consider the influence of internal and external factors of the battery on a battery model, the calculation burden of a nonlinear battery model is high, the information redundancy of an estimation algorithm is high, and the anti-noise capability is weak.

Description

Technical Field

[0001] The present invention belongs to the field of automotive power lithium battery modeling and SOC estimation, and relates to a power battery adaptive linear modeling and charge state estimation method taking into account internal and external information of the battery. Background Art

[0002] Lithium-ion batteries have become the core power source of the electric vehicle industry due to their significant advantages of low self-discharge rate, long cycle life, high energy density and continuously decreasing costs. With the vigorous development of the electric vehicle market and technological iteration, the widespread use of lithium-ion batteries is also accompanied by a series of safety challenges, especially the frequent vehicle accidents caused by internal failures, which has attracted great attention from the industry. In this context, the importance of automotive battery management systems (BMS) has become increasingly prominent, among which the accurate estimation of the battery state of charge (SOC) has become the core task of BMS management. A reliable and accurate SOC evaluation system can not only provide key information on the remaining available energy of the battery in real time to ensure the safety of the battery during driving, but also effectively extend the overall service life of the battery and improve the operating efficiency of the battery system by optimizing the charging and discharging strategy, which is of great significance to promoting the sustainable development of the electric vehicle industry.

[0003] Existing estimation methods for lithium battery SoC can be generally divided into model-based methods and data-driven methods. For data-driven (deep learning) methods, the calculation time is long, a huge amount of data is required to train the model, and a lot of manpower and material resources are consumed; for model-based methods, there are the following shortcomings: First, the traditional lithium battery electrochemical model is difficult to consider the impact of internal and external factors of the battery; second, the generalization ability of the traditional lithium battery electrochemical model is low; third, the traditional lithium battery electrochemical model is a nonlinear model with a high computational burden; fourth, the traditional estimation algorithm has high information redundancy and weak noise resistance. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a method for adaptive linear modeling and state of charge estimation of a power battery taking into account internal and external information of the battery, so as to overcome the problems that the existing lithium battery electrochemical model cannot consider the influence of internal and external factors of the battery on the battery model, the nonlinear battery model has a high computational burden, the estimation algorithm has high information redundancy, and the anti-noise ability is weak.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery, the method comprising the following steps:

[0007] S1. Collect dynamic data of the power battery, pre-process the data, and preliminarily calculate the SOC of the power battery;

[0008] S2, using piecewise linear processing to establish an open circuit voltage V oc Linear relationship with SOC;

[0009] S3. Establish a linear model of the power battery and its discrete state space equation considering the internal and external information of the battery;

[0010] S4, using adaptive forgetting factor least squares algorithm to identify battery parameters online;

[0011] S5. Estimate the SOC based on the online identified battery parameters using a noise adaptive Kalman algorithm.

[0012] Furthermore, in S1, the real-time data of the power battery obtained includes at least: current, open circuit voltage, and temperature; wherein the SOC is preliminarily calculated using the ampere-hour integration method:

[0013]

[0014] Where soc(t) is the SOC at time t; soc(0) is the initial SOC; C is the rated capacity of the battery in ampere-hours; and I(t) is the current at time t.

[0015] Furthermore, in S2, a piecewise linear regression method is used to deal with high-order nonlinear V oc -SOC relationship, and obtain linear V oc -SOC relationship, the formula is:

[0016] V oc (k)=f(soc)=b1(k)*soc(k)+b0(k)

[0017] Among them, soc(k) represents the SOC value at time step k, b1(k) and b0(k) represent the V corresponding to soc(k) respectively. oc -Slope and intercept values ​​of the SOC linear segment curve.

[0018] Further, in S3, considering the influence of internal and external factors of the power battery, a linear equivalent model of the power battery is established, which includes the following steps:

[0019] S31, design temperature weighting factor T weigh , aging weighting factor A weight And SOC weighting factor SOC weight , where the temperature weighting factor T weight It is expressed as:

[0020]

[0021] Among them, α Tis the temperature influence coefficient, T ref is the reference temperature;

[0022] Aging weighting factor A weight It is expressed as:

[0023]

[0024] Among them, α A is the aging influence coefficient, A is the aging degree;

[0025] SOC weighting factor SOC weig It is expressed as:

[0026]

[0027] Among them, α c is the SOC influence coefficient, SOC(k) is the SOC value at K time steps;

[0028] S32. Use the Sigmoid activation function to limit the value of each weighting factor to a range of 0-1:

[0029]

[0030] V t =V oc (k)-Vd k -R0(p k ).I k =b1(k)*soc(k)+b0(k)-Vd k -R0(p k ).I k

[0031] Among them, p k It is defined as the total impact factor, which represents the comprehensive impact of internal and external factors on model parameters; Soc k and Soc k+1 Respectively represent the SOC at the current moment and the next moment; Vd k and Vd k+1 Represents the RC dynamic voltage at the current moment and the next moment; R0(p k ) is considered p k Influence of battery internal resistance, R d (p k ) is considered p k The influence of polarization internal resistance, C d (p k ) is considered p k Affected polarization capacitance; I k 、V t Respectively represent the input current and output voltage at k time step; Ts represents the sampling time, η represents the battery charge and discharge rate, C n represents the nominal capacity of the battery, b1(k) and b0(k) represent soc(k) corresponding to V oc -Slope and intercept values ​​of the SOC linear segment curve.

[0032] Further, in S4, the process of online identifying battery parameters using the adaptive forgetting factor least squares algorithm includes the following steps:

[0033] S41, combined with the adaptive influence gain matrix M k Discretize the system output equation into the least squares form;

[0034] S42, using the recursive forgetting factor least squares algorithm to identify the parameter matrix θ online k ;

[0035] S43, design an adaptive strategy to adaptively update the forgetting factor λ;

[0036] S44. Calculate the battery internal resistance and polarization network parameters affected by internal and external factors based on the identified parameters.

[0037] Further, in S41, let y=V oc -V t , then the system output equation is expressed as:

[0038] V oc (k)-V t (k) = b1(V oc (k-1)-V t (k-1))+a1I k +a2I k-1

[0039] The discretized least squares form is:

[0040]

[0041] In the formula, and θ k =[b1, a1, a2] are the data matrix and the parameter vector matrix to be identified respectively;

[0042] Adaptive influence gain matrix M k It is expressed as:

[0043]

[0044] Among them, m1, m2, and m3 are the combination of SOC weighting factor, temperature weighting factor, and aging weighting factor;

[0045] In S42, there are:

[0046]

[0047] Where P k-1 and P k Respectively represent the covariance matrix at time k-1 and time k; represents the input matrix; K LS，k Represents the gain value of the RLS algorithm at time k;

[0048] In S43, there are:

[0049]

[0050] Among them, λ(k) is the forgetting factor, λ min Represents the minimum value of the set forgetting factor, λ max Indicates the maximum value of the set forgetting factor, e base represents the expected error value, e(k) represents the estimated error at K time steps, ε(k) represents the output voltage estimation error at time k; h is a constant value, which represents the sensitivity of the algorithm to the error;

[0051] In S44, there are:

[0052]

[0053] Among them, b1, a1, a2 are the parameters identified by the RLS algorithm and are used to calculate the battery parameter R0 (p k ), R d (p k ), C d (p k ), R0(p k ) is the internal resistance of the battery affected by internal and external factors, R d (p k ) is the polarization internal resistance affected by internal and external factors, C d (p k ) is the polarized capacitance affected by internal and external factors.

[0054] Further, in S5, the identified battery parameters are transmitted to the noise adaptive Kalman filter algorithm for SOC estimation, which includes the following process:

[0055] S51, obtaining a discretized state space equation of a battery model for SOC estimation while taking noise into consideration;

[0056] S52, using the recursive process of the standard Kalman filter algorithm to obtain the optimal estimated value of the SOC at each moment;

[0057] S53: Use a residual-based noise adaptive strategy to adjust the noise value.

[0058] Further, in S51, the discretized state space equation is expressed as:

[0059]

[0060] In the formula, w k 、V k are system state noise and system measurement noise respectively, A k , B k , C k Respectively represent A k , B k , C k Respectively represent the state transfer matrix, input matrix, and observation matrix, and have:

[0061]

[0062] C k =[-1b1(k)]

[0063] The R0(p k ), R d (p k ), C d (p k ) Parameters are updated at all times A k , B k matrix.

[0064] Further, in S52, the process of obtaining the optimal estimated value of SOC at each moment by using the standard Kalman filter algorithm includes:

[0065] Status prediction: x k - =A k-1 x k-1 +B k-1 μ k-1 ;

[0066] State prediction covariance: P k - =A k-1 P k-1 A k-1 T +Q k-1 ;

[0067] Calculate the Kalman gain: K k =P k - C k T (C k P k - C kT +R k ) -1 ;

[0068] Status update: x k + =x k - +K k (y k -C k x k - );

[0069] Covariance update: P k + =(IK k C k ) k - .

[0070] Further, in S53, the noise adaptation strategy based on the residual is expressed as:

[0071] ero k =V k -V e

[0072]

[0073] Among them, V k Represents the true value of the terminal voltage at time step k, V e Represents the terminal voltage estimate output by the Kalman filter at time step k, ero k represents the error value between the Kalman estimated voltage and the actual terminal voltage at time k; H e The residual sum of squares is calculated using the sliding window technique, where M1 represents the size of the sliding window value and H e represents the mean of the residual sum of squares, K k represents the Kalman gain, R k is the adaptively updated input noise value, Q k is the output noise value updated adaptively.

[0074] The beneficial effects of the present invention are:

[0075] The present invention establishes a power lithium-ion battery model that takes into account internal and external information of the battery, invents weighting factors of internal and external factors such as SOC and temperature, and invents an adaptive gain weighting matrix to take into account the dynamic influence of internal and external factors such as SOC and temperature on the battery model, thereby improving the model accuracy and estimation precision; in addition, the present invention establishes a linear power battery model, reduces the amount of calculation of the automobile BMS, and improves real-time applicability; and adopts an adaptive forgetting factor least squares algorithm for parameter identification, dynamically adjusts the size of the forgetting factor in real time, solves information redundancy, and improves algorithm stability; and adopts a noise adaptive Kalman filter algorithm for SOC estimation, invents a noise adaptive strategy to adjust noise parameters, and solves the problems of low accuracy and weak anti-noise performance of the lithium battery model caused by environmental noise changes.

[0076] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0078] Figure 1 It is a simplified flow chart of the method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to the present invention;

[0079] Figure 2 is the linear V obtained by piecewise linear method oc -SOC relationship diagram;

[0080] Figure 3 This is a power battery model diagram that takes into account internal and external factors such as SOC and temperature;

[0081] Figure 4 It is a detailed flow chart of the method for adaptive linear modeling and state of charge estimation of a power battery taking into account internal and external information of the battery according to the present invention. DETAILED DESCRIPTION

[0082] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0083] Among them, the drawings are only used for illustrative explanations, and they only represent schematic diagrams rather than actual pictures, and should not be understood as limitations on the present invention. In order to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0084] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right", "front", "rear", etc. indicate the orientation or position relationship, they are based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the terms describing the position relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0085] See also Figure 1 to Figure 4 , which is an adaptive linear modeling and state of charge estimation method for power batteries that takes into account internal and external information of the battery.

[0086] Example

[0087] This embodiment proposes a power battery adaptive linear modeling and charge state estimation method considering the internal and external information of the battery. Figure 1 As shown, it includes the following steps:

[0088] S1: Collect dynamic data of the power lithium-ion battery, pre-process the input current data and voltage data, obtain pre-processed battery data, and preliminarily calculate the SOC of the power battery;

[0089] S2: Piecewise linear processing of open circuit voltage V oc -SOC data, get linear V oc -SOC relationship;

[0090] S3: Establish a linear model of the power battery taking into account the internal and external information of the battery, and obtain the discrete state space equation of the model;

[0091] S4: Online identification of battery parameters using adaptive forgetting factor least squares algorithm,

[0092] S5: The identified battery parameters are passed to the noise adaptive Kalman algorithm to estimate SOC online.

[0093] In S1 of this embodiment, battery data and information are acquired in real time, and data processing includes: collecting data of actual lithium-ion batteries to obtain information such as current, open circuit voltage, temperature, etc., and preliminarily calculating SOC using the ampere-hour integration method:

[0094]

[0095] Where soc(t) is the SOC at time t, soc(0) is the initial SOC (usually ranging from 0 to 1 or 0% to 100%), C is the rated capacity of the battery (in ampere-hours, Ah), and I(t) is the current at time t (positive during charging and negative during discharging).

[0096] In S2 of this embodiment, as Figure 2 As shown, the piecewise linear regression method is used to deal with high-order nonlinear V oc -SOC relationship, and obtain linear V oc -SOC relationship, the formula is:

[0097] V oc (k)=f(soc)=b1(k)*soc(k)+b0(k)

[0098] Among them, soc(k) represents the SOC value at time step k, b1(k) and b0(k) represent the SOC value corresponding to v at time step k, respectively. oc -Slope and intercept values ​​of the SOC linear segment curve.

[0099] In S3 of this embodiment, the influence of internal and external factors such as SOC and temperature on the power battery is considered, and combined with the LPV theory, a linear equivalent model of the power battery is established that considers the influence of factors such as temperature change and SOC on battery parameters, such as Figure 3 As shown, the specific steps include:

[0100] S31, using p k The Sigmoid activation function quantifies the impact of internal and external information on the power battery and designs the temperature weighting factor:

[0101]

[0102] Among them, α Tis the temperature influence coefficient, T ref is the reference temperature.

[0103] Design aging weighting factor:

[0104]

[0105] Among them, α A is the aging influence coefficient. A is the degree of aging, and its value is between (0,1). For new batteries, the number of cycles is very small: A≈0; moderate aging: A≈0.5-0.8; severe aging: A>0.8.

[0106] Design SOC weighting factor:

[0107]

[0108] Among them, α c is the SOC influence coefficient, and SOC(k) is the SOC value at K time steps.

[0109] S32. The value of each weighting factor is limited to the range of 0-1 through the Sigmoid activation function:

[0110]

[0111] V t =V oc (k)-Vd k -R0(p k ).I k =b1(k)*soc(k)+b0(k)-Vd k -R0(p k ).I k

[0112] Among them, p k Defined as the total influencing factor, it represents the comprehensive influence of internal factors such as soc, aging degree, and external factors such as working environment temperature and noise on the model parameters. k+1 Indicates the SOC at the next moment, Vd k+1 Indicates the RC dynamic voltage at the next moment, Soc k Indicates the current SOC, Vd k Represents the RC dynamic voltage at the current moment, R0(p k ) is considered p k Influence of battery internal resistance, R d (p k ) is considered p k The influence of polarization internal resistance, C d (p k ) is considered p k The polarization capacitance is affected.k 、V t Respectively represent the input current and output voltage at k time step. s represents the sampling time, η represents the battery charge and discharge rate, C n represents the nominal capacity of the battery, b1(k) and b0(k) represent soc(k) corresponding to V oc -Slope and intercept values ​​of the SOC linear segment curve.

[0113] In S4 of this embodiment, the battery parameters are identified online using an adaptive forgetting factor least squares algorithm (AFFRLS), which specifically includes the following steps:

[0114] S41, using adaptive influence gain matrix M k Quantify the impact of internal and external factors such as SOC and temperature on the power battery model. The impact of SOC, temperature, aging and other factors on model parameters varies, and actual applications should be considered according to needs;

[0115] Combined with M k , let y = V oc -V t , discretize the system output equation into the least square form, where the system output equation is expressed as:

[0116] V oc (k)-V t (k) = b1(V oc (k-1)-V t (k-1))+a1I k +a2I k-1

[0117] The discretized least squares form is:

[0118]

[0119] In the formula, and θ k =[b1, a1, a2] are the data matrix and the parameter vector matrix to be identified respectively.

[0120] Adaptive influence gain matrix M k It is expressed as:

[0121]

[0122] Among them, m1, m2, and m3 are the combination of SOC, temperature, and aging weighting factors. When considering the impact of SOC on the battery, SOC mainly affects the dynamic response rate of current to voltage, m1 = 1,

[0123] S42, using the recursive forgetting factor least squares algorithm (FFRLS) to identify the parameter matrix θ online k ;

[0124]

[0125] Where P k-1 and P k Respectively represent the covariance matrix at time k-1 and time k; represents the input matrix; K LS,k Represents the gain value of the RLS algorithm at time k;

[0126] S43, design an adaptive strategy to adaptively update the forgetting factor λ;

[0127]

[0128] Among them, λ min Represents the minimum value of the set forgetting factor, λ max Indicates the maximum value of the set forgetting factor. Usually, λ max Take 0.999. base represents the expected error value, and e(k) represents the estimated error at K time steps. h is a constant value, which represents the sensitivity of the algorithm to the error. When h tends to 1, the change of λ tends to slow down, resulting in a slower response. On the contrary, when h is close to 0, λ oscillates violently, the response time is fast, but the accuracy is low.

[0129] S44, calculating the battery internal resistance and polarization network parameters affected by internal and external factors such as SOC and temperature based on the identified parameters;

[0130]

[0131] In S5 of this embodiment, the battery parameters identified by the adaptive forgetting factor least squares algorithm (AFFRLS) are transferred to the noise adaptive Kalman filter algorithm for SOC estimation, which specifically includes the following steps:

[0132] S51. Considering the noise, the discretized state space equation of the battery model for SOC estimation is obtained:

[0133]

[0134] Among them, w k 、V k are system state noise and system measurement noise respectively, A k , B k , C k They represent the state transfer matrix, input matrix, and observation matrix respectively.

[0135]

[0136] C k =[-1b1(k)]

[0137] Using the identified R0(p k ), R d (p k ), C d (p k ) Update A at all times k , B k matrix.

[0138] S52. Using the recursive process of the standard Kalman filter algorithm, the optimal estimated value of the SOC at each moment can be obtained, including:

[0139] Status prediction: x k - =A k-1 x k-1 +B k-1 μ k-1

[0140] State prediction covariance: P k - =A k-1 P k-1 A k-1 T +Q k-1

[0141] Calculate the Kalman gain: K k =P k - C k T (C k P k - C k T +R k ) -1

[0142] Status update: x k + =x k - +K k (y k -C k x k - ).

[0143] S53. In the actual operation state of the vehicle, the noise at the input and output ends of the power lithium-ion battery is not Gaussian white noise in the ideal state. Therefore, in order to improve the estimation accuracy of the algorithm in the actual noise change environment and enhance the robustness of the algorithm, a noise adaptive strategy based on residual is invented;

[0144] ero k =V k -V e

[0145]

[0146] Among them, V k Represents the true value of the terminal voltage at time step k, V e Represents the terminal voltage estimate output by the Kalman filter at time step k, ero k It represents the error value between the Kalman estimated voltage and the actual terminal voltage at time k. e The residual sum of squares is calculated using the sliding window technique, where M1 represents the size of the sliding window value and H e represents the mean of the residual sum of squares, K k represents the Kalman gain, R k is the adaptively updated input noise value, Q k is the output noise value updated adaptively. Figure 4 As shown, by combining a low computational burden, highly accurate model with a highly robust, accurate algorithm, the real-time SOC estimation value can be output quickly, accurately, and stably.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery, characterized in that: The method comprises the following steps: S1. Collect dynamic data of the power battery, pre-process the data, and preliminarily calculate the SOC of the power battery; S2, using piecewise linear processing to establish an open circuit voltage V oc Linear relationship with SOC; S3. Establish a linear model of the power battery and its discrete state space equation considering the internal and external information of the battery; S4, using adaptive forgetting factor least squares algorithm to identify battery parameters online; S5. Estimate the SOC based on the online identified battery parameters using a noise adaptive Kalman algorithm.

2. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 1, characterized in that: In S1, the real-time data of the power battery obtained includes at least: current, open circuit voltage, and temperature; among them, the SOC is preliminarily calculated using the ampere-hour integration method: Where soc(t) is the SOC at time t; soc(0) is the initial SOC; C is the rated capacity of the battery in ampere-hours; and I(t) is the current at time t.

3. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 2, characterized in that: In S2, a piecewise linear regression method is used to deal with high-order nonlinear V oc -SOC relationship, and obtain linear V oc -SOC relationship, the formula is: In oc (k)=f(soc)=b1(k)*soc(k)+b0(k) Among them, soc(k) represents the SOC value at time step k, b1(k) and b0(k) represent the V corresponding to soc(k) respectively. oc -Slope and intercept values ​​of the SOC linear segment curve.

4. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 3, characterized in that: In S3, considering the influence of internal and external factors of the power battery, a linear equivalent model of the power battery is established, which includes the following steps: S31, design temperature weighting factor T weig , aging weighting factor A weight And the SOC weighting factor SOC weight , where the temperature weighting factor T weigh It is expressed as: Among them, α T is the temperature influence coefficient, T ref is the reference temperature; Aging weighting factor A weight It is expressed as: Among them, α A is the aging influence coefficient, A is the aging degree; SOC weighting factor SOC weight It is expressed as: Among them, α c is the SOC influence coefficient, SOC(k) is the SOC value at K time steps; S32, through Sigmoid(1 / (1+e x ))The activation function limits the value of each weighting factor to a range of 0-1: V t =V oc (k)-Vd k -R0(p k ).I k =b1(k)*soc(k)+b0(k)-Vd k -R0(p k ).I k Among them, p k It is defined as the total impact factor, which represents the comprehensive impact of internal and external factors on model parameters; Soc k and Soc k+1 Respectively represent the SOC at the current moment and the next moment; Vd k and Vd k+1 Represents the RC dynamic voltage at the current moment and the next moment; R0(p k ) is considered p k Influence of battery internal resistance, R d (p k ) is considered p k The influence of polarization internal resistance, c d (p k ) is considered p k Affected polarization capacitance; I k 、V t Respectively represent the input current and output voltage at k time step; T s represents the sampling time, η represents the battery charge and discharge rate, C n represents the nominal capacity of the battery, b1(k) and b0(k) represent soc(k) corresponding to V oc -Slope and intercept values ​​of the SOC linear segment curve.

5. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 4, characterized in that: In S4, the process of online identification of battery parameters using the adaptive forgetting factor least squares algorithm includes the following steps: S41, combined with the adaptive influence gain matrix M k Discretize the system output equation into the least squares form; S42, using the recursive forgetting factor least squares algorithm to identify the parameter matrix θ online k ; S43, design an adaptive strategy to adaptively update the forgetting factor λ; S44. Calculate the battery internal resistance and polarization network parameters affected by internal and external factors based on the identified parameters.

6. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 5, characterized in that: In S41, let y = V oc -V t , then the system output equation is expressed as: In oc (k)-V t (k)=b1(V oc (k-1)-V t (k-1))+a1I k +a2I k-1 The discretized least squares form is: In the formula, and θ k =[b1, a1, a2] are the data matrix and the parameter vector matrix to be identified respectively; Adaptive influence gain matrix M k It is expressed as: Among them, m1, m2, and m3 are the combination of SOC weighting factor, temperature weighting factor, and aging weighting factor; In S42, there are: Where P k-1 and P k Respectively represent the covariance matrix at time k-1 and time k; represents the input matrix; K LS,k Represents the gain value of the RLS algorithm at time k; In S43, there are: Among them, λ(k) is the forgetting factor, λ min Represents the minimum value of the set forgetting factor, λ max Indicates the maximum value of the set forgetting factor, e base represents the expected error value, e(k) represents the estimated error at K time steps, ε(k) represents the output voltage estimation error at time k; h is a constant value, which represents the sensitivity of the algorithm to the error; In S44, there are: Among them, b1, a1, a2 are the parameters identified by the RLS algorithm and are used to calculate the battery parameter R0 (p k ), R d (p k ), C d (p k ), R0(p k ) is the internal resistance of the battery affected by internal and external factors, R d (p k ) is the polarization internal resistance affected by internal and external factors, C d (p k ) is the polarized capacitance affected by internal and external factors.

7. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 6, characterized in that: In S5, the identified battery parameters are passed to the noise adaptive Kalman filter algorithm for SOC estimation. The process includes: S51, obtaining a discretized state space equation of a battery model for SOC estimation while taking noise into consideration; S52, using the recursive process of the standard Kalman filter algorithm to obtain the optimal estimated value of the SOC at each moment; S53: Use a residual-based noise adaptive strategy to adjust the noise value.

8. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 7, characterized in that: In S51, the discretized state space equation is expressed as: In the formula, w k 、V k are system state noise and system measurement noise respectively, A k , B k , C k Respectively represent the state transfer matrix, input matrix, and observation matrix, and have: FC k =[-1 b1(k)] The R0(p k ), R d (p k ), C d (p k ) Parameters are updated at all times A k , B k matrix.

9. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 8, characterized in that: In S52, the process of obtaining the optimal estimated value of SOC at each moment by using the standard Kalman filter algorithm includes: Status prediction: x k - =A k-1 x k-1 +B k-1 μ k-1 ; State prediction covariance: P k - =A k-1 P k-1 A k-1 T +Q k-1 ; Calculate the Kalman gain: K k =P k - C k T (C k P k - C k T +R k ) -1 ; Status update: x k + =x k - +K k (y k -C k x k - ); Covariance update: P k + =(IK k C k ) k - .

10. The method for adaptive linear modeling and state of charge estimation of a power battery considering internal and external information of the battery according to claim 9, characterized in that: In S53, the residual-based noise adaptation strategy is expressed as: hero k =V k -V e Among them, V k Represents the true value of the terminal voltage at time step k, V e Represents the terminal voltage estimate output by the Kalman filter at time step k, ero k represents the error value between the Kalman estimated voltage and the actual terminal voltage at time k; H e The residual sum of squares is calculated using the sliding window technique, where M1 represents the size of the sliding window value and H e represents the mean of the residual sum of squares, K k represents the Kalman gain, R k is the adaptively updated input noise value, Q k is the output noise value updated adaptively.

Images