Anti-noise lithium battery charge state prediction method, device, equipment and medium

Abstract

The invention discloses an anti-noise lithium battery charge state prediction method and device, equipment and a medium, and relates to the technical field of battery detection, and the method comprises the steps: updating the integer order capacitance of a polarization unit in a preset equivalent circuit model into a fractional order capacitance, and constructing a fractional order battery model; obtaining time sequence measurement data of the target lithium battery based on a charging and discharging experiment of the target lithium battery, and fitting the open-circuit voltage-SOC value curve by using static parameters of the time sequence measurement data to obtain a polynomial fitting function; performing initial parameter identification on the fractional-order battery model by using a preset optimization algorithm based on the measured voltage and the measured current at different sampling moments and a polynomial fitting function to obtain an initial fractional-order battery model, and performing secondary dynamic parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model; and the FOMHPF-EKF detection algorithm is used to detect the state of charge of the target fractional order battery model. And accurate state-of-charge prediction is realized.

Description

Technical Field

[0001] The present invention relates to the field of battery detection technology, and in particular to a noise-resistant lithium battery state of charge prediction method, device, equipment and medium. Background Art

[0002] In practical applications, lithium batteries exhibit typical characteristics such as nonlinearity, time-varying behavior, complex influencing factors, and uncertainty, influenced by various factors such as temperature fluctuations, aging, and charge-discharge cycles. This makes SOC (State of Charge) estimation difficult, inaccurate, and lacking adaptability. In the study of SOC estimation models with noise suppression capabilities, the Kalman filter (KF) demonstrates good robustness against Gaussian-distributed system noise. The derived extended Kalman filter (EKF) enables state estimation for nonlinear systems. However, linearization can introduce significant errors in highly nonlinear systems. The unscented Kalman filter (UKF) improves state estimation for nonlinear systems to a certain extent by selecting a specific set of sampling points (unscented transformation). To address the issue of accurate estimation for nonlinear systems, filters such as the Adaptive Extended Kalman Filter (AEKF) and the Adaptive Cubic Kalman Filter (ACKF) have been introduced. However, these methods only offer good suppression capabilities against Gaussian-distributed system noise, failing to address the noise-type interference requirements of real-world applications. Furthermore, they have not been validated in actual embedded hardware. Furthermore, existing algorithms often require expensive hardware resources to acquire current and voltage data, significantly increasing costs and making them unsuitable for widespread use. Therefore, designing a method that can be efficiently adapted to low-cost embedded devices is crucial for achieving rapid portable SOC detection. Summary of the Invention

[0003] In view of this, the present invention aims to provide a noise-resistant lithium battery state-of-charge prediction method, device, equipment, and medium that can accurately predict the state of charge of lithium batteries with different noise types and is suitable for low-cost embedded devices. The specific solution is as follows:

[0004] In a first aspect, the present application discloses a noise-resistant lithium battery state of charge prediction method, comprising:

[0005] Updating the integer-order capacitance of the polarization unit in the preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model including at least two polarization units;

[0006] Obtaining time series measurement data of the target lithium battery based on a charge and discharge experiment of the target lithium battery, and fitting an open circuit voltage-SOC value curve using static parameters of the time series measurement data to obtain a corresponding polynomial fitting function; wherein the time series measurement data includes the open circuit voltage, the corresponding SOC value, and the measured voltage and measured current at different sampling times;

[0007] Using a preset optimization algorithm, and based on the measured voltage and measured current at the different sampling moments and the polynomial fitting function, performing initial model parameter identification on the fractional-order battery model to obtain an initial fractional-order battery model, and performing secondary dynamic model parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model;

[0008] The FOMHPF-EKF detection algorithm is used to detect the state of charge of the target fractional-order battery model.

[0009] Optionally, the preset equivalent circuit model includes an open-circuit voltage source, an ohmic internal resistance, and at least two groups of polarization units connected in parallel, each polarization unit including an integer-order resistor and an integer-order capacitor;

[0010] Correspondingly, the fractional-order capacitor includes an open-circuit voltage source, an ohmic internal resistor, and at least two groups of new polarization units connected in parallel, and each new polarization unit includes an integer-order resistor and a fractional-order capacitor connected in parallel.

[0011] Optionally, constructing a fractional-order battery model of the target lithium battery includes:

[0012] Constructing a system equation of state variables at a current moment based on the first parameter matrix, the second parameter matrix, the system state at a previous moment, the system noise, and the input current at a previous moment, wherein the state variables include the SOC value and the voltage value of the integer-order capacitor;

[0013] Constructing a measurement equation for the terminal voltage at the current moment according to the system state at the current moment, the input current at the current moment, the ohmic internal resistance, the observation matrix, and the observation noise;

[0014] A fractional-order battery model of the target lithium battery is constructed based on the system equation and the measurement equation.

[0015] Optionally, the preset optimization algorithm is a whale optimization algorithm.

[0016] Accordingly, the method of using a preset optimization algorithm and performing initial model parameter identification on the fractional-order battery model based on the measured voltage and measured current at different sampling times and the polynomial fitting function to obtain an initial fractional-order battery model includes:

[0017] Initially and randomly generating a set of candidate initial model parameters as an initial population; wherein the initial population is a population including multiple sets of candidate initial model parameters;

[0018] Predicting a predicted voltage at each sampling moment using the fractional-order battery model and based on parameter values ​​of each group of candidate initial model parameters in the initial population;

[0019] Calculating the sum of squares of the differences between the predicted voltage and the measured voltage at the corresponding sampling moment as fitness information, updating the parameter values ​​of the candidate initial model parameters in the population based on the fitness information and the polynomial fitting function, obtaining an updated initial population, and jumping to the step of predicting the predicted voltage at each sampling moment using the fractional-order battery model and based on the parameter values ​​of each group of candidate initial model parameters in the initial population;

[0020] The parameter values ​​of the candidate initial model parameters that meet the preset iteration end condition are used as the initial model parameters; wherein the initial model parameters include ohmic internal resistance, fractional order internal resistance and fractional order parameters.

[0021] Optionally, updating the parameter values ​​of the candidate initial model parameters in the population based on the fitness information and the polynomial fitting function to obtain an updated initial population includes:

[0022] Using the whale optimization algorithm and adjusting the control coefficient based on the polynomial fitting function to switch the search method of the model parameter identification process;

[0023] The parameter values ​​of the candidate initial model parameters in the population are updated based on the current search mode and the fitness information to obtain an updated initial population.

[0024] Optionally, the search method includes a first search method and a second search method, wherein:

[0025] When the search coefficient is less than the preset coefficient threshold, the shrinking and surrounding strategy is set to update the position in the first search mode;

[0026] When the search coefficient is greater than or equal to the preset coefficient threshold, the global search strategy is set to update the position in the second search method.

[0027] Optionally, the detecting the state of charge of the target fractional-order battery model by using the FOMHPF-EKF detection algorithm includes:

[0028] Set the initial value of state estimation and error covariance based on prior knowledge;

[0029] According to the state estimation initial value, the real-time input current and the system noise, and by predicting the predicted value of the current state vector through the target fractional-order battery model, then determining and generating an initial particle set based on the predicted value and the error covariance;

[0030] Calculating the residual between the measured voltage and the predicted value, dynamically adjusting the weight of each particle in the initial particle set according to the residual, and performing normalization processing to obtain the target particle weight corresponding to each particle;

[0031] Calculating the effective particle number based on the distribution of the target particle weights, and performing resampling when the effective particle number is lower than a preset threshold until the effective particle number meets the preset threshold condition;

[0032] An average value is taken for each effective particle, and the average value result is output as the state of charge.

[0033] In a second aspect, the present application discloses a noise-resistant lithium battery state of charge prediction device, comprising:

[0034] A model creation module, configured to update the integer-order capacitance of the polarization unit in a preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model including at least two polarization units;

[0035] a curve fitting module for acquiring time series measurement data of a target lithium battery based on a charge and discharge experiment of the target lithium battery, and fitting an open circuit voltage-SOC value curve using static parameters of the time series measurement data to obtain a corresponding polynomial fitting function; wherein the time series measurement data includes the open circuit voltage, the corresponding SOC value, and the measured voltage and measured current at different sampling times;

[0036] a parameter identification module, configured to perform initial model parameter identification on the fractional-order battery model using a preset optimization algorithm and based on the measured voltage and measured current at different sampling times and the polynomial fitting function to obtain an initial fractional-order battery model, and perform secondary dynamic model parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model;

[0037] The state prediction module is used to detect the state of charge of the target fractional-order battery model using the FOMHPF-EKF detection algorithm.

[0038] In a third aspect, the present application discloses an electronic device, comprising:

[0039] Memory, used to store computer programs;

[0040] The processor is used to execute the computer program to implement the steps of the aforementioned noise-resistant lithium battery state of charge prediction method disclosed above.

[0041] In a fourth aspect, the present application discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the steps of the aforementioned noise-resistant lithium battery state of charge prediction method are implemented.

[0042] It can be seen that the present application discloses a noise-resistant lithium battery state of charge prediction method, including: updating the integer-order capacitance of the polarization unit in the preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model including at least two polarization units; obtaining the time series measurement data of the target lithium battery based on the charge and discharge experiment of the target lithium battery, and fitting the open circuit voltage-SOC value curve using the static parameters of the time series measurement data to obtain a corresponding polynomial fitting function; wherein the time series measurement data includes the open circuit voltage, the corresponding SOC value, the measured voltage and the measured current at different sampling times; using a preset optimization algorithm, and based on the measured voltage and the measured current at the different sampling times and the polynomial fitting function, performing initial model parameter identification on the fractional-order battery model to obtain an initial fractional-order battery model, and performing secondary dynamic model parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model; using the FOMHPF-EKF detection algorithm to detect the state of charge of the target fractional-order battery model. It can be seen that by constructing a fractional-order battery model, the relaxation process of lithium batteries under noise interference can be more accurately described. By fitting the OCV-SOC curve with a high-order polynomial, the static mapping error is eliminated, and the voltage-SOC mapping accuracy in the full SOC range is improved. Furthermore, through two-stage parameter identification, the model parameter robustness of the fractional-order battery model is improved, the parameter sensitivity is reduced, and the model can also operate effectively in an inherent noise platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0044] Figure 1 This is a flow chart of a noise-resistant lithium battery state of charge prediction method disclosed in this application;

[0045] Figure 2 This is a flow chart of a model parameter identification method disclosed in this application;

[0046] Figure 3 This is an error diagram of the terminal voltage predicted by the FOMHPF-EKF algorithm disclosed in this application under two working conditions; wherein, Figure 3 (a) is the error diagram of the terminal voltage predicted by the FOMHPF-EKF algorithm under the condition of uniformly distributed random noise current. Figure 3 (b) Error diagram of the terminal voltage predicted by the FOMHPF-EKF algorithm under the condition of zero-mean random noise current;

[0047] Figure 4 The battery SOC estimation results and estimation error diagrams of three algorithms disclosed in this application under NEDC and DST conditions are shown; wherein, Figure 4 (a) is the battery SOC estimation result diagram of FOMHPF, UKF, and PF algorithms under NEDC conditions; Figure 4 (b) is the battery SOC estimation error diagram of FOMHPF, UKF, and PF algorithms under NEDC conditions; Figure 4 (c) is the battery SOC estimation result diagram of FOMHPF, UKF, and PF algorithms under DST conditions; Figure 4 (d) is the battery SOC estimation error diagram of FOMHPF, UKF, and PF algorithms under DST conditions;

[0048] Figure 5 This is a structural diagram of a noise-resistant lithium battery state of charge prediction and verification system disclosed in this application;

[0049] Figure 6 Schematic diagram of two artificially added random noise currents disclosed in this application;

[0050] Figure 7 The comparison diagram of random noise on SOC prediction results and SOC prediction error under NEDC conditions disclosed in this application; wherein, Figure 7 (a) is a comparison chart of the SOC prediction results of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and uniformly distributed random noise current. Figure 7 (b) is a comparison chart of the SOC prediction results of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and random noise current with zero mean. Figure 7 (c) Comparison of SOC prediction results of FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and system noise; Figure 7(d) is a comparison chart of the SOC prediction errors of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and uniformly distributed random noise current. Figure 7 (e) is a comparison chart of the SOC prediction errors of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and random noise current with zero mean. Figure 7 (f) Comparison of SOC prediction errors of FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and system noise;

[0051] Figure 8 The comparison diagram of the impact of random noise on SOC prediction results and the comparison diagram of SOC prediction error under DST conditions disclosed in this application are as follows; Figure 8 (a) is a comparison chart of the SOC prediction results of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and uniformly distributed random noise current. Figure 8 (b) is a comparison chart of the SOC prediction results of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and random noise current with zero mean. Figure 8 (c) Comparison of SOC prediction results of FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST condition and system noise; Figure 8 (d) is a comparison chart of the SOC prediction errors of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and uniformly distributed random noise current. Figure 8 (e) is a comparison chart of the SOC prediction errors of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and random noise current with zero mean. Figure 8 (f) Comparison of SOC prediction errors of FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and system noise;

[0052] Figure 9 This is a schematic structural diagram of a noise-resistant lithium battery state of charge prediction device disclosed in this application;

[0053] Figure 10 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0055] With the growing global demand for renewable energy, batteries are increasingly being used as energy storage devices in electric vehicles, renewable energy generation systems, and energy storage systems. Battery management systems (BMS) play a vital role in ensuring battery safety, improving system reliability, and extending battery life. One of the core functions of a BMS is to accurately estimate the battery's state of charge (SOC), a key parameter for measuring the battery's current remaining capacity. However, due to the complexity of the battery's internal electrochemical reactions, changes in operating conditions, and the impact of battery aging, accurately estimating SOC is an extremely challenging task. In this context, the introduction of fractional-order modeling provides a new approach to accurately modeling SOC, accurately describing the battery's internal state, and thus optimizing the overall performance of the battery management system.

[0056] In practical applications, lithium-ion batteries exhibit typical characteristics such as nonlinearity, time-varying behavior, complexity of influencing factors, and uncertainty, influenced by various factors such as temperature fluctuations, aging, and charge-discharge cycles. This makes state-of-charge (SOC) estimation difficult, inaccurate, and insufficiently adaptable. Research on SOC estimation models with noise suppression capabilities has shown that the Kalman filter (KF) demonstrates robustness against Gaussian-distributed system noise. The extended Kalman filter (EKF), derived from this, achieves state estimation for nonlinear systems. However, linearization can introduce significant errors in highly nonlinear systems. The unscented Kalman filter (UKF) improves state estimation for nonlinear systems to a certain extent by selecting a specific set of sampling points (unscented transformation). To address accurate estimation of nonlinear systems, filters such as the AEKF and ACKF have been introduced. However, these methods only offer good suppression capabilities for Gaussian-distributed system noise and cannot meet the requirements of real-world noise interference. Furthermore, they have not been validated in actual embedded hardware. Furthermore, existing algorithms often require expensive hardware resources to acquire current and voltage data, significantly increasing costs and making them unsuitable for widespread use.

[0057] To this end, the present invention provides a noise-resistant lithium battery state-of-charge prediction solution, which can accurately predict the state of charge of lithium batteries with different noise types and is suitable for low-cost embedded devices.

[0058] Reference Figure 1As shown, the present invention provides a noise-resistant lithium battery state of charge prediction method, comprising:

[0059] Step S11: updating the integer-order capacitance of the polarization unit in the preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model including at least two polarization units.

[0060] In this embodiment, the preset equivalent circuit model includes an open circuit voltage source, an ohmic internal resistance, and at least two groups of polarization units connected in parallel, each polarization unit includes an integer-order resistor and an integer-order capacitor; correspondingly, the fractional-order capacitor includes an open circuit voltage source, an ohmic internal resistance, and at least two groups of new polarization units connected in parallel, each new polarization unit includes an integer-order resistor and a fractional-order capacitor connected in parallel. It can be understood that a fractional-order second-order equivalent circuit model is established based on the second-order RC equivalent physical model of the target lithium battery, that is, a fractional-order battery model is updated; specifically, the preset equivalent circuit model includes an open circuit voltage source, an ohmic internal resistance, and at least two groups of new polarization units connected in parallel, each new polarization unit includes an integer-order resistor and a fractional-order capacitor connected in parallel. , terminal voltage , ohmic internal resistance , Polarization internal resistance (integer resistance) and , polarized capacitance (integer order capacitance) and Polarization internal resistance With polarized capacitance Parallel, polarization internal resistance With polarized capacitance In parallel, the open circuit voltage is Ohmic internal resistance , polarization internal resistance after parallel connection With polarized capacitance , polarization internal resistance after parallel connection With polarized capacitance Then, based on the second-order RC equivalent physical model of the lithium battery, fractional-order operators α and β are introduced to obtain the fractional-order battery model.

[0061] In this embodiment, a state-space discrete system equation (system equation) and a measurement equation are obtained. Specifically, a system equation of the state variables at the current moment is constructed based on the first parameter matrix, the second parameter matrix, the system state at the previous moment, the system noise, and the input current at the previous moment. The state variables include the SOC value and the voltage value of the integer-order capacitor; a measurement equation of the terminal voltage at the current moment is constructed based on the system state at the current moment, the input current at the current moment, the ohmic internal resistance, the observation matrix, and the observation noise; and a fractional-order battery model of the target lithium battery is constructed based on the system equation and the measurement equation. It can be understood that the mathematical expressions of the state-space discrete system equation and the measurement equation are as follows:

[0062] ;

[0063] ;

[0064] ;

[0065] in, represents the state vector, and are the system process noise and observation noise respectively. is the variable at the current moment, 、 、 are parameter matrices, specifically the state transfer matrix, input matrix, and observation matrix, for Current flowing through the ohmic internal resistance The current value, Indicates the rated capacity of the battery. for Estimated value of the moment, polarization capacitance The order is α, the polarized capacitance The order of is β, Represents the open circuit voltage compensation term.

[0066] Specifically, the present invention selects the second-order RC equivalent model as the lithium battery equivalent circuit model in terms of the fractional-order model, including the open circuit voltage , terminal voltage , ohmic internal resistance , polarization internal resistance and , polarized capacitors and Polarization internal resistance With polarized capacitance Parallel, polarization internal resistance With polarized capacitance In parallel, the open circuit voltage is and the battery internal resistance , polarization internal resistance after parallel connection With polarized capacitance , polarization internal resistance after parallel connection With polarized capacitance The circuit equation of the second-order RC equivalent model is:

[0067] ;

[0068] in, is the load current. and is the polarization voltage corresponding to the RC network.

[0069] According to the definition of the fractional ladder equation Gel'fand-Levitan (GL), the discrete form of fractional calculus can be expressed as follows:

[0070] ;

[0071] in, Representation function The α-order derivative at point t, where α is a non-negative real number. is the coefficient of the generalized Newton binomial, The state value at a historical moment, representing the polarization voltage of the battery at a past moment (e.g., th, t-2h), is defined as:

[0072] ;

[0073] when =0, Among them, α determines the type of fractional calculus equation. When the value of α is greater than 0, the equation is in differential form; when the value of α is less than 0, the equation is in integral form:

[0074] ;

[0075] 、 The capacitor model in the frequency domain is established based on the Westerlund theory to represent the fractional characteristics. The function being operated is, Represents the dynamic change function of the state parameters (polarization voltage, SOC, etc.) of the corresponding lithium battery over time; Represents the weight coefficient of fractional calculus, which is related to the order a and index j. and The fractional capacitance is expressed as:

[0076] ;

[0077] Where α and β are fractional orders of polarization capacitance, their values ​​are decimals, j is an imaginary number, ω is the angular frequency, 、 Polarized capacitance and impedance.

[0078] The state space equation of the fractional second-order equivalent circuit model is:

[0079] ;

[0080] ;

[0081] Where α is the polarization capacitance The order of β is the polarization capacitance The order of Polarization internal resistance The voltage on Polarization internal resistance The voltage on the MOSFET is U, and the terminal voltage is U.

[0082] SOC is defined as the percentage of remaining capacity to nominal discharge capacity, and the calculation formula is:

[0083] ;

[0084] in, For the next moment, Battery SOC at the moment, represents the sampling interval (s), represents the load current at the kth sampling moment, represents Columbia efficiency, Indicates the battery capacity.

[0085] Combined formula and , establish the discretization equation:

[0086] ;

[0087] in, represents the filter gain matrix;

[0088] ;

[0089] ;

[0090] ;

[0091] Assuming the true value With the state vector The relationship is:

[0092] ;

[0093] Add boundary condition definition matrix To rearrange the equation, the filtering formula is as follows:

[0094] ;

[0095] in, and are the state transfer matrix and the observation matrix, is the gain matrix, is the error matrix, Represents the state error covariance matrix, which reflects the degree of deviation between the state estimate and the true value. represents the identity matrix, represents the weighting coefficient, represents the weighting matrix, represents the model parameter vector.

[0096] Step S12: Obtaining timing measurement data of the target lithium battery based on a charge and discharge experiment of the target lithium battery, and fitting an open circuit voltage-SOC value curve using static parameters of the timing measurement data to obtain a corresponding polynomial fitting function; wherein the timing measurement data includes the open circuit voltage, the corresponding SOC value, the measured voltage and the measured current at different sampling times.

[0097] In this embodiment, a high-pressure charge-discharge (HPPC) test was performed on the target lithium battery to obtain time-series measurement data of the target lithium battery. An OCV-SOC curve was fitted based on the open-circuit voltage and corresponding SOC values ​​in the time-series measurement data obtained from the battery charge-discharge test. The OCV-SOC curve was fitted to obtain a corresponding 8th-order polynomial fitting function. Table 1 lists the battery-related parameters of the target lithium battery.

[0098] Table 1 Battery parameters of target lithium battery

[0099]

[0100] Specifically, the battery was charged in constant current and constant voltage mode, with constant current and constant voltage values ​​of 2A and 4.2V, respectively. It was then discharged at a constant current of 6.5A for 3 minutes. The battery was then left to rest for 3 hours to reach electrochemical equilibrium. The constant current and constant voltage, constant current discharge, and resting steps were repeated until the cutoff voltage was reached. A polynomial fit was performed on the experimental data to produce an OCV-SOC fitting curve. Analysis of the results of fittings of different orders revealed that the 8th-order fit had higher accuracy and the resulting curve was closer to the actual result. The function expression for the 8th-order fit is as follows:

[0101] ;

[0102] Step S13: Using a preset optimization algorithm, and based on the measured voltage and measured current at the different sampling times and the polynomial fitting function, the initial model parameter identification of the fractional-order battery model is performed to obtain an initial fractional-order battery model, and the secondary dynamic model parameter identification of the initial fractional-order battery model is performed to obtain a target fractional-order battery model.

[0103] In this embodiment, the preset optimization algorithm is a whale optimization algorithm. Accordingly, the preset optimization algorithm is used, and the initial model parameter identification of the fractional-order battery model is performed based on the measured voltage and measured current at different sampling moments and the polynomial fitting function to obtain an initial fractional-order battery model, including: initially and randomly generating a group of candidate initial model parameters as an initial population; wherein the initial population is a population containing multiple groups of candidate initial model parameters; using the fractional-order battery model and based on the parameter value of each group of candidate initial model parameters in the initial population, predicting the predicted voltage at each sampling moment; calculating The sum of the squares of the differences between the predicted voltage and the measured voltage at the corresponding sampling moment is used as fitness information. The parameter values ​​of the candidate initial model parameters in the population are updated based on the fitness information and the polynomial fitting function to obtain an updated initial population. The process then jumps to the step of predicting the predicted voltage at each sampling moment using the fractional-order battery model based on the parameter values ​​of each group of candidate initial model parameters in the initial population. The parameter values ​​of the candidate initial model parameters that meet a preset iteration termination condition are used as the initial model parameters. The initial model parameters include ohmic internal resistance, fractional-order internal resistance, and fractional-order parameters. The whale optimization algorithm is used to adjust the control coefficient based on the polynomial fitting function to switch the search mode of the model parameter identification process. The parameter values ​​of the candidate initial model parameters in the population are updated based on the current search mode and the fitness information to obtain an updated initial population. When the search coefficient is less than a preset coefficient threshold, the shrinking and surrounding strategy position update mode is set to the first search mode. When the search coefficient is greater than or equal to the preset coefficient threshold, the global search strategy position update mode is set to the second search mode.

[0104] It is understandable that the whale algorithm is used to perform offline parameter identification on the model to obtain the parameters of the fractional-order model, wherein the parameters optimized in the initial model parameter identification specifically include the ohmic internal resistance , polarization internal resistance 、 , polarized capacitance 、 , fractional-order operators α and β, and their results are used as the initial model parameters. The whale algorithm is used to optimize the initial parameters of the fractional-order lithium battery model (including ohmic internal resistance, polarization internal resistance, polarization capacitance, and fractional-order parameters). Based on the fractional-order lithium battery model and the functional relationship between open-circuit voltage and SOC, the algorithm switches between the shrinking and random search phases by adjusting the control coefficients a and c to adapt to the nonlinear optimization process of lithium battery parameters. Specifically, the steps for offline parameter identification of the model are as follows:

[0105] Reference Figure 2 As shown, the whale population is initialized, and a set of model parameter solutions (including the initial parameters of ohmic internal resistance, polarization internal resistance, polarization capacitance, and fractional-order parameters) are randomly generated as the initial population;

[0106] Calculate the fitness value of each individual in the current population;

[0107] Based on the current individual parameter values, the predicted voltage at each sampling time point is calculated using the fractional-order battery model;

[0108] The sum of the squares of the difference between the predicted voltage and the actual measured voltage is used as the fitness function value, and the normalized fitness is calculated;

[0109] Based on the normalized fitness and the coefficients a and c, the individual positions of the population are updated:

[0110] When |a|<1, update the position according to the shrinking and surrounding strategy;

[0111] When |a|≥1, update the position according to the global search strategy;

[0112] Iterative execution steps: Calculate the fitness value of each individual in the current population to step: Based on the normalized fitness and according to the coefficients a and c, update the position of the individual in the population until the termination condition is met, and output the model parameters corresponding to the optimal individual as the initial model parameters, that is, obtain the initial fractional-order battery model after parameter identification.

[0113] After offline parameter identification and updating, the initial fractional-order battery model is obtained. The initial model parameters are then used as the initial values ​​for the online parameter identification operation. The EKF is used to perform online parameter identification and update, and the target fractional-order battery model after secondary dynamic parameter identification is obtained. Table 2 shows the parameter results corresponding to the various model parameters of the target fractional-order battery model after parameter identification.

[0114] Table 2 Model parameters of the target fractional-order battery model

[0115]

[0116] The relevant formulas involved in the above whale optimization algorithm are as follows:

[0117] When |a|<1, the algorithm mainly performs shrinkage and encirclement operations: ;in, Represents the candidate parameter combination at time t, specifically the parameter set of the lithium battery model;

[0118] When |a|≥1, the algorithm randomly selects a global search: ;in, represents a randomly selected candidate parameter combination.

[0119] The fitness function is measured by the difference between the predicted voltage and the actual voltage, which can be defined as:

[0120] ;

[0121] in, is the voltage at the i-th time point through the fractional-order model, is the actual measured voltage at the i-th time point. T represents the number of sampling time points.

[0122] Fitness can be expressed as:

[0123] ;

[0124] in, represents the maximum value of the fitness function, Represents the fitness value at the i-th time point.

[0125] Step S14: Utilizing the FOMHPF-EKF detection algorithm to detect the state of charge of the target fractional-order battery model.

[0126] In this embodiment, the initial state estimation value and error covariance are set based on prior knowledge; the predicted value of the current state vector is predicted based on the initial state estimation value, real-time input current and system noise, and through the target fractional-order battery model, and then the initial particle set is determined and generated based on the predicted value and the error covariance; the residual between the measured voltage and the predicted value is calculated, and the weight of each particle in the initial particle set is dynamically adjusted according to the residual, and normalized to obtain the target particle weight corresponding to each particle; the number of effective particles is calculated based on the distribution of the target particle weight, and resampling is performed when the effective particle number is lower than the preset threshold until the effective particle number meets the preset threshold condition; the average value of each effective particle is taken to output the average value result as the state of charge. It can be understood that the target fractional-order lithium battery model after parameter identification adopts the fractional-order H∞ filter and particle filter (PF) fusion algorithm and combines it with the EKF online update, that is, the FOMHPF-EKF algorithm to detect and analyze SOC;

[0127] Estimating battery state of charge using the FOMHPF-EKF algorithm:

[0128] k∈{1,...,N}, where N represents the total sampling time of the system.

[0129] The cost function of HPF is:

[0130] ;

[0131] in, and Represent the true value and estimated value at time k respectively. Reflects the initial state value With the initial estimate Prior information on the degree of closeness. is a user-defined weighting matrix.

[0132] Initialization: Set the initial state and state error covariance.

[0133] ;

[0134] in, and are the initial parameter estimates and the error covariance matrix, and are the initial state estimates and the error covariance matrix.

[0135] Prior estimates:

[0136] ;

[0137] Filter gain:

[0138] ;

[0139] in, represents the prior error covariance matrix of particles i at time k, reflecting the deviation between the predicted state and the actual state, represents the observation noise covariance matrix at time k, represents the weighting coefficient at time k, Represents the user-defined weight matrix at time k.

[0140] Posterior estimate:

[0141] ;

[0142] in, represents the internal resistance of the battery at time k, represents the observation error of the i-th particle at time k, Represents the measured terminal voltage (measured value) at time k.

[0143] Error covariance update:

[0144] ;

[0145] in, represents the system process noise covariance matrix at time k.

[0146] Normalize particle weights:

[0147] ;

[0148] Calculating the threshold :

[0149] ;

[0150] Perform residual resampling ( < ). Calculate the average value and get SOC:

[0151] ;

[0152] like Figure 3 As shown in Figure 2, the error diagram of the terminal voltage predicted by the FOMHPF-EKF algorithm under two working conditions (uniformly distributed random noise current and random noise current with zero mean), where: Figure 3 (a) is the error diagram of the terminal voltage predicted by the FOMHPF-EKF algorithm under the condition of uniformly distributed random noise current. Figure 3 (b) is the error diagram of the terminal voltage predicted by the FOMHPF-EKF algorithm under the condition of random noise current with zero mean; further, Figure 4 (a) and Figure 4 (c) shows the battery SOC estimation results of FOMHPF, UKF, and PF algorithms under NEDC and DST conditions respectively. Figure 4 (b) and Figure 4 (d) shows the error in battery SOC estimation for the FOHIPF, UKF, and PF algorithms under NEDC and DST conditions. The FOMHPF algorithm achieves an RMSE of 0.22% and a Mean Absolute Error (MAE) of 0.018% under NEDC conditions, while its RMSE and MAE under DST conditions are 0.5% and 0.12%. FOHIPF demonstrates high accuracy and robustness in SOC estimation.

[0153] In this embodiment, after predicting the state of charge of the target lithium battery, the predicted state of charge is further verified for noise immunity. The verification method is to build a dual-source noise system to verify its noise immunity. Specifically, refer to Figure 5 As shown, the present application discloses a lithium battery embedded management system based on fractional-order fusion filtering, which is used to execute a noise-resistant lithium battery state of charge prediction method and noise resistance verification steps, wherein the management system includes two platforms, platform one: a battery testing instrument; platform two: using an STM32F407ZG main control chip, platform one is connected to a host computer PC via CAN communication, and platform two is connected to a host computer PC via a USB serial port. It should be noted that, in addition to executing the steps of the above-mentioned anti-noise lithium battery state of charge prediction method, the current management system also has the function of verifying the anti-noise performance of this solution. Specifically, Platform 2 is connected to the PC via a USB serial port to collect the measured voltage of the low-cost equipment system and the real-time input current information carrying system noise, and then input the above-mentioned measured information with system noise into the host PC. The host PC runs the FOMHPF-EKF algorithm proposed in this solution in real time to estimate the SOC, directly realizing the above-mentioned steps of constructing the target fractional-order battery model and outputting the anti-noise SOC prediction value. The steps for anti-noise verification are as follows: three types of random noise currents are selected, namely, the uniformly distributed random noise current added artificially, the random noise current with zero mean (such as Figure 6 As shown in the figure, and the inherent noise current of platform 2, under the influence of the above noise signal, the corresponding noise-resistant SOC prediction value will be output through platform 2. At the same time, the benchmark SOC value will be provided by platform 1. By comparing the SOC value results of the two, its noise-resistant performance is verified, which also proves that this scheme can accurately predict the SOC value under noise signals on low-cost equipment.

[0154] like Figure 7 and Figure 8 As shown, Figure 7 and Figure 8 The influence of random noise on SOC prediction results under NEDC and DST conditions is demonstrated. The SOC estimation accuracy of the FOMHPF-EKF algorithm developed under random noise current is verified. Figure 7 (a) is a comparison chart of the SOC prediction results of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and uniformly distributed random noise current. Figure 7 (b) is a comparison chart of the SOC prediction results of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and random noise current with zero mean. Figure 7 (c) Comparison of SOC prediction results of FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and system noise; Figure 7 (d) is a comparison chart of the SOC prediction errors of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and uniformly distributed random noise current. Figure 7 (e) is a comparison chart of the SOC prediction errors of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and random noise current with zero mean. Figure 7 (f) Comparison of SOC prediction errors of FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under NEDC conditions and system noise; Figure 8 (a) is a comparison chart of the SOC prediction results of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and uniformly distributed random noise current. Figure 8 (b) is a comparison chart of the SOC prediction results of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and random noise current with zero mean. Figure 8 (c) Comparison of SOC prediction results of FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST condition and system noise; Figure 8 (d) is a comparison chart of the SOC prediction errors of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and uniformly distributed random noise current. Figure 8 (e) is a comparison chart of the SOC prediction errors of the FOMHPF-EKF algorithm, UKF algorithm, and PF algorithm under DST conditions and random noise current with zero mean. Figure 8 (f) Comparison of the SOC prediction errors of the FOMHPF-EKF algorithm, the UKF algorithm, and the PF algorithm under DST conditions and system noise. It can be observed that under the influence of three different types of random current noise, the UKF and PF algorithms experience increasing errors and begin to fail to converge. However, the FOMHPF-EKF algorithm consistently achieves excellent SOC predictions, with the maximum SOC error being less than 0.01%. This demonstrates the high prediction accuracy of the FOMHP-EKF algorithm even under uncertain current disturbances.

[0155] It can be seen that the present application discloses a noise-resistant lithium battery state of charge prediction method, including: updating the integer-order capacitance of the polarization unit in the preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model including at least two polarization units; obtaining the time series measurement data of the target lithium battery based on the charge and discharge experiment of the target lithium battery, and fitting the open circuit voltage-SOC value curve using the static parameters of the time series measurement data to obtain a corresponding polynomial fitting function; wherein the time series measurement data includes the open circuit voltage, the corresponding SOC value, the measured voltage and the measured current at different sampling times; using a preset optimization algorithm, and based on the measured voltage and the measured current at the different sampling times and the polynomial fitting function, performing initial model parameter identification on the fractional-order battery model to obtain an initial fractional-order battery model, and performing secondary dynamic model parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model; using the FOMHPF-EKF detection algorithm to detect the state of charge of the target fractional-order battery model. It can be seen that by constructing a fractional-order battery model, the relaxation process of lithium batteries under noise interference can be more accurately described. By fitting the OCV-SOC curve with a high-order polynomial, the static mapping error is eliminated, and the voltage-SOC mapping accuracy in the full SOC range is improved. Furthermore, through two-stage parameter identification, the model parameter robustness of the fractional-order battery model is improved, the parameter sensitivity is reduced, and the model can also operate effectively in an inherent noise platform.

[0156] Reference Figure 9 As shown, the present invention also discloses a noise-resistant lithium battery state of charge prediction device, comprising:

[0157] A model creation module 11 is used to update the integer-order capacitance of the polarization unit in the preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model including at least two polarization units;

[0158] a curve fitting module 12 for acquiring time series measurement data of a target lithium battery based on a charge and discharge experiment of the target lithium battery, and fitting an open circuit voltage-SOC value curve using static parameters of the time series measurement data to obtain a corresponding polynomial fitting function; wherein the time series measurement data includes the open circuit voltage, the corresponding SOC value, and the measured voltage and measured current at different sampling times;

[0159] a parameter identification module 13, configured to perform initial model parameter identification on the fractional-order battery model using a preset optimization algorithm and based on the measured voltage and measured current at different sampling times and the polynomial fitting function to obtain an initial fractional-order battery model, and perform secondary dynamic model parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model;

[0160] The state prediction module 14 is configured to detect the state of charge of the target fractional-order battery model using a FOMHPF-EKF detection algorithm.

[0161] It can be seen that the present application discloses updating the integer-order capacitance of the polarization unit in the preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model comprising at least two polarization units; obtaining the timing measurement data of the target lithium battery based on the charge and discharge experiment of the target lithium battery, and fitting the open circuit voltage-SOC value curve using the static parameters of the timing measurement data to obtain a corresponding polynomial fitting function; wherein the timing measurement data includes the open circuit voltage, the corresponding SOC value, the measured voltage and the measured current at different sampling times; using a preset optimization algorithm, and based on the measured voltage and measured current at the different sampling times and the polynomial fitting function, performing initial model parameter identification on the fractional-order battery model to obtain an initial fractional-order battery model, and performing secondary dynamic model parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model; using the FOMHPF-EKF detection algorithm to detect the state of charge of the target fractional-order battery model. It can be seen that by constructing a fractional-order battery model, the relaxation process of lithium batteries under noise interference can be more accurately described. By fitting the OCV-SOC curve with a high-order polynomial, the static mapping error is eliminated, and the voltage-SOC mapping accuracy in the full SOC range is improved. Furthermore, through two-stage parameter identification, the model parameter robustness of the fractional-order battery model is improved, the parameter sensitivity is reduced, and the model can also operate effectively in an inherent noise platform.

[0162] Furthermore, the embodiment of the present application also discloses an electronic device, Figure 10 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content in the diagram should not be considered as any limitation to the scope of application of the present application.

[0163] Figure 10This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps of the noise-resistant lithium battery state of charge prediction method disclosed in any of the aforementioned embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0164] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device. The communication protocol it follows is any communication protocol that can be applied to the technical solution of this application and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world. Its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0165] Among them, the processor 21 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 21 can be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 21 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 21 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 21 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0166] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or CD, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0167] The operating system 221 is used to manage and control the hardware devices and computer program 222 on the electronic device 20, so as to enable the processor 21 to calculate and process the massive amount of data 223 in the memory 22. The operating system 221 can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of implementing the noise-resistant lithium battery state of charge prediction method performed by the electronic device 20 as disclosed in any of the aforementioned embodiments, the computer program 222 can further include computer programs capable of performing other specific tasks. In addition to data received by the electronic device from an external device, the data 223 can also include data collected by its own input and output interface 25.

[0168] Furthermore, this application discloses a computer-readable storage medium for storing a computer program; wherein, when executed by a processor, the computer program implements the aforementioned noise-resistant lithium battery state-of-charge prediction method. The specific steps of this method can be referred to the corresponding contents disclosed in the aforementioned embodiments and will not be repeated here.

[0169] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. Reference can be made to the descriptions of the identical or similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and the relevant parts can be referred to the descriptions of the methods.

[0170] Professionals may further appreciate that the units and algorithmic steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application. The steps of the method or algorithm described in conjunction with the embodiments disclosed herein can be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module can be placed in a random access memory RAM (Random Access Memory), memory, read-only memory ROM (Read Only Memory), electrically programmable EPROM (Electrically Programmable Read Only Memory), electrically erasable programmable EEPROM (Electric Erasable Programmable Read Only Memory), registers, hard disk, removable disk, CD-ROM (Compact Disc-Read Only Memory), or any other form of storage medium known in the technical field.

[0171] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0172] The above is a detailed introduction to the solution provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core ideas. At the same time, for those skilled in the art, according to the ideas of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A noise-resistant lithium battery state of charge prediction method, characterized in that: include: Updating the integer-order capacitance of the polarization unit in the preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model including at least two polarization units; Obtaining time series measurement data of the target lithium battery based on a charge and discharge experiment of the target lithium battery, and fitting an open circuit voltage-SOC value curve using static parameters of the time series measurement data to obtain a corresponding polynomial fitting function; wherein the time series measurement data includes the open circuit voltage, the corresponding SOC value, and the measured voltage and measured current at different sampling times; Using a preset optimization algorithm, and based on the measured voltage and measured current at the different sampling moments and the polynomial fitting function, performing initial model parameter identification on the fractional-order battery model to obtain an initial fractional-order battery model, and performing secondary dynamic model parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model; The FOMHPF-EKF detection algorithm is used to detect the state of charge of the target fractional-order battery model.

2. The noise-resistant lithium battery state of charge prediction method according to claim 1, characterized in that: The preset equivalent circuit model includes an open-circuit voltage source, an ohmic internal resistance, and at least two groups of polarization units connected in parallel, each polarization unit including an integer-order resistor and an integer-order capacitor; Correspondingly, the fractional-order capacitor includes an open-circuit voltage source, an ohmic internal resistor, and at least two groups of new polarization units connected in parallel, and each new polarization unit includes an integer-order resistor and a fractional-order capacitor connected in parallel.

3. The noise-resistant lithium battery state of charge prediction method according to claim 2, characterized in that: The constructing of a fractional-order battery model of the target lithium battery includes: Constructing a system equation of state variables at a current moment based on the first parameter matrix, the second parameter matrix, the system state at a previous moment, the system noise, and the input current at a previous moment, wherein the state variables include the SOC value and the voltage value of the integer-order capacitor; Constructing a measurement equation for the terminal voltage at the current moment according to the system state at the current moment, the input current at the current moment, the ohmic internal resistance, the observation matrix, and the observation noise; A fractional-order battery model of the target lithium battery is constructed based on the system equation and the measurement equation.

4. The noise-resistant lithium battery state of charge prediction method according to claim 1, characterized in that: The preset optimization algorithm is the whale optimization algorithm. Accordingly, the method of using a preset optimization algorithm and performing initial model parameter identification on the fractional-order battery model based on the measured voltage and measured current at different sampling times and the polynomial fitting function to obtain an initial fractional-order battery model includes: Initially and randomly generating a set of candidate initial model parameters as an initial population; wherein the initial population is a population including multiple sets of candidate initial model parameters; Predicting a predicted voltage at each sampling moment using the fractional-order battery model and based on parameter values ​​of each group of candidate initial model parameters in the initial population; Calculating the sum of squares of the differences between the predicted voltage and the measured voltage at the corresponding sampling moment as fitness information, updating the parameter values ​​of the candidate initial model parameters in the population based on the fitness information and the polynomial fitting function, obtaining an updated initial population, and jumping to the step of predicting the predicted voltage at each sampling moment using the fractional-order battery model and based on the parameter values ​​of each group of candidate initial model parameters in the initial population; The parameter values ​​of the candidate initial model parameters that meet the preset iteration end condition are used as the initial model parameters; wherein the initial model parameters include ohmic internal resistance, fractional order internal resistance and fractional order parameters.

5. The noise-resistant lithium battery state of charge prediction method according to claim 4, characterized in that: The updating of the parameter values ​​of the candidate initial model parameters in the population based on the fitness information and the polynomial fitting function to obtain an updated initial population includes: Using the whale optimization algorithm and adjusting the control coefficient based on the polynomial fitting function to switch the search method of the model parameter identification process; The parameter values ​​of the candidate initial model parameters in the population are updated based on the current search mode and the fitness information to obtain an updated initial population.

6. The noise-resistant lithium battery state of charge prediction method according to claim 5, characterized in that: The search method includes a first search method and a second search method, wherein: When the search coefficient is less than the preset coefficient threshold, the shrinking and surrounding strategy is set to update the position in the first search mode; When the search coefficient is greater than or equal to the preset coefficient threshold, the global search strategy is set to update the position in the second search method.

7. The noise-resistant lithium battery state of charge prediction method according to any one of claims 1 to 6, characterized in that: The detecting of the state of charge of the target fractional-order battery model by using the FOMHPF-EKF detection algorithm includes: Set the initial value of state estimation and error covariance based on prior knowledge; According to the state estimation initial value, the real-time input current and the system noise, and by predicting the predicted value of the current state vector through the target fractional-order battery model, then determining and generating an initial particle set based on the predicted value and the error covariance; Calculating the residual between the measured voltage and the predicted value, dynamically adjusting the weight of each particle in the initial particle set according to the residual, and performing normalization processing to obtain the target particle weight corresponding to each particle; Calculating the effective particle number based on the distribution of the target particle weights, and performing resampling when the effective particle number is lower than a preset threshold until the effective particle number meets the preset threshold condition; An average value is taken for each effective particle, and the average value result is output as the state of charge.

8. A noise-resistant lithium battery state of charge prediction device, characterized in that: include: A model creation module, configured to update the integer-order capacitance of the polarization unit in a preset equivalent circuit model of the target lithium battery to a fractional-order capacitance to construct a fractional-order battery model of the target lithium battery; wherein the preset equivalent circuit model is an equivalent physical model including at least two polarization units; a curve fitting module for acquiring time series measurement data of a target lithium battery based on a charge and discharge experiment of the target lithium battery, and fitting an open circuit voltage-SOC value curve using static parameters of the time series measurement data to obtain a corresponding polynomial fitting function; wherein the time series measurement data includes the open circuit voltage, the corresponding SOC value, and the measured voltage and measured current at different sampling times; a parameter identification module, configured to perform initial model parameter identification on the fractional-order battery model using a preset optimization algorithm and based on the measured voltage and measured current at different sampling times and the polynomial fitting function to obtain an initial fractional-order battery model, and perform secondary dynamic model parameter identification on the initial fractional-order battery model to obtain a target fractional-order battery model; The state prediction module is used to detect the state of charge of the target fractional-order battery model using the FOMHPF-EKF detection algorithm.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor is used to execute the computer program to implement the steps of the noise-resistant lithium battery state of charge prediction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that Used to store a computer program; wherein, when the computer program is executed by a processor, the steps of the noise-resistant lithium battery state of charge prediction method according to any one of claims 1 to 7 are implemented.

Images