Lithium-ion battery SOC estimation method and system based on unknown input observer

Abstract

This invention proposes a lithium-ion battery SOC estimation method and system based on an unknown input observer, relating to the field of state estimation. The method comprises: constructing a battery state-space model based on the equivalent circuit model and terminal voltage principle of a lithium battery; defining an observer equation using the excitation current as an unknown input, and constructing an error dynamic equation based on the battery state-space model and the observer equation; simplifying the error dynamic equation based on stability constraints, and verifying the stability by constructing a Lyapunov function; obtaining the open-circuit voltage based on the current state of charge using a lookup table of open-circuit voltage and state of charge, and updating the observation matrix; iteratively solving the observer equation based on the updated observation matrix and the measured voltage to estimate the excitation current, and inputting the excitation current into the battery state space to complete the state of charge estimation. By introducing voltage measurement deviation as a state variable into the battery system model and designing a stable observer, the accuracy and robustness of SOC estimation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of state estimation, and in particular to a lithium-ion battery SOC estimation method and system based on an unknown input observer. Background Art

[0002] Lithium-ion batteries (Li-ion batteries), due to their high energy density, lightweight design, and long cycle life, have become the primary energy storage unit in portable devices such as smartphones, tablets, and wearables. In these devices, the Battery Management System (BMS) must accurately estimate the Li-ion battery's state of charge (SOC) to optimize power usage, extend battery life, and ensure stable device operation. Traditional SOC estimation methods are typically based on equivalent circuit models, measuring voltage and current data and combining parameter identification algorithms (such as the least squares method) or state estimation algorithms (such as the Kalman filter).

[0003] However, these methods usually assume that the terminal voltage measurement is accurate and rely on the current sensor to provide accurate current input. In practical applications, due to factors such as the external environment and the accuracy of the measuring instrument, the terminal voltage measurement often has a fixed deviation. In addition, in order to reduce costs or simplify hardware design, current sensors may not be used in some scenarios. These factors cause the traditional method to have a significant decrease in the accuracy of SOC estimation when the voltage measurement deviation is large or there is no current data. For example, the traditional current sensor-free method usually directly derives the SOC from the voltage signal, but does not fully consider the impact of the voltage measurement deviation on the estimation result, which can easily cause the SOC estimation error to accumulate or even diverge. Summary of the Invention

[0004] To solve the above problems, the present invention proposes a lithium-ion battery SOC estimation method and system based on an unknown input observer. By introducing the voltage measurement deviation as a state variable into the battery system model and designing a stable observer, the limitations of measurement deviation and no current input are effectively overcome, thereby improving the accuracy and robustness of SOC estimation.

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

[0006] In a first aspect, the present invention provides a lithium-ion battery SOC estimation method based on an unknown input observer, comprising:

[0007] Based on the equivalent circuit model and terminal voltage principle of lithium batteries, a battery state space model is constructed;

[0008] An observer equation is defined by taking the excitation current as an unknown input, and the battery state space model and the observer equation are substituted into the observer error equation to construct an error dynamic equation;

[0009] Based on the observer noise immunity and stability constraints and the gain matrix, the error dynamic equation is simplified to obtain the final error dynamic equation, and the stability is verified by constructing the Lyapunov function;

[0010] According to the lookup table of open circuit voltage and state of charge, the open circuit voltage information is obtained based on the current state of charge, and the observation matrix in the battery state space model is updated;

[0011] Based on the updated observation matrix and the measured voltage, the observer equation is iteratively solved to estimate the excitation current, which is then input into the battery state space to complete the state of charge estimation.

[0012] In a second aspect, the present invention provides a lithium-ion battery SOC estimation system based on an unknown input observer, comprising:

[0013] State modeling module, used to build a battery state space model based on the equivalent circuit model and terminal voltage principle of lithium batteries;

[0014] An observer design module is used to define an observer equation using the excitation current as an unknown input, substitute the battery state space model and the observer equation into the observer error equation, and construct an error dynamic equation;

[0015] a gain stabilization module for simplifying the error dynamics equation based on observer noise immunity and stability constraints and a gain matrix to obtain a final error dynamics equation and verifying stability by constructing a Lyapunov function;

[0016] A table lookup and tuning module is used to obtain open circuit voltage information based on the current state of charge according to a lookup table of open circuit voltage and state of charge, and to update the observation matrix in the battery state space model;

[0017] The estimation module is used to iteratively solve the observer equation to estimate the excitation current based on the updated observation matrix and the measured voltage, input the excitation current into the battery state space, and complete the estimation of the state of charge.

[0018] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the lithium-ion battery SOC estimation method based on an unknown input observer described in the first aspect.

[0019] In a fourth aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the lithium-ion battery SOC estimation method based on an unknown input observer described in the first aspect are implemented.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] (1) Traditional methods rely on accurate terminal voltage measurement and precise current input, but in practice, the terminal voltage has a fixed deviation, and some scenarios do not have current sensors, resulting in a decrease in the accuracy of SOC estimation. To solve the above problems, the present invention constructs a battery state space model, taking into account the error factors that may exist in actual situations; the excitation current is used as an unknown input to define the observer equation, getting rid of the absolute dependence on the current sensor. The reliability of the estimation process is ensured by constructing an error dynamic equation and verifying the stability. The observation matrix is ​​updated according to the open circuit voltage and state of charge lookup table to reduce the influence of voltage measurement deviation. The observer equation is iteratively solved to estimate the excitation current and complete the state of charge estimation, effectively avoiding the problems of error accumulation and divergence in traditional methods and improving the SOC estimation accuracy in complex actual scenarios.

[0022] (2) By using the fixed measurement deviation as a state variable and introducing an unknown input observer, the present invention can effectively overcome the influence of the terminal voltage measurement deviation and achieve SOC estimation. Experimental results show that the estimated root mean square error is significantly lower than that of the traditional current sensorless method.

[0023] (3) The present invention does not require a current sensor and only relies on voltage measurement data. There is no need to install an additional current sensor, which reduces hardware costs and simplifies system design.

[0024] (4) Through reasonable observer design and dynamic adjustment of the observation matrix, the present invention can maintain stable estimation performance under different working conditions and measurement deviations and has strong robustness.

[0025] (5) The present invention is based on matrix operations, where K is a predetermined constant. It does not require complex online optimization and has moderate computational complexity, making it suitable for embedded system applications.

[0026] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their description are used to explain the present invention but do not constitute a limitation of the present invention.

[0028] Figure 1 A main flow chart of a lithium-ion battery SOC estimation method based on an unknown input observer provided by an embodiment of the present invention;

[0029] Figure 2 Schematic diagram of a first-order RC equivalent circuit model provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0031] Example 1

[0032] like Figure 1 As shown, this embodiment discloses a lithium-ion battery SOC estimation method based on an unknown input observer, comprising the following steps:

[0033] S1: Construct a battery state space model based on the equivalent circuit model and terminal voltage principle of lithium batteries;

[0034] S2: defining an observer equation using the excitation current as an unknown input, substituting the battery state space model and the observer equation into the observer error equation, and constructing an error dynamic equation;

[0035] S3: Based on the observer noise immunity and stability constraints and the gain matrix, the error dynamics equation is simplified to obtain the final error dynamics equation, and the stability is verified by constructing the Lyapunov function;

[0036] S4: According to the lookup table of open circuit voltage and state of charge, the open circuit voltage information is obtained based on the current state of charge, and the observation matrix in the battery state space model is updated;

[0037] S5: Based on the updated observation matrix and the measured voltage, the observer equation is iteratively solved to estimate the excitation current, and the excitation current is input into the battery state space to complete the state of charge estimation.

[0038] Next, combine Figure 1 , a lithium-ion battery SOC estimation method based on an unknown input observer disclosed in this embodiment is described in detail.

[0039] (1) Establishing a battery state space model

[0040] To achieve this goal, we first construct a state space representation of the battery system based on the equivalent circuit model of the lithium-ion battery. This model comprehensively considers the battery's state of charge. , polarization voltage and fixed measurement deviation The dynamic evolution of the excitation current and process noise. Considering the calculation accuracy and complexity, the model adopts the first-order RC equivalent circuit, such as Figure 2 shown.

[0041] Based on first-order Equivalent circuit model, the state variables of the battery system include 、 and . Its dynamic expressions are as follows:

[0042] 1. The discretization expression of is:

[0043] (1)

[0044] in, is the sampling period, For The associated dynamic time constant, is the equivalent resistance, is the excitation current, is the process noise component.

[0045] 2. Polarization voltage The discretization expression of is:

[0046] (2)

[0047] in, is the polarization time constant, , is the polarization resistance, is the polarized capacitance; is the process noise component. Figure 2 middle, is the terminal voltage, is the equivalent capacitance.

[0048] 3. Fixed measurement deviation is a constant value, and its expression is:

[0049] (3)

[0050] According to the first-order Equivalent circuit model, terminal voltage It can be expressed as:

[0051] (4)

[0052] in, is the open circuit voltage, and It is a nonlinear relationship and can be approximately linearized as ,in is the linear slope, is the intercept; is the polarization voltage; is the internal resistance voltage drop; To measure noise.

[0053] To facilitate observer design, define the output for:

[0054] (5)

[0055] Substituting into formula (4), we get:

[0056] (6)

[0057] Combining the dynamics of the above state variables, the system is represented as a state space model as follows:

[0058] (7)

[0059] Among them, the state vector is defined as:

[0060] (8)

[0061] The process noise and measurement noise vectors are:

[0062] (9)

[0063] in, 、 、 is the process noise component, 、 、 Measure the noise component.

[0064] The state transfer matrix, input matrix, observation matrix, and internal resistance are as follows:

[0065] ; ; ; (10)

[0066] The process noise and measurement noise distribution matrix are as follows:

[0067] (11)

[0068] Due to internal resistance (about 0.04794Ω) is relatively small compared to the magnitude of the open circuit voltage and polarization voltage, and under typical operating conditions of portable devices, the excitation current The amplitude is limited, resulting in Item pair output The contribution of is usually at the millivolt level. To simplify the observer design and improve computational efficiency, this embodiment ignores the influence of this term when designing the unknown input observer and simplifies the state space model to:

[0069] (12)

[0070] This embodiment comprehensively considers the dynamic evolution of the battery's state of charge, polarization voltage, and fixed measurement bias, and uses a first-order RC equivalent circuit to construct a model. Compared to traditional methods, this method is more realistic, laying a solid foundation for subsequent accurate SOC estimation, avoiding estimation errors caused by model simplification, and improving the accuracy of SOC estimation in complex scenarios.

[0071] 2. Design of unknown input observer

[0072] Estimated state , design the following observer equation:

[0073] (13)

[0074] in, is the intermediate state variable of the observer, is the estimated state vector, 、 、 、 Design the matrix for the observer, which must satisfy the stability condition; represents the state variable vector of the battery state estimation system at time k, 、 Represent the measured voltages at time k and k+1 respectively. is the estimated excitation current, which is treated as an unknown input. The observer error is defined as:

[0075] (14)

[0076] In this embodiment, the excitation current is used as an unknown input to define the observer equation, eliminating the absolute reliance on current sensors and resolving the challenge of accurate SOC estimation in some scenarios where current sensors are unavailable. Even when precise current data is unavailable, the observer equation can be used to estimate SOC, reducing hardware requirements and simplifying system design. This also avoids the accumulation and divergence of estimation errors caused by inaccurate current measurements or the absence of current data.

[0077] To ensure that the observer is stable and consistent with the unknown input and process noise It is irrelevant and requires detailed derivation of the error dynamic equation and design of a suitable matrix. The following is the derivation process of the error equation: Substitute the state equation of Equation (12) and the observer equation of (13) into Equation (14) and expand it step by step:

[0078] (15)

[0079] Where I is the identity matrix.

[0080] By constructing an error dynamics equation, this embodiment clearly demonstrates the variation in the error between the observer's estimated value and the actual value. This facilitates in-depth analysis of system performance and identifies the root causes of the error. This provides a key basis for optimizing observer design and improving SOC estimation accuracy, and is crucial for achieving accurate SOC estimation.

[0081] In order to eliminate the influence of noise on the observer, this embodiment proposes observer noise immunity and stability constraints, specifically:

[0082] To make the observer and process noise No relation, need to meet the following requirements:

[0083] (16)

[0084] To make the observer and measurement noise No relation, need to meet the following requirements:

[0085] (17)

[0086] To make the observer and the unknown input No relation, need to meet the following requirements:

[0087] (18)

[0088] Where T is the intermediate transition matrix introduced in the observer design and is determined by I-EC.

[0089] Substitute equations (16)-(18) into equation (15) and continue to simplify:

[0090] (19)

[0091] When the observer error reaches 0, and The value of may be arbitrary. To solve this problem, the gain matrix is ​​introduced , increasing design freedom:

[0092] (20)

[0093] In this embodiment, by setting 、 、 Constraints such as these are used, and the gain matrix is ​​introduced to effectively eliminate process noise. , measurement noise and unknown input This significantly improves the stability of the observer, removes obstacles for the derivation of the dynamic error equation, enables more accurate acquisition of the final form of the dynamic error equation, and enhances the system's adaptability and reliability to complex working conditions.

[0094] Note that, assuming When , the equation becomes:

[0095] (twenty one)

[0096] remember:

[0097] (twenty two)

[0098] At this point, the error dynamics equation is finally simplified to:

[0099] (twenty three)

[0100] (3) Stability conditions and Lyapunov analysis

[0101] After the above observer design is completed, the error dynamic equation is finally simplified to:

[0102] (twenty four)

[0103] in, is the state estimation error; is the observer dynamic matrix; is the gain matrix to be designed; is the disturbance distribution matrix; To measure the noise, assume Bounded noise.

[0104] This example uses the Lyapunov method to analyze the stability of the error dynamics. The goal is to prove that the system is exponentially stable when there is no disturbance and satisfies the bounded disturbance. stability while quantifying the performance level γ.

[0105] 1. Lyapunov function construction

[0106] Define the Lyapunov function:

[0107] (25)

[0108] Where P is a symmetric positive definite matrix, and the function is used to measure The energy is analyzed and the stability is judged by the change of the analyzer over time.

[0109] 2. Stability without disturbance

[0110] When measuring noise When , the error dynamics is:

[0111] (26)

[0112] like After designing a suitable K, the K value makes the eigenvalues ​​of the N matrix less than 1, satisfying Schur stability, and the system is globally exponentially stable in the absence of disturbances.

[0113] 3. Under disturbance conditions stability

[0114] When measurement noise exists, the error dynamics are as shown in Equation (24).

[0115] Define the disturbance term:

[0116] (27)

[0117] but:

[0118] (28)

[0119] The Lyapunov function difference is:

[0120] (29)

[0121] To ensure Stability, introducing attenuation conditions:

[0122] (30)

[0123] in, is the decay rate; define the positive definite matrix is the disturbance gain. By designing appropriate K and P, the impact of disturbance on the error can be controlled.

[0124] It should be understood that the design of appropriate K and P is achievable by those skilled in the art.

[0125] This embodiment uses the Lyapunov method to construct a function to analyze stability, which can theoretically and rigorously prove the stability of the system in both unperturbed and disturbed conditions. This provides a reliable theoretical basis for the SOC estimation process, ensuring the accuracy and stability of the estimation results, avoiding fluctuations in estimation errors caused by system instability, and making the entire estimation method more scientific and reliable.

[0126] (IV) Dynamically adjust the observation matrix and estimate the state

[0127] Since the open circuit voltage and There is a nonlinear relationship between the observation matrix The elements in Dynamically adjust to changes. The specific methods are:

[0128] Pre-established and A lookup table (e.g., 20 points uniformly distributed from 0 to 1 and their corresponding value);

[0129] At each time step , according to current estimates , calculate the corresponding value by linear interpolation Slope and intercept ; where the observer estimates the current , there is an initial To start the estimation, each step K will observe a new ;

[0130] Update observation matrix .

[0131] In this embodiment, the observation matrix is ​​updated using a lookup table and linear interpolation, effectively reducing the impact of the nonlinear relationship between open-circuit voltage and SOC, as well as voltage measurement deviations, on the estimation results. This allows the observation matrix to be dynamically adjusted based on the real-time battery state, improving the observer's ability to respond to changes in battery state.

[0132] Then, using the measured voltage and the observer equation, iteratively calculating the state estimate , and the excitation current is estimated based on the updated observation matrix using the following formula:

[0133] (31)

[0134] The estimated current Substitute into formula (13) to complete Estimates.

[0135] This embodiment iteratively solves the observer equation to estimate the excitation current and inputs it into the battery state space, continuously optimizing the SOC estimate. This iterative calculation method can track changes in the battery state in real time, promptly correct estimation errors, avoid error accumulation, and achieve accurate SOC estimation, meeting the device's need for precise battery charge monitoring.

[0136] This embodiment addresses the problem that traditional methods rely on accurate voltage measurement and current input, and in practice, measurement deviation and the lack of current sensors lead to reduced accuracy. This embodiment constructs a state-space model based on the lithium battery equivalent circuit model and terminal voltage principle, taking into account more practical factors. The excitation current is used as an unknown input to define the observer equation, eliminating excessive reliance on current sensors and reducing hardware costs. By constructing an error dynamic equation and verifying stability, the error is effectively controlled. The observation matrix is ​​dynamically adjusted to reduce the impact of the nonlinear relationship between open-circuit voltage and SOC and voltage deviation, significantly improving the accuracy of SOC estimation.

[0137] This invention overcomes the limitations of traditional methods by employing a unique unknown input observer architecture to innovatively address measurement bias and the challenges of no current data. It incorporates the Lyapunov method for stability analysis and dynamically adjusts the observation matrix using open-circuit voltage and an SOC lookup table. The combined application of these methods provides a new and effective approach for accurately estimating SOC.

[0138] Example 2

[0139] This embodiment provides a lithium-ion battery SOC estimation system based on an unknown input observer, including:

[0140] State modeling module, used to build a battery state space model based on the equivalent circuit model and terminal voltage principle of lithium batteries;

[0141] An observer design module is used to define an observer equation using the excitation current as an unknown input, substitute the battery state space model and the observer equation into the observer error equation, and construct an error dynamic equation;

[0142] a gain stabilization module for simplifying the error dynamics equation based on observer noise immunity and stability constraints and a gain matrix to obtain a final error dynamics equation and verifying stability by constructing a Lyapunov function;

[0143] A table lookup and tuning module is used to obtain open circuit voltage information based on the current state of charge according to a lookup table of open circuit voltage and state of charge, and to update the observation matrix in the battery state space model;

[0144] The estimation module is used to iteratively solve the observer equation to estimate the excitation current based on the updated observation matrix and the measured voltage, input the excitation current into the battery state space, and complete the estimation of the state of charge.

[0145] Example 3

[0146] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps of the lithium-ion battery SOC estimation method based on an unknown input observer as described in the first embodiment above are implemented.

[0147] Example 4

[0148] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the method for estimating the SOC of a lithium-ion battery based on an unknown input observer as described in the first embodiment above are implemented.

[0149] The steps or modules involved in Examples 2 to 4 above correspond to those in Example 1. For detailed implementations, please refer to the relevant description of Example 1. The term "computer-readable storage medium" should be understood to mean a single medium or multiple media that includes one or more instruction sets; it should also be understood to include any medium that can store, encode, or carry an instruction set for execution by a processor and cause the processor to perform any method of the present invention.

[0150] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A lithium-ion battery SOC estimation method based on an unknown input observer, characterized in that: include: Based on the equivalent circuit model and terminal voltage principle of lithium batteries, a battery state space model is constructed, which specifically includes: A first-order equivalent circuit model is used, with the state of charge, polarization voltage and fixed measurement deviation as state variables, and an expression for the terminal voltage is constructed. Determine the output equation in the state-space model by defining the observer output and substituting it into the terminal voltage expression; Construct a battery state space model based on the output equation and the dynamic expressions of each state variable; An observer equation is defined by taking the excitation current as an unknown input, and the battery state space model and the observer equation are substituted into the observer error equation to construct an error dynamic equation; The observer equation is: ； in, is the intermediate state variable of the observer, is the estimated state vector; 、 、 、 Design the matrix for the observer; is the estimated excitation current, which is treated as an unknown input; represents the state variable vector of the battery state estimation system at time k, 、 Represent the measured voltages at time k and k+1 respectively; The error dynamic equation is: ； in, represents the observer error, represents the state vector, represents the process noise vector, represents the measurement noise vector, A represents the state transfer matrix, B represents the input matrix, C represents the observation matrix, D represents the internal resistance, F represents the process noise distribution matrix, G represents the measurement noise distribution matrix, and I represents the identity matrix; Based on the observer noise immunity and stability constraints and the gain matrix, the error dynamic equation is simplified to obtain the final error dynamic equation, and the stability is verified by constructing a Lyapunov function; the observer noise immunity and stability constraints include making the observer independent of process noise, measurement noise, and unknown inputs; According to the lookup table of open circuit voltage and state of charge, the open circuit voltage information is obtained based on the current state of charge, and the observation matrix in the battery state space model is updated; Based on the updated observation matrix and the measured voltage, the observer equation is iteratively solved to estimate the excitation current, which is then input into the battery state space to complete the state of charge estimation.

2. The lithium-ion battery SOC estimation method based on an unknown input observer according to claim 1, characterized in that: The Lyapunov function is constructed using a symmetric positive definite matrix, and its stability is determined by analyzing how the function changes over time. Specifically: When there is no measurement noise, the corresponding error dynamics are obtained. If the observer dynamic matrix satisfies Schur stability after adjusting the gain matrix, the battery state estimation system is globally exponentially stable in the absence of disturbances. When measurement noise exists, a disturbance term is defined to obtain the error dynamics in the presence of disturbance. The Lyapunov function difference is calculated, and an attenuation condition is introduced. By adjusting the gain matrix and the symmetric positive definite matrix, the influence of the disturbance on the error is controlled to ensure stability in the presence of disturbance.

3. The lithium-ion battery SOC estimation method based on an unknown input observer according to claim 1, characterized in that: The method of obtaining open circuit voltage information based on the current state of charge according to the lookup table of open circuit voltage and state of charge and updating the observation matrix in the battery state space model specifically includes: pre-constructing a lookup table of open circuit voltage and state of charge according to historical data, calculating the current open circuit voltage information corresponding to the current state of charge through linear interpolation based on the lookup table, and updating the observation matrix.

4. The lithium-ion battery SOC estimation method based on an unknown input observer according to claim 1, characterized in that: The method of iteratively solving the observer equation to estimate the excitation current based on the updated observation matrix and the measured voltage, inputting the excitation current into the battery state space, and completing the state of charge estimation specifically includes: Based on the measured voltage and the observer equation, the state estimate is iteratively calculated, and the current excitation current is estimated based on the updated observation matrix; The current excitation current is input into the observer equation to complete the estimation of the state of charge.

5. A lithium-ion battery SOC estimation system based on an unknown input observer, characterized in that: include: The state modeling module is used to build a battery state space model based on the equivalent circuit model and terminal voltage principle of the lithium battery, specifically including: A first-order equivalent circuit model is used, with the state of charge, polarization voltage and fixed measurement deviation as state variables, and an expression for the terminal voltage is constructed. Determine the output equation in the state-space model by defining the observer output and substituting it into the terminal voltage expression; Construct a battery state space model based on the output equation and the dynamic expressions of each state variable; An observer design module is used to define an observer equation using the excitation current as an unknown input, substitute the battery state space model and the observer equation into the observer error equation, and construct an error dynamic equation; The observer equation is: ； in, is the intermediate state variable of the observer, is the estimated state vector; 、 、 、 Design the matrix for the observer; is the estimated excitation current, which is treated as an unknown input; represents the state variable vector of the battery state estimation system at time k, 、 Represent the measured voltages at time k and k+1 respectively; The error dynamic equation is: ； in, represents the observer error, represents the state vector, represents the process noise vector, represents the measurement noise vector, A represents the state transfer matrix, B represents the input matrix, C represents the observation matrix, D represents the internal resistance, F represents the process noise distribution matrix, G represents the measurement noise distribution matrix, and I represents the identity matrix; a gain stabilization module for simplifying the error dynamics equation based on observer noise immunity and stability constraints and a gain matrix to obtain a final error dynamics equation and verifying stability by constructing a Lyapunov function; the observer noise immunity and stability constraints include making the observer independent of process noise, measurement noise, and unknown inputs; A table lookup and tuning module is used to obtain open circuit voltage information based on the current state of charge according to a lookup table of open circuit voltage and state of charge, and to update the observation matrix in the battery state space model; The estimation module is used to iteratively solve the observer equation to estimate the excitation current based on the updated observation matrix and the measured voltage, input the excitation current into the battery state space, and complete the estimation of the state of charge.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the lithium-ion battery SOC estimation method based on an unknown input observer are implemented as described in any one of claims 1 to 4.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the lithium-ion battery SOC estimation method based on an unknown input observer are implemented as described in any one of claims 1 to 4.

Images