Lithium-ion battery state joint estimation method and system

Through the combined state estimation method of lithium-ion battery based on fractional coupling model, the problem of unstable SOC estimation of lithium-ion batteries is solved, and higher accuracy state estimation and battery safety improvement are achieved.

CN117074962BActive Publication Date: 2024-11-01JIANGNAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311060583.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-22
Publication Date
2024-11-01
Estimated Expiration
2043-08-22

AI Technical Summary

Technical Problem

The internal structure of lithium-ion batteries is complex and greatly affected by external factors, resulting in unstable SOC estimates, affecting the battery life and safety performance.

Method used

Using the combined state estimation method of lithium-ion battery based on fractional-order coupling model, the second-order equivalent circuit and two-state thermal model are constructed, combined with dynamic interest least squares method and MMFCKF algorithm, parameter estimation and state joint estimation are performed to accurately describe the battery. nonlinear changes.

Benefits of technology

It improves the accuracy of lithium-ion battery state estimation, reduces battery model errors, enhances the safety performance and stability of the battery, and avoids shortening of service life.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117074962B_ABST
    Figure CN117074962B_ABST
Patent Text Reader

Abstract

The present application provides a method and system for jointly estimating the state of a lithium-ion battery, the method comprising: constructing an electrical model of the lithium-ion battery based on a second-order equivalent circuit according to the input charging current; constructing a two-state thermal model of the lithium-ion battery according to the impedance parameters of the electrical model and the heat generation and convection heat transfer relationship of the lithium-ion battery itself during the charging process, and constructing a fractional-order coupling model of the lithium-ion battery based on the bidirectional coupling effect of the electrical model and the two-state thermal model; based on the fractional-order coupling model, using the dynamic discounted least squares method to estimate the parameters of the fractional-order coupling model; based on the fractional-order coupling model, using the MMFCKF algorithm to jointly estimate the SOC and SOH states of the fractional-order coupling model. The accuracy of the battery model is improved by establishing a fractional-order electrothermal coupling model method, which has good robustness. Its parameter identification method can identify the model parameters with high accuracy, and the lithium battery state joint estimation method can achieve joint estimation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of battery technology, and in particular to a method and system for jointly estimating the state of a lithium-ion battery. Background Art

[0002] In recent years, environmental pollution and energy security have prompted the world to promote the development and utilization of new energy. Promoting the development of new energy vehicles such as hybrid and pure electric vehicles is an effective means to achieve this goal. Lithium-ion batteries are one of the core components of the power system of new energy vehicles. Energy storage and release are mainly completed through chemical reactions. Their safety and power performance are affected by many factors such as charge state and health state. Therefore, SOC and SOH estimation have become a hot topic of widespread concern and research.

[0003] The working process of lithium-ion batteries is very diverse. It is a complex, nonlinear, time-varying dynamic system. Its dynamic characteristics are also affected by many factors, including: battery electrode materials, internal electrolytes of the battery, and ambient temperature. Therefore, for a battery with complex dynamic characteristics, accurate modeling is a very difficult problem. At present, there are three types of battery models commonly used by domestic and foreign scholars to study batteries: electrochemical model, equivalent circuit model, and black box model.

[0004] According to the charging and discharging current data of lithium-ion batteries, different parameter identification methods can be used to identify the parameters of the established model. Through parameter identification, the real-time voltage, current, internal resistance and other data of lithium-ion batteries can be obtained, so as to better grasp the operating rules of lithium-ion batteries and increase the service life of batteries. According to the difference between the model parameter identification results and the experimental data, parameter identification can be divided into two categories: offline identification and online identification. Using a parameter identification method suitable for the battery model to accurately identify the parameters in the model can effectively improve the accuracy of the model.

[0005] The state of charge of lithium-ion batteries is a key performance indicator of battery management systems, and its accurate estimation is the basis for calculating and obtaining other state variables of lithium-ion battery systems. High-precision SOC and SOH values ​​can accurately describe the internal characteristics of lithium-ion batteries, and play a key role in the efficient management of power and energy of battery systems. Accurate SOC estimation of lithium-ion batteries can avoid the shortening of lithium-ion battery service life and enhance the safety and stability of lithium-ion batteries. However, due to the complex internal structure of lithium-ion batteries, variable operating temperatures, and susceptibility to external factors, the estimated SOC value of the battery varies.

[0006] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0007] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present application provides a lithium-ion battery state joint estimation method and system based on a fractional-order coupling model, which can accurately describe the nonlinear changes of the lithium-ion battery system during use, reduce the error of the battery model, and improve the accuracy of lithium battery state estimation.

[0008] The present invention provides a method for jointly estimating the state of a lithium-ion battery based on a fractional-order coupling model, the method comprising:

[0009] According to the input charging current, an electrical model of the lithium-ion battery is constructed based on a second-order equivalent circuit;

[0010] A two-state thermal model of the lithium-ion battery is constructed according to the impedance parameters of the electrical model and the heat generation and convection heat transfer relationship of the lithium-ion battery itself during charging, and a fractional-order coupling model of the lithium-ion battery is constructed based on the bidirectional coupling effect between the electrical model and the two-state thermal model;

[0011] Based on the fractional-order coupling model, a dynamic discounted least squares method is used to estimate parameters of the fractional-order coupling model;

[0012] Based on the fractional-order coupling model, the MMFCKF algorithm is used to jointly estimate the SOC and SOH states of the fractional-order coupling model.

[0013] In one embodiment of the present invention, the state space equation of the electrical model is:

[0014]

[0015] Where k is a discrete time point, x k is the state vector at discrete time point k, A k-1 is the matrix A at time k-1, x k-1 is the state variable at time k-1, B k-1 is the matrix B at time k-1, I k-1 is the output current at time k-1, w k-1 is the lithium-ion battery noise, k j is a constant, x k-j is a matrix variable, U k is the output voltage of the lithium-ion battery, is the C matrix, R0 is the internal resistance of the lithium-ion battery, I k is the current at time k, E k is the equivalent voltage source of the lithium-ion battery, U h is the hysteresis voltage of the lithium-ion battery, v k Measuring process noise for lithium-ion batteries.

[0016] In one embodiment of the present invention, the heat generated by the lithium-ion battery itself during the charging process is expressed as:

[0017]

[0018] Among them, Q is the heat generated by the lithium-ion battery, I is the current of the lithium-ion battery, u is the equivalent voltage of the lithium-ion battery, and U ocv is the open circuit voltage of the lithium-ion battery, T is the temperature of the lithium-ion battery, is the lithium-ion battery voltage, r i is the reaction rate of i, is the partial molar enthalpy of sample j, is the partial molar enthalpy of sample j at the average volume, c j is the concentration, t is the time, and i is the number of times.

[0019] In one embodiment of the present invention, the step of estimating parameters of the fractional-order coupling model using the dynamic discounted least squares method includes:

[0020] The parameters of the fractional-order coupling model are estimated by using the parameter identification method of the dynamic discounted least squares method to obtain the optimal solution of the fractional-order coupling model.

[0021] In one embodiment of the present invention, based on the fractional-order coupling model, the MMFCKF algorithm is used to jointly estimate the SOC and SOH states of the fractional-order coupling model, including:

[0022] Constructing a discrete state space model of a lithium-ion battery, wherein the discrete state space model is composed of a recursive form of a model state equation and an observation equation;

[0023] Defining the discrete state space model as a function of the SOC on the state space equation of the electrical model;

[0024] The SOC and SOH values ​​of the lithium-ion battery are taken as state variables, and the load voltage of the lithium-ion battery is taken as the observation variable, and the state equation and observation equation of the lithium-ion battery SOH are reconstructed;

[0025] Define the SOC of the fractional-order model of lithium-ion batteries according to the ampere-hour integration method;

[0026] In combination with the state-space equation, the MMFCKF algorithm is used to estimate the SOC and SOH of the lithium-ion battery.

[0027] In one embodiment of the present invention, the state equation and observation equation of the reconstructed lithium-ion battery SOH are:

[0028]

[0029] Among them, SOC(k+1) is the SOC at time k+1, SOC(k) is the SOC at time k, I(k) is the current at time k, Q max is the maximum capacity of lithium-ion battery, w k is the system noise, Q k+1 is the energy released by the lithium-ion battery at time k+1, Q k is the energy released by the lithium-ion battery at time k, r k is the process noise of the system, U k is the output voltage of the lithium-ion battery, is the C matrix, x k is the state vector at discrete time point k, R0 is the internal resistance of the lithium-ion battery, E k is the lithium-ion battery equivalent voltage source, v k To measure process noise.

[0030] In one embodiment of the present invention, the SOC of the fractional-order model of a lithium-ion battery is defined as follows according to the ampere-hour integration method:

[0031]

[0032] Where SOC0 is the initial SOC value of the lithium-ion battery, η is the coulombic efficiency of the lithium-ion battery, Q N is the electrical energy that can be stored in the lithium-ion battery when it leaves the factory, t is the time, and i is the output current of the lithium-ion battery.

[0033] In one embodiment of the present invention, the identification model of the dynamic discounted interest least squares method is:

[0034] z(k)=h T (k)θ(k)+e(k);

[0035] Among them, z(k) is the system output, h T (k) is the observable data vector, θ(k) is the parameter vector to be identified, e(k) is the noise and identification error, k is 1, 2, …, N, and N is the data length.

[0036] The present invention also provides a lithium-ion battery state joint estimation system based on a fractional-order coupling model, comprising the following modules:

[0037] A coupling model module is used to construct an electrical model of the lithium-ion battery based on a second-order equivalent circuit according to an input charging current; construct a two-state thermal model of the lithium-ion battery according to the impedance parameters of the electrical model and the heat generation and convection heat transfer relationship of the lithium-ion battery itself during the charging process, and construct a fractional-order coupling model of the lithium-ion battery based on the bidirectional coupling effect between the electrical model and the two-state thermal model;

[0038] A parameter identification module, used for estimating parameters of the fractional-order coupling model of the lithium-ion battery by using a dynamic discounted least squares method;

[0039] The state joint estimation module is used to perform SOC and SOH state joint estimation on the fractional-order coupling model based on the fractional-order coupling model by using the MMFCKF algorithm.

[0040] In one embodiment of the present invention, the identification model of the dynamic discounted least squares method in the parameter identification module is:

[0041] z(k)=h T (k)θ(k)+e(k);

[0042] Among them, z(k) is the system output, h T (k) is the observable data vector, θ(k) is the parameter vector to be identified, e(k) is the noise and identification error, k is 1, 2, …, N, and N is the data length.

[0043] The above technical solution of the present invention has the following advantages compared with the prior art:

[0044] The present invention discloses a method and system for jointly estimating the state of a lithium-ion battery based on a fractional-order coupling model. First, the fractional-order electrothermal coupling model of the lithium battery is constructed using fractional-order theory to target the hysteresis effect of the lithium battery and the influence of temperature on the battery model. Secondly, a parameter identification method of a dynamic discounted least squares method is used to introduce a dynamic discounted factor into the least squares method, and the model is adjusted in real time through the error, thereby overcoming the problem of data saturation of the traditional recursive least squares method. Finally, a robust fractional-order cubature Kalman filter based on the Masreliez-Martin (MM) method is proposed for the state estimation problem of a fractional-order discrete-time nonlinear system with non-Gaussian measurement noise. On the basis of the fractional-order discrete nonlinear dynamic system, the state prediction formula is derived using the cubature principle, and the state measurement update is realized using the MM method, thus forming a robust fractional-order cubature Kalman tracking algorithm based on the MM method (MM based robustfactional CKF, MMFCKF). The present application can accurately describe the nonlinear changes of the lithium-ion battery system during use, reduce the error of the battery model, and improve the accuracy of lithium battery state estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below according to specific embodiments of the present invention in conjunction with the accompanying drawings, wherein:

[0046] Figure 1A flow chart of a lithium-ion battery state joint estimation method based on a fractional-order coupling model provided in an embodiment of the present application is shown;

[0047] Figure 2 A functional module diagram of a lithium-ion battery state joint estimation system based on a fractional-order coupling model provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0048] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art, under the guidance of the content of the present application, can add one or more other operations to the flowchart, or remove one or more operations from the flowchart.

[0049] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0050] In order to enable those skilled in the art to use the contents of this application, the following implementation is given in combination with the specific application scenario "Joint estimation of lithium-ion battery state based on fractional-order coupling model". For those skilled in the art, the general principles defined here can be applied to other embodiments and application scenarios without departing from the spirit and scope of this application.

[0051] The following method of the embodiment of the present application can be applied to any scenario that requires joint estimation of the lithium-ion battery state based on a fractional-order coupling model. The embodiment of the present application is not limited to the specific application scenario. Any joint estimation of the lithium-ion battery state based on the fractional-order coupling model provided by the embodiment of the present application is within the scope of protection of the present application.

[0052] To facilitate the understanding of the present application, the technical solution provided by the present application is described in detail below in conjunction with specific embodiments.

[0053] Figure 1 A flowchart of a lithium-ion battery state joint estimation method based on a fractional-order coupling model provided in an embodiment of the present application. The method provided in an embodiment of the present application comprises the following steps:

[0054] S101: Constructing an electrical model of the lithium-ion battery based on a second-order equivalent circuit according to the input charging current.

[0055] For example, in order to accurately describe the dynamic and static characteristics of the lithium battery terminal voltage and reduce the model error, the present invention is based on a second-order RC circuit. In the second-order RC circuit: U is the output terminal voltage of the integer circuit; R0 is the internal resistance of the lithium-ion battery; U ocv is the open circuit voltage of the lithium-ion battery; R1, R2, C1 and C2 are the resistance and capacitance parameters of the second-order circuit; U1 and U2 are the voltages across capacitors C1 and C2 respectively; U h is the hysteresis voltage of lithium-ion battery.

[0056] In the second-order RC fractional equivalent circuit model, the impedance of the fractional capacitor can be defined as:

[0057]

[0058] in, is the impedance of the fractional capacitor, C i is a fractional-order capacitor, s is an operator, p is a power, and i is a constant 1, 2.

[0059] By using the inverse Laplace transform, the differential form of the impedance of the fractional capacitor is obtained, as shown in the following equation (2). Equation (2) describes the time domain relationship between the current flowing through the fractional capacitor and the voltage flowing through the fractional capacitor.

[0060]

[0061] Where I is the output current of the lithium-ion battery, t is the time, i is a constant 1, 2, R i is the resistance, is the voltage of the fractional capacitor.

[0062] According to Kirchhoff's voltage law, we can get:

[0063]

[0064] Among them, U ocv is the open circuit voltage of the lithium-ion battery, V0 is the equivalent voltage of the lithium-ion battery, I0 is the output current, R0 is the internal resistance of the lithium-ion battery, is the fractional capacitor 1 voltage, is the fractional capacitor 2 voltage, U his the hysteresis voltage of lithium-ion battery.

[0065] Therefore, the differential equation expression of the voltage of the second-order RC fractional equivalent circuit is:

[0066]

[0067] Among them, D m is the m-order differential form, D n is the nth-order differential form, C i is the fractional capacitor, R1 and R2 are the resistances of the second-order circuit, C1 and C2 are the capacitances of the second-order circuit, and m and n are the orders of the fractional capacitor elements.

[0068] The SOC of a lithium battery changes during use, and its differential form is described as:

[0069]

[0070] In formula (5), η is 1, which means that the lithium battery efficiency is 100%, Q N is the rated capacity of the lithium battery.

[0071] Therefore, the state space equation of the fractional-order lithium battery electrical model is:

[0072]

[0073] Where d is the derivative, U is the output voltage of the integer-order circuit, U1 and U2 are the voltages across capacitors C1 and C2 respectively. is the fractional capacitor voltage, is the fractional capacitor voltage, Q n is the electrical energy of lithium-ion battery, and S is SOC.

[0074] In some possible implementations, the state space equation of the electrical model is:

[0075]

[0076] In formula (7), k is a discrete time point, x k is the state vector at discrete time point k, A k-1 is the k-1 moment matrix A, x k-1 is the system state variable at time k-1, B k-1 is the k-1 moment matrix B, I k-1 is the current at time k-1, w k-1 is the lithium-ion battery noise, k j For x k-j For U k is the output voltage of the lithium-ion battery, is the C matrix, R0 is the internal resistance of the lithium-ion battery, I k is the current at time k, E k is the battery equivalent voltage source, U h is the hysteresis voltage of lithium-ion battery, v k Measuring process noise for lithium-ion batteries.

[0077] Exemplarily, the second-order RC fractional-order equivalent circuit model is discretized using the GL definition to obtain the above formula (7).

[0078] In formula (7), the weighting coefficient matrix x k is the state vector [U1(k), U2(k), SOC(k)] T ,

[0079] Furthermore, during the charging process, the heat generated by lithium-ion batteries is mainly divided into two parts: reversible heat and irreversible heat. Reversible heat mainly comes from the entropy change heat generated by the chemical reaction inside the battery, while irreversible heat is mainly caused by the heating of the ohmic internal resistance and polarization resistance inside the battery. Therefore, combined with the lithium-ion battery electrical model, the spatial expression of the lithium-ion battery thermal model is:

[0080]

[0081]

[0082] In formula (8) and formula (9), C c is the heat capacity of the lithium-ion battery; T c is the core temperature of the battery; T s is the battery surface temperature; R c It combines the thermal conductivity and contact thermal resistance of dense and inhomogeneous materials to simulate the heat exchange between the lithium-ion battery core and the surface; Q is the heat generated in the lithium-ion battery core; C s is the heat capacity of the battery casing; T f is the ambient temperature; R u is the equivalent convection resistance.

[0083] Assuming the sampling time is Δt, k is the discrete moment, the thermal model space expression is discretized to obtain:

[0084]

[0085]

[0086] S102: constructing a two-state thermal model of the lithium-ion battery according to the impedance parameters of the electrical model and the heat generation and convection heat transfer relationship of the lithium-ion battery itself during charging, and constructing a fractional-order coupling model of the lithium-ion battery based on the bidirectional coupling between the electrical model and the two-state thermal model.

[0087] In some possible implementations, the heat generated by the lithium-ion battery itself during the charging process is expressed as:

[0088]

[0089] Among them, Q is the heat generated by the lithium-ion battery, I is the current of the lithium-ion battery, u is the equivalent voltage of the lithium-ion battery, and U ocv is the open circuit voltage of the lithium-ion battery, T is the temperature of the lithium-ion battery, is the lithium-ion battery voltage, r i is the reaction rate of i, is the partial molar enthalpy of sample j, is the partial molar enthalpy of sample j at the average volume, c j is the concentration, t is the time, and i is the number of times.

[0090] For example, based on the above-mentioned coupling model, in order to solve the problem of low accuracy of parameter identification of lithium battery fractional-order coupling model in the solution process, a parameter identification method of dynamic discounted least squares method is adopted, a dynamic discount factor is introduced into the least squares method, and the model is adjusted in real time through the error, thus overcoming the problem of data saturation of traditional recursive least squares method.

[0091] S103: Based on the fractional-order coupling model, use a dynamic discounted least squares method to estimate parameters of the fractional-order coupling model.

[0092] For example, the dynamic discounted least squares method can not only solve the data saturation problem by reducing the trust of the original data and ensure the reliability of the algorithm's real-time tracking of parameters, but also balance the stability of the data's identification accuracy while maintaining the rapidity of identification. Among them, the dynamic forgetting factor can be set as a variable, and the weighting factor can be obtained through experiments to obtain a suitable constant. They are used as excitation inputs into the system, and the system can be regarded as a stable system with slowly changing parameters. In general, the dynamic range of the dynamic forgetting factor is 0.95-1.00, and the weighting factor selection range is 0.2-1.0. In practical applications, the dynamic discounted factor obtained through repeated experiments will significantly improve the identification accuracy of the recursive least squares method.

[0093] In some possible implementations, the identification model of the dynamic discounted least squares method is:

[0094] z(k)=h T(k)θ(k)+e(k) (13);

[0095] Among them, z(k) is the system output, h T (k) is the observable data vector, θ(k) is the parameter vector to be identified, e(k) is the noise and identification error, k is 1, 2, …, N, and N is the data length.

[0096] Further, wherein:

[0097] h(k)=[-z(k-1), ..., -z(kn a ), u(k-1), …, u(kn b )] (14);

[0098]

[0099]

[0100] In the above equations (14), (15), and (16), n is the number of parameters a and b to be identified, Γ(k, i) is the discount factor, and its relationship with the dynamic forgetting factor μ(j) and the weighting factor Λ(i) is as follows:

[0101]

[0102] By minimizing the objective function J(θ), the model parameters closest to the true value can be obtained. Parameter estimates at time k for:

[0103]

[0104]

[0105]

[0106] In the above formulas (18), (19), and (20), is the estimated value of the parameter at time k, K(k) is the gain matrix, P(k) is the covariance matrix, h (i) is the observable data, z (i) is the system output, h (k) For observation data.

[0107] The absolute value of the difference between the theoretical value and the actual value is ε(k), which can be used as the key value for adjusting the size of the dynamic forgetting factor. The value of ε(k) at time k is:

[0108]

[0109] During the identification process, the dynamically adjusted ε(k) can be used as an indicator to adjust μ(k). When ε(k) is large, the value of μ(k) can be reduced to allow the algorithm to converge quickly. When ε(k) is small, the value of μ(k) can be increased to allow the algorithm to maintain high robustness and identification accuracy.

[0110] Based on the above analysis, a dynamic forgetting factor function can be established:

[0111] μ(k)=α+(1-α)e -γε(k) (twenty two);

[0112] In formula (22), α is a positive adjustable number close to and less than 1, γ is a positive adjustable parameter, and e is the base.

[0113] It can be seen from the above formula (22) that the larger the value of ε(k), the closer the value of μ(k) is to α; the smaller the value of ε(k), the closer the value of μ(k) is to 1.

[0114] The following dynamic discounted interest recursive least squares equation is obtained:

[0115]

[0116] K(k)=P(k-1)h(k)×[h T (k)P(k-1)h(k)+μ(k) / Λ(k)] -1 (twenty four);

[0117] P(k)=1 / μ(k)[IK(k)h T (k)]P(k) (25);

[0118] In summary, the basic steps of coupling model parameter identification are:

[0119] 1) Obtaining initial observation data parameters through measurement;

[0120] 2) Calculate the parameter estimate θ;

[0121] 3) By collecting the observed data h(k), the data of the last estimate is corrected and continuously updated

[0122] 4) Determine whether the estimated parameter value meets the required accuracy. If so, proceed to 5); otherwise, go to 3);

[0123] 5) Output the final result and the algorithm ends.

[0124] In some possible implementations, the using the dynamic discounted least squares method to estimate the parameters of the fractional-order coupling model includes:

[0125] The parameters of the fractional-order coupling model are estimated by using the parameter identification method of the dynamic discounted least squares method to obtain the optimal solution of the fractional-order coupling model.

[0126] Exemplarily, the lithium battery parameter estimation module uses the parameter identification method of the dynamic discounted least squares method to obtain the optimal solution of the coupling model: R0, R1, R2, C1, C2, m<1, n<1, R c , R u It is found that the coupling model is fractional-order. Compared with the integer-order battery model, the maximum error and root mean square error of the fractional-order model are smaller. It can be seen that the fractional-order model has better accuracy and robustness than the integer-order model under complex working conditions. On this basis, the influence of temperature on the model accuracy is considered to further improve the accuracy of the coupling model.

[0127] S104: Based on the fractional-order coupling model, use the MMFCKF algorithm to jointly estimate the SOC and SOH states of the fractional-order coupling model.

[0128] In some possible implementations, based on the fractional-order coupling model, the MMFCKF algorithm is used to jointly estimate the SOC and SOH states of the fractional-order coupling model, including:

[0129] Constructing a discrete state space model of a lithium-ion battery, wherein the discrete state space model is composed of a recursive form of a model state equation and an observation equation;

[0130] Defining the discrete state space model as a function of the SOC on the state space equation of the electrical model;

[0131] The SOC and SOH values ​​of the lithium-ion battery are taken as state variables, and the load voltage of the lithium-ion battery is taken as the observation variable, and the state equation and observation equation of the lithium-ion battery SOH are reconstructed;

[0132] Define the SOC of the fractional-order model of lithium-ion batteries according to the ampere-hour integration method;

[0133] In combination with the state-space equation, the MMFCKF algorithm is used to estimate the SOC and SOH of the lithium-ion battery.

[0134] For example, a robust fractional-order cubature Kalman filter based on the Masreliez-Martin (MM) method is proposed for the state estimation problem of fractional-order discrete-time nonlinear systems with non-Gaussian measurement noise. On the basis of the fractional-order discrete nonlinear dynamic system, the state prediction formula is derived using the third-order cubature principle, and the state measurement update is realized using the MEM method, forming a robust fractional-order cubature Kalman tracking algorithm based on the MM method (i.e., M-Mbased robust factional CKF, MMFCKF algorithm).

[0135] First, a discrete state space model of lithium-ion batteries is established, which consists of the recursive form of the model state equation and the observation equation. It is defined as a function of SOC based on the state space equation of the fractional-order lithium-ion battery electrical model, with the lithium-ion battery SOC and SOH values ​​as state variables and the lithium-ion battery load voltage as the observation variable, to build a new equation. Its expression is as follows (26).

[0136] In some possible implementations, the state equation and observation equation of the reconstructed lithium-ion battery SOH are:

[0137]

[0138] Among them, SOC(k+1) is the SOC at time k+1, SOC(k) is the SOC at time k, I(k) is the current at time k, Q max is the maximum capacity of lithium-ion battery, W k is the system noise, Q k+1 is the energy released by the lithium-ion battery at time k+1, Q k is the energy released by the lithium-ion battery at time k, r k is the process noise of the system, U k is the output voltage of the lithium-ion battery, is the C matrix, x k is the state vector at discrete time point k, R0 is the internal resistance of the lithium-ion battery, E k is the lithium-ion battery equivalent voltage source, v k To measure process noise.

[0139] In some possible implementations, the SOC of the fractional-order model of a lithium-ion battery is defined as follows according to the ampere-hour integration method:

[0140]

[0141] Where SOC0 is the initial SOC value of the lithium-ion battery, η is the coulombic efficiency of the lithium-ion battery, Q Nis the electrical energy that can be stored in the lithium-ion battery when it leaves the factory, t is the time, and i is the output current of the lithium-ion battery.

[0142] Combining the state space equation of the fractional-order lithium battery model, the steps of estimating the SOC and SOH of the lithium battery using the robust fractional-order volumetric Kalman tracking algorithm based on the MM method are as follows:

[0143]

[0144]

[0145] In formula (28) and formula (29), is the battery model mean, is the battery model variance, E is the expectation, is the state variable, and P is the system variance.

[0146] Calculate Sigma points (k = 0, 1, 2...), select Sigma points:

[0147]

[0148] Among them, x is the state variable, χ0 is the initial Sigma point mean, χ i is the mean of the ith Sigma point, χ i+n is the mean of the i+nth Sigma point, n is the dimension, λ is a constant, P x is the system variance.

[0149] calculate Calculate volume point ξ j And the volume point calculation value X j,k-1 Propagation through the system function f(x):

[0150] X j,k-1 =S k-1 ξ j +x^ k-1 (31);

[0151]

[0152] Among them, X is the calculated value of the volume point, ξ is the volume point, S k-1 is the Cholesky decomposition matrix, ξ j Volume point set, is the predicted value at the next moment. k represents the current moment, i and j are positive integers, (·) j,k It represents the predicted value j steps ahead relative to time k.

[0153]

[0154] In formula (33), To estimate the predicted value,

[0155]

[0156] Among them, P k is the prediction error covariance, R k is the system noise, Z is the range prediction value, and X is the volume point calculation value.

[0157] Compute the square root of the prediction variance:

[0158]

[0159] in, is the square root of the system variance, T is the matrix transpose, for chol{P k}.

[0160] Calculate volume points:

[0161]

[0162] Among them, X j,k is the volume point calculation value, is the estimated predicted value, ξ is the volume point.

[0163] Compute volume points propagating through nonlinear measurement equations:

[0164] γ j,k =h(X j,k ) (37);

[0165] Among them, γ j,k is the volume point of the nonlinear measurement equation propagation, h is the transformation function, X j,k Calculates values ​​for volume points.

[0166] Compute measurement forecasts, innovation variance, and covariance estimates:

[0167]

[0168]

[0169]

[0170] in, For measurement prediction, P z,k is the variance of the new interest, P xz,k is the covariance, R 1,k System measurement noise.

[0171] Using the MM method, calculate the state estimate and covariance:

[0172]

[0173]

[0174] in, is the state estimation, P k is the system covariance, v k is the state vector, K k For gain.

[0175]

[0176]

[0177] Among them, K k For gain, is the volume point calculation function, P xz,k is the covariance.

[0178] ψ(v k ) and its derivative are defined as:

[0179] ψ(v k )=[ψ(v k,1 ), ψ(v k,2 ),…,ψ(v k,m )] T (45);

[0180]

[0181] In equations (45) and (46), m is the dimension of the vector, and ψ(·) is the Huber function, which is defined as:

[0182]

[0183] In formula (47), L is a constant and t is a time constant.

[0184] Based on the same application concept, the embodiment of the present application also provides a lithium-ion battery state joint estimation system based on a fractional-order coupling model corresponding to the lithium-ion battery state joint estimation method based on a fractional-order coupling model provided in the above embodiment. Since the principle of solving the problem of the system in the embodiment of the present application is similar to the method in the above embodiment of the present application, the implementation of the system can refer to the implementation of the method, and the repeated parts will not be repeated.

[0185] like Figure 2 As shown, Figure 2 A functional module diagram of a lithium-ion battery state joint estimation system based on a fractional-order coupling model provided in an embodiment of the present application includes the following modules:

[0186] The coupling model module 100 is used to construct an electrical model of the lithium-ion battery based on a second-order equivalent circuit according to the input charging current; construct a two-state thermal model of the lithium-ion battery according to the impedance parameters of the electrical model and the heat generation and convection heat transfer relationship of the lithium-ion battery itself during the charging process, and construct a fractional-order coupling model of the lithium-ion battery based on the bidirectional coupling effect between the electrical model and the two-state thermal model;

[0187] A parameter identification module 200 is used to estimate the parameters of the fractional-order coupling model based on the fractional-order coupling model of the lithium-ion battery by using a dynamic discounted least squares method;

[0188] The state joint estimation module 300 is used to perform SOC and SOH state joint estimation on the fractional-order coupling model based on the fractional-order coupling model by using the MMFCKF algorithm.

[0189] In some possible implementations, the identification model of the dynamic discounted least squares method in the parameter identification module 200 is: z(k)=h T (k)θ(k)+e(k); where z(k) is the system output, h T (k) is the observable data vector, θ(k) is the parameter vector to be identified, e(k) is the noise and identification error, k is 1, 2, …, N, and N is the data length.

[0190] In summary, the lithium-ion battery state joint estimation method and system provided in the embodiment of the present application establishes a fractional-order electrothermal coupling model of lithium batteries based on hysteresis voltage in view of the battery hysteresis effect and the influence of temperature on the battery model; on the basis of the coupling model, in view of the problem that the traditional recursive least squares method has poor accuracy in identifying model parameters, a parameter identification method of a dynamic discounted least squares method is adopted, a dynamic discounted factor is introduced into the least squares method, and the model is adjusted in real time through the error, thereby overcoming the problem of data saturation of the traditional recursive least squares method; in view of the state estimation problem of fractional-order discrete-time nonlinear systems with non-Gaussian measurement noise, a robust fractional-order cubature Kalman filter (MMFCKF) based on the Masreliez-Martin (MM for short) method is proposed, on the basis of the fractional-order discrete nonlinear dynamic system, the state prediction formula is derived using the cubature principle, and the state measurement update is realized using the MM method, thereby forming a robust fractional-order cubature Kalman tracking algorithm based on the MM method. Its advantages are: the method of establishing a fractional-order electro-thermal coupling model based on the battery hysteresis effect improves the accuracy of the battery model and has good robustness. The proposed parameter identification method can identify the model parameters with high accuracy. The proposed lithium battery state joint estimation method can realize the joint estimation of S0C and SOH of lithium batteries.

[0191] Those skilled in the art should understand that the embodiments of the present application can be provided as methods and systems. Therefore, the present application can be in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware.

[0192] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.

Claims

1. A lithium-ion battery state joint estimation method based on a fractional-order coupling model, characterized in that: The method comprises: According to the input charging current, the electrical model of the lithium-ion battery is constructed based on the second-order equivalent circuit; A thermal model of the lithium-ion battery is constructed based on the impedance parameters of the electrical model and the heat generation and convection heat transfer relationship of the lithium-ion battery itself during charging, and a fractional-order coupling model of the lithium-ion battery is constructed based on the bidirectional coupling effect between the electrical model and the thermal model; Based on the fractional-order coupling model, a parameter identification method of a dynamic discounted least square method is used to estimate the parameters of the fractional-order coupling model to obtain an optimal solution of the fractional-order coupling model; The dynamic discount least squares equation is: K(k)=P(k-1)h(k)×[h T (k)P(k-1)h(k)+μ(k) / Λ(k)] -1 ; P(k)=1 / μ(k)[IK(k)h T (k)]P(k) in, is the estimated value of the parameter at time k; is the estimated value of the parameter at time k-1; K(k) is the gain matrix; z(k) is the system output; h T (k) is the observable data vector; P(k) is the covariance matrix; μ(k) is the dynamic forgetting factor; h(k) is the observed data; Λ(k) is the weighting factor; The step of estimating the coupling model parameters comprises: 1) Obtaining initial observation data parameters through measurement; 2) Calculate the parameter estimate θ; 3) By collecting the observed data h(k), the data of the last estimate is corrected and continuously updated 4) Determine whether the estimated parameter value meets the required accuracy. If so, proceed to 5); otherwise, go to 3); 5) Output the final result and the algorithm ends; Based on the fractional-order coupling model, the SOC and SOH states of the fractional-order coupling model are jointly estimated using a robust fractional-order cubic Kalman filter (MMFCKF) algorithm. The steps of performing joint estimation of SOC and SOH states on the fractional-order coupling model using the robust fractional-order cubic Kalman filter (MMFCKF) algorithm are as follows: 1) Establish the mean value of battery model and system variance P; 2) Calculate Sigma points; 3) Calculate the square root of the prediction variance With volume point X j,k ; 4) Calculate measurement forecasts, innovation variance and covariance estimates; 5) Using the Masreliez-Martin method, calculate the state estimate and variance: is the state estimation, P k is the system covariance, v k is the state vector, K k For gain.

2. The lithium-ion battery state joint estimation method based on the fractional-order coupling model according to claim 1 is characterized in that: The state space equation of the electrical model is: Where k is a discrete time point, x k is the state vector at discrete time point k, A k-1 is the matrix A at time k-1, x k-1 is the state variable at time k-1, B k-1 is the matrix B at time k-1, I k-1 is the output current at time k-1, w k-1 is the lithium-ion battery noise, k j is a constant, x k-j is a matrix variable, U k is the output voltage of the lithium-ion battery, is the C matrix, R0 is the internal resistance of the lithium-ion battery, I k is the current at time k, E k is the equivalent voltage source of the lithium-ion battery, U h is the hysteresis voltage of the lithium-ion battery, v k Measuring process noise for lithium-ion batteries.

3. The lithium-ion battery state joint estimation method based on the fractional-order coupling model according to claim 1 is characterized in that: The heat generated by the lithium-ion battery itself during the charging process is expressed as: Among them, Q is the heat generated by the lithium-ion battery, I is the current of the lithium-ion battery, u is the voltage of the lithium-ion battery, and U ocv is the open circuit voltage of the lithium-ion battery, T is the temperature of the lithium-ion battery, is the lithium-ion battery voltage, r i is the reaction rate of i, is the partial molar enthalpy of sample j, is the partial molar enthalpy of sample j at the average volume, c j is the concentration, t is the time, and i is the number of times.

4. The lithium-ion battery state joint estimation method based on the fractional-order coupling model according to claim 1 is characterized in that: Based on the fractional-order coupling model, a robust fractional-order cubature Kalman filter algorithm is used to jointly estimate the SOC and SOH states of the fractional-order coupling model, including: Constructing a discrete state space model of a lithium-ion battery, wherein the discrete state space model is composed of a recursive form of a model state equation and an observation equation; Defining the discrete state space model as a function of the SOC on the state space equation of the electrical model; The SOC and SOH values ​​of the lithium-ion battery are taken as state variables, and the load voltage of the lithium-ion battery is taken as the observation variable, and the state equation and observation equation of the lithium-ion battery SOH are reconstructed; Define the SOC of the fractional-order model of lithium-ion batteries based on the ampere-hour integration method; In combination with the state-space equation, the MMFCKF algorithm is used to estimate the SOC and SOH of the lithium-ion battery.

5. The lithium-ion battery state joint estimation method based on the fractional-order coupling model according to claim 4 is characterized in that: The state equation and observation equation of the reconstructed lithium-ion battery SOH are: Among them, SOC(k+1) is the SOC at time k+1, SOC(k) is the SOC at time k, I(k) is the current at time k, Q max is the maximum capacity of lithium-ion battery, w k is the system noise, Q k+1 is the energy released by the lithium-ion battery at time k+1, Q k is the energy released by the lithium-ion battery at time k, r k is the process noise of the system, U k is the output voltage of the lithium-ion battery, is the C matrix, x k is the state vector at discrete time point k, R0 is the internal resistance of the lithium-ion battery, E k is the lithium-ion battery equivalent voltage source, v k To measure process noise.

6. The lithium-ion battery state joint estimation method based on the fractional-order coupling model according to claim 4 is characterized in that: According to the ampere-hour integration method, the SOC of the fractional-order model of lithium-ion batteries is defined as: Where SOC0 is the initial SOC value of the lithium-ion battery, η is the coulombic efficiency of the lithium-ion battery, Q N is the electrical energy that can be stored in the lithium-ion battery when it leaves the factory, t is the time, and i is the output current of the lithium-ion battery.

7. The lithium-ion battery state joint estimation method based on the fractional-order coupling model according to claim 1 is characterized in that: The identification model of the dynamic discounted least squares method is: with(k)=h T (k)θ(k)+e(k); Among them, z(k) is the system output, h T (k) is the observable data vector, θ(k) is the parameter vector to be identified, e(k) is the noise and identification error, k is 1, 2, …, N, and N is the data length.

8. A lithium-ion battery state joint estimation system based on a fractional-order coupling model, characterized in that: Includes the following modules: A coupling model module is used to construct an electrical model of the lithium-ion battery based on a second-order equivalent circuit according to an input charging current; construct a two-state thermal model of the lithium-ion battery according to the impedance parameters of the electrical model and the heat generation and convection heat transfer relationship of the lithium-ion battery itself during the charging process, and construct a fractional-order coupling model of the lithium-ion battery based on the bidirectional coupling effect between the electrical model and the two-state thermal model; A parameter identification module, used for estimating parameters of the fractional-order coupling model of the lithium-ion battery by using a dynamic discounted least squares method; A state joint estimation module, used for performing SOC and SOH state joint estimation on the fractional-order coupling model based on the fractional-order coupling model by using a robust fractional-order cubature Kalman filter algorithm; The parameters of the fractional-order coupling model are estimated by using the parameter identification method of the dynamic discounted least squares method to obtain the optimal solution of the fractional-order coupling model; The dynamic discount recursive least squares equation is: K(k)=P(k-1)h(k)×[h T (k)P(k-1)h(k)+μ(k) / Λ(k)] -1 ; P(k)=1 / μ(k)[IK(k)h T (k)]P(k); in, is the estimated value of the parameter at time k; is the estimated value of the parameter at time k-1; K(k) is the gain matrix; z(k) is the system output; h T (k) is the observable data vector; P(k) is the covariance matrix; μ(k) is the dynamic forgetting factor; h(k) is the observed data; Λ(k) is the weighting factor; The step of estimating the coupling model parameters comprises: 1) Obtaining initial observation data parameters through measurement; 2) Calculate the parameter estimate θ; 3) By collecting the observed data h(k), the data of the last estimate is corrected and continuously updated 4) Determine whether the estimated parameter value meets the required accuracy. If so, proceed to 5); otherwise, go to 3); 5) Output the final result and the algorithm ends; Based on the fractional-order coupling model, the SOC and SOH states of the fractional-order coupling model are jointly estimated using a robust fractional-order cubic Kalman filter (MMFCKF) algorithm. The steps of performing joint estimation of SOC and SOH states on the fractional-order coupling model using the robust fractional-order cubic Kalman filter (MMFCKF) algorithm are as follows: 6) Establish the battery model mean and system variance P; 7) Calculate Sigma points; 8) Calculate the square root of the prediction variance With volume point X j,k ; 9) Calculate measurement forecasts, innovation variance and covariance estimates; 10) Using the Masreliez-Martin method, calculate the state estimate and variance: The state estimates and covariances are calculated based on: is the state estimation, P k is the system covariance, v k is the state vector, K k For gain.

9. The lithium-ion battery state joint estimation system based on fractional-order coupling model according to claim 8, characterized in that: The identification model of the dynamic discounted least squares method in the parameter identification module is: with(k)=h T (k)θ(k)+e(k); Among them, z(k) is the system output, h T (k) is the observable data vector, θ(k) is the parameter vector to be identified, e(k) is the noise and identification error, k is 1, 2, …, N, and N is the data length.