Electric vehicle battery SOC prediction method and system and computer readable storage medium

Abstract

The invention provides an electric vehicle battery SOC prediction method and system and a computer readable storage medium, and the method comprises the steps: firstly, building a second-order equivalent circuit model of a battery, and combining an extended Kalman filtering algorithm to convert SOC estimation into prediction and correction of a state space; the process noise covariance matrix and the measurement noise covariance matrix are optimized in an off-line mode through the Harris eagle optimization algorithm, the optimized matrix value is applied to the extended Kalman filtering algorithm in the SOC prediction process, and accurate prediction of the SOC is achieved. The method can guarantee the prediction stability and accuracy under the dynamic load and various discharge conditions, facilitates the early warning of the electric vehicle before the electric quantity is exhausted, and improves the user experience.

Description

Technical Field

[0001] The present invention belongs to the technical field of power batteries, and in particular relates to a method, system and computer-readable storage medium for predicting the SOC of an electric vehicle battery. Background Art

[0002] Lithium-ion batteries are widely used in new energy vehicles due to their advantages such as long cycle life, high specific energy, and low self-discharge rate. As power batteries, they have discharge voltage characteristics, so it is necessary to estimate the driving range of electric vehicles. Among them, the estimation of the driving range of electric vehicles is mainly based on the estimation of the SOC (State of Charge) of the power battery.

[0003] In related technologies, SOC calculations can be performed using a vehicle's onboard BMS (Battery Management System), using methods such as open-circuit voltage, ampere-hour integration, and neural networks. For example, patent application CN110673039A (A Big Data-Based Online Correction Method for SOC Charging of Lithium Iron Phosphate Batteries) utilizes the one-to-one correspondence between the peak position of the dQ / dV value and the SOC to estimate the SOC. By establishing a BP (Back Propagation) neural network model, neural network calculations are performed based on the peak value of the battery capacity differential curve dQ / dv, combined with battery parameters such as temperature, voltage, and SOH (State of Health). This allows for correction of the SOC during vehicle charging, eliminating the cumulative error caused by shallow charging and discharging. Patent application CN113625175A (A Cloud-Based Big Data Platform-Based SOC Estimation Method and System) achieves accurate SOC estimation and correction through data cleaning and the establishment of different SOC estimation modules under different operating conditions. Furthermore, weights are assigned based on the feasibility of the estimation algorithm in different regions to achieve accurate SOC estimation and correction. This method can adjust the algorithm's estimation weights and estimation modules at any time.

[0004] However, the methods in the related art are that the former requires a large amount of data for training of the neural network model, and the vehicle data upload cycle is long and the charging process is in a variable current state, so the calculation accuracy of dQ / dv is low, which increases the difficulty of training the neural network model; the latter estimation idea still follows the method of the vehicle-side BMS (Battery Management System), and the quality of cloud data is worse than that of the vehicle-side, which may cause the problem of reduced estimation accuracy. In addition, this method simplifies the vehicle-side algorithm, but requires frequent communication between the cloud and the vehicle, and has requirements for timeliness of communication. If there is a problem with the communication or the data is delayed, it may cause the algorithm to deviate. In summary, in the related art, when making SOC predictions, vehicle storage is limited, the model used for the estimation result is short-lived and has low reliability; and the BMS software algorithm is rarely changed after the vehicle leaves the factory, resulting in an increase in the SOC estimation deviation during operation, affecting the user experience. Summary of the Invention

[0005] In view of the deficiencies in the prior art, the present invention provides a method, system and computer-readable storage medium for predicting the SOC of an electric vehicle battery.

[0006] In a first aspect, the present invention provides a method for predicting the SOC of an electric vehicle battery, comprising:

[0007] A second-order equivalent circuit model is used to construct a battery circuit model to determine the relationship between the battery's open circuit voltage and SOC;

[0008] Construct the state transfer equation and observation equation according to the battery's SOC and polarization voltage;

[0009] Initialize the estimated value and error covariance matrix of the battery SOC;

[0010] Construct process noise covariance matrix and measurement noise covariance matrix;

[0011] The extended Kalman filter algorithm is used to predict the battery SOC based on the state transfer equation, the observation equation, the initial estimated value of the battery SOC, the initial error covariance matrix, the process noise covariance matrix, and the measurement noise covariance matrix to obtain a first SOC estimate of the battery;

[0012] The Harris Eagle optimization algorithm is used to optimize the process noise covariance matrix and the measurement noise covariance matrix according to the first SOC estimation value;

[0013] Determine a priori estimated value of the battery SOC according to the state transition equation as a first state vector;

[0014] Determine the prior estimate of the error covariance matrix based on the state transfer equation and the optimized process noise covariance matrix;

[0015] Determining the Kalman gain based on the optimized measurement noise covariance matrix and the prior estimation of the error covariance matrix;

[0016] Update the first state vector according to the observed value of the battery SOC and the first SOC estimated value to obtain a second state vector;

[0017] Update the error covariance matrix based on the Kalman gain, the observation equation, and the prior estimate of the error covariance matrix;

[0018] The extended Kalman filter algorithm is used to predict the battery SOC based on the optimized process noise covariance matrix and measurement noise covariance matrix, the second state vector and the updated error covariance matrix to obtain the second SOC estimate of the battery, which is used as the final battery SOC prediction value.

[0019] Optionally, constructing a state transfer equation and an observation equation based on the SOC and polarization voltage of the battery includes:

[0020] Construct the expression of the state transition equation:

[0021]

[0022] Among them, SOC k′ is the battery state of charge at time k'; V p1,k′ is the first polarization voltage at time k'; V p2,k′ is the second polarization voltage at time k'; e is a natural constant; t is the sampling interval; R p1 is the resistance of the first resistor-capacitor pair in the second-order equivalent circuit model; C p1 is the capacitance of the first resistor-capacitor pair in the second-order equivalent circuit model; R p2 C is the resistance of the second resistor-capacitor pair in the second-order equivalent circuit model; p2 is the capacitance value of the capacitor in the second resistor-capacitor pair of the second-order equivalent circuit model; SOC k'-1 is the battery state of charge at time k'-1; V p1,k'-1 is the first polarization voltage at time k'-1; V p2,k'-1 is the second polarization voltage at time k'-1; Q u is the effective capacity of the battery; I d is the discharge current of the battery;

[0023] Construct an expression for the observation equation:

[0024]

[0025] Among them, V dis the output voltage of the second-order equivalent circuit model; OCV is the open circuit voltage of the battery; Indicates the rate of change of OCV with SOC; R s is the resistance of the series resistor in the second-order equivalent circuit model.

[0026] Optionally, the adopting the Harris Eagle optimization algorithm and optimizing the process noise covariance matrix and the measurement noise covariance matrix according to the first SOC estimation value includes:

[0027] The escape energy E is calculated according to the following formula:

[0028]

[0029] Among them, E initial is the initial escape energy; t' is the current iteration number of the Harris Hawk optimization algorithm; t' max The maximum number of iterations for the Harris Hawk optimization algorithm; f best is the global optimal fitness value found currently; f current is the fitness value of the current individual;

[0030] Construct the expression of Levy flight distance LF(D):

[0031]

[0032] Where u is a random variable drawn from a normal distribution; μ is a random variable drawn from a Cauchy distribution; β is a parameter of the Levy distribution used to control the shape of the distribution; α is a weight factor that controls the step size;

[0033] The Harris Eagle optimization algorithm updates the process noise covariance matrix and the measurement noise covariance matrix in each iteration by minimizing the mean square error (MSE) according to the following formula:

[0034]

[0035] Where N is the number of samples; SOC estimated (n) is the first SOC estimate of the nth iteration; SOC ture (n) is the true value of SOC at the nth iteration;

[0036] After multiple iterations, the process noise covariance matrix and measurement noise covariance matrix corresponding to the minimum MSE are taken as the final process noise covariance matrix and measurement noise covariance matrix.

[0037] Optionally, determining a priori estimated value of the battery SOC according to the state transfer equation as the first state vector includes:

[0038] The prior estimate of the battery SOC is calculated according to the following formula:

[0039]

[0040] Where A is the state transfer matrix used to describe the time variation of SOC and polarization voltage; is the a posteriori estimate of the battery SOC at time k'-1; B is the control matrix used to reflect the influence of current on the state; I d is the discharge current of the battery.

[0041] Optionally, determining a priori estimation of the error covariance matrix based on the state transition equation and the optimized process noise covariance matrix includes:

[0042] The prior estimate P of the error covariance matrix is calculated according to the following formula - :

[0043] P - =A·P k'-1 ·A T +Q;

[0044] Where A is the state transfer matrix used to describe the time variation of SOC and polarization voltage; P k'-1 is the posterior estimate of the error covariance matrix at time k'-1; T represents the transpose of the matrix; Q is the optimized process noise covariance moment.

[0045] Optionally, determining the Kalman gain based on the optimized prior estimates of the measurement noise covariance matrix and the error covariance matrix includes:

[0046] The Kalman gain K is calculated according to the following formula:

[0047] K=P - ·H T (H.P - ·H T +R) -1 ;

[0048] Among them, P - is the prior estimate of the error covariance matrix; H is the observation matrix; T represents the transpose of the matrix; R is the optimized measurement noise covariance matrix.

[0049] Optionally, updating the first state vector according to the observed value of the battery SOC and the first SOC estimated value to obtain the second state vector includes:

[0050] Update the first state vector according to the following formula:

[0051]

[0052] in, is the second state vector; is the prior estimate of battery SOC; K is the Kalman gain; Z is the observed value of battery SOC; and H is the observation matrix.

[0053] Optionally, updating the error covariance matrix according to the Kalman gain, the observation equation, and the prior estimate of the error covariance matrix includes:

[0054] Update the error covariance matrix according to the following formula:

[0055] P=(I-KH)P - ;

[0056] Where P is the updated error covariance matrix; I is the identity matrix; K is the Kalman gain; P - is the prior estimate of the error covariance matrix.

[0057] In a second aspect, the present invention provides an electric vehicle battery SOC prediction system, comprising:

[0058] The first building module is used to build a circuit model of the battery using a second-order equivalent circuit model to determine the relationship between the open circuit voltage and SOC of the battery;

[0059] The second building module is used to construct a state transfer equation and an observation equation according to the SOC and polarization voltage of the battery;

[0060] Initialization module, used to initialize the estimated value of battery SOC and error covariance matrix;

[0061] The third building module is used to build the process noise covariance matrix and the measurement noise covariance matrix;

[0062] A first SOC prediction module is configured to use an extended Kalman filter algorithm to predict the SOC of the battery based on a state transfer equation, an observation equation, an initial estimated value of the battery SOC, an initial error covariance matrix, a process noise covariance matrix, and a measurement noise covariance matrix to obtain a first SOC estimated value of the battery;

[0063] an optimization module for optimizing a process noise covariance matrix and a measurement noise covariance matrix based on a first SOC estimate using a Harris Eagle optimization algorithm;

[0064] a first determining module, configured to determine a priori estimated value of the battery SOC according to a state transition equation as a first state vector;

[0065] a second determination module, configured to determine a priori estimates of an error covariance matrix based on a state transfer equation and an optimized process noise covariance matrix;

[0066] a third determination module, configured to determine the Kalman gain based on the optimized measurement noise covariance matrix and the prior estimation of the error covariance matrix;

[0067] A first updating module is configured to update the first state vector according to the observed value of the battery SOC and the first SOC estimated value to obtain a second state vector;

[0068] A second updating module is used to update the error covariance matrix according to the Kalman gain, the observation equation and the prior estimate of the error covariance matrix;

[0069] The second SOC prediction module is used to use the extended Kalman filter algorithm to predict the battery SOC based on the optimized process noise covariance matrix and measurement noise covariance matrix, the second state vector and the updated error covariance matrix, to obtain the second SOC estimation value of the battery, which is used as the final battery SOC prediction value.

[0070] In a third aspect, the present invention provides a computer-readable storage medium for storing a computer program; when the computer program is executed by a processor, the steps of the electric vehicle battery SOC prediction method described in the first aspect are implemented.

[0071] The present invention provides a method, system and computer-readable storage medium for predicting the SOC of an electric vehicle battery. The method predicts the SOC based on the Harris Eagle optimization algorithm and the extended Kalman filter algorithm. First, by establishing a second-order equivalent circuit model of the battery, the SOC estimation is converted into a prediction and correction problem in the state space in combination with the extended Kalman filter algorithm. The process noise covariance matrix and the measurement noise covariance matrix are optimized offline by the Harris Eagle optimization algorithm, and the optimized matrix values are applied to the extended Kalman filter algorithm in the SOC prediction process to achieve accurate prediction of the SOC. The present invention can ensure the stability and accuracy of the prediction under dynamic loads and various discharge conditions, which helps electric vehicles to warn before the battery is exhausted and improve the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. 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 these drawings without creative work.

[0073] Figure 1 A flow chart of a method for predicting SOC of an electric vehicle battery provided by an embodiment of the present invention;

[0074] Figure 2 A schematic diagram of the structure of an electric vehicle battery SOC prediction system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0075] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0076] Example 1

[0077] like Figure 1 As shown, an embodiment of the present invention provides a method for predicting the SOC of an electric vehicle battery, comprising:

[0078] Step 101: construct a circuit model of the battery using a second-order equivalent circuit model to determine the relationship between the open circuit voltage and the SOC of the battery.

[0079] In this step, a second-order equivalent circuit model is used to describe the dynamic behavior of the battery under different discharge states. The model includes the battery's open circuit voltage OCV, series resistance R s and two parallel resistor-capacitor pairs (R p1 , C p1 and R p2 , C p2 ), the output voltage V of the second-order equivalent circuit model d Expressed as:

[0080] V d =OCV-V p1 -V p2 -R s I d .

[0081] Using OCV and polarization voltage V p1 and V p2 Describe the dynamic behavior of the battery, the polarization voltage is represented by an exponential decay function to simulate the polarization effect of the battery, the polarization voltage V p1 and V p2 Expressed as:

[0082]

[0083] Among them, Q u is the effective capacity of the battery; I d It is the discharge current of the battery, which indicates the current intensity when the battery supplies power to the outside world.

[0084] Step 102: construct a state transfer equation and an observation equation based on the SOC and polarization voltage of the battery.

[0085] For example, the expression of the state transition equation is constructed:

[0086]

[0087] Among them, SOC k′ is the battery state of charge at time k'; V p1,k′ is the first polarization voltage at time k'; V p2,k′ is the second polarization voltage at time k'; e is a natural constant; t is the sampling interval; R p1 is the resistance of the first resistor-capacitor pair in the second-order equivalent circuit model; C p1 is the capacitance of the first resistor-capacitor pair in the second-order equivalent circuit model; R p2 C is the resistance of the second resistor-capacitor pair in the second-order equivalent circuit model; p2 is the capacitance value of the capacitor in the second resistor-capacitor pair of the second-order equivalent circuit model; SOC k'-1 is the battery state of charge at time k'-1; V p1,k'-1 is the first polarization voltage at time k'-1; V p2,k'-1 is the second polarization voltage at time k'-1; and Both are exponential decay terms, indicating the exponential decay behavior of the polarization voltage.

[0088] Construct an expression for the observation equation:

[0089]

[0090] Among them, V d is the output voltage of the second-order equivalent circuit model; OCV is the open circuit voltage of the battery; It represents the rate of change of open circuit voltage OCV with state of charge (SOC), that is, the partial derivative of OCV with respect to SOC; R s is the resistance of the series resistor in the second-order equivalent circuit model.

[0091] Step 103: Initialize the estimated value of the battery SOC and the error covariance matrix.

[0092] In this step, the error covariance matrix P0 after initialization is:

[0093]

[0094] Among them, σ SOC 、 and They represent the initial state estimation at battery SOC, first polarization voltage V p1 and the second polarization voltage V p2 uncertainty on the .

[0095] Step 104: construct a process noise covariance matrix and a measurement noise covariance matrix.

[0096] For example, the expressions of process noise covariance matrix Q and measurement noise covariance matrix R are:

[0097]

[0098] R=r meas .

[0099] Among them, q SOC 、 and Respectively represent the battery SOC, the first polarization voltage V p1 and the second polarization voltage V p2 Process noise on r meas represents the measurement noise.

[0100] Step 105 , using an extended Kalman filter algorithm and predicting the battery SOC based on the state transfer equation, the observation equation, the initial estimated value of the battery SOC, the initial error covariance matrix, the process noise covariance matrix, and the measurement noise covariance matrix, to obtain a first SOC estimate of the battery.

[0101] Step 106 : Using the Harris Eagle optimization algorithm, optimize the process noise covariance matrix and the measurement noise covariance matrix according to the first SOC estimation value.

[0102] In this step, we first apply an adaptive escape energy adjustment strategy to dynamically adjust the hawks' escape energy based on the iteration progress. Secondly, we introduce a Lévy flight strategy with dynamic weighting factors, making the algorithm more exploratory in the early stages and more exploitative in the later stages, thereby improving the accuracy and convergence speed of SOC estimation.

[0103] Adopting an adaptive escape energy adjustment strategy, the escape energy E reflects the exploration intention of the eagle group. By adaptively adjusting the escape energy, more flexibility is provided in switching between the exploration and exploitation phases. For example, the escape energy E is calculated according to the following formula:

[0104]

[0105] Among them, E initial is the initial escape energy; t' is the current iteration number of the Harris Hawk optimization algorithm; t' max The maximum number of iterations for the Harris Hawk optimization algorithm; f best is the global optimal fitness value found currently; f current is the fitness value of the current individual;

[0106] The Levy flight strategy with a dynamic weight factor is introduced to make the step size of the Levy flight adaptively adjusted during the iteration process, so that it has stronger exploration in the early stage and more stable convergence in the later stage. For example, the expression of the Levy flight distance LF(D) is constructed as follows:

[0107]

[0108] Wherein, u is a random variable drawn from a normal distribution; μ represents a random variable drawn from a Cauchy distribution; β is a parameter of the Levy distribution used to control the shape of the distribution. In this embodiment, 1<β≤2; α is a weight factor that controls the step size.

[0109] The Harris Eagle optimization algorithm updates the process noise covariance matrix and the measurement noise covariance matrix in each iteration by minimizing the mean square error (MSE) according to the following formula:

[0110]

[0111] Where N is the number of samples; SOC estimated (n) is the first SOC estimate of the nth iteration; SOC ture (n) is the true value of SOC at the nth iteration. n represents a natural number from 1 to N, used for counting.

[0112] After multiple iterations (set according to actual needs, that is, the number of times can be preset), the process noise covariance matrix and measurement noise covariance matrix corresponding to the minimum MSE are taken as the final process noise covariance matrix and measurement noise covariance matrix.

[0113] Step 107 : Determine a priori estimated value of the battery SOC according to the state transition equation to serve as a first state vector.

[0114] In this step, the SOC and polarization voltage are predicted a priori based on the state transfer equation.

[0115] For example, the prior estimate of the battery SOC is calculated according to the following formula:

[0116]

[0117] Where A is the state transfer matrix used to describe the time variation of SOC and polarization voltage; is the a posteriori estimate of the battery SOC at time k'-1; B is the control matrix used to reflect the influence of current on the state; I d is the discharge current of the battery.

[0118] Step 108 : determining a priori estimation of the error covariance matrix based on the state transfer equation and the optimized process noise covariance matrix.

[0119] The noise covariance matrix Q is used to calculate the prior estimate of the error covariance, which represents the uncertainty of the predicted state.

[0120] Exemplarily, the prior estimate P of the error covariance matrix is calculated according to the following formula - :

[0121] P - =A·P k'-1 ·A T +Q.

[0122] Where A is the state transfer matrix used to describe the time variation of SOC and polarization voltage; P k'-1 is the posterior estimate of the error covariance matrix at time k'-1; T represents the transpose of the matrix; Q is the optimized process noise covariance moment.

[0123] Step 109 : determining the Kalman gain based on the optimized measurement noise covariance matrix and the prior estimation of the error covariance matrix.

[0124] The Kalman gain is used to balance the weights between the predicted value and the observed value. For example, the Kalman gain K is calculated according to the following formula:

[0125] K=P - ·H T (H.P - ·H T +R) -1 .

[0126] Among them, P - is the prior estimate of the error covariance matrix; H is the observation matrix; T represents the transpose of the matrix; R is the optimized measurement noise covariance matrix.

[0127] Step 1010: Update the first state vector according to the observed value of the battery SOC and the first SOC estimated value to obtain a second state vector.

[0128] According to the observed value Z of battery SOC and the predicted observed value (i.e., the predicted value of SOC), correct the prior state estimate, and obtain the updated SOC estimate Exemplarily, the first state vector is updated according to the following formula:

[0129]

[0130] in, is the second state vector; is the prior estimate of battery SOC; K is the Kalman gain; H is the observation matrix.

[0131] Step 1011: Update the error covariance matrix based on the Kalman gain, the observation equation, and the prior estimate of the error covariance matrix.

[0132] By updating the error covariance matrix, the uncertainty of state estimation is reduced, providing a basis for the next prediction.

[0133] Exemplarily, the error covariance matrix is updated according to the following formula:

[0134] P=(I-KH)P - .

[0135] Where P is the updated error covariance matrix; I is the identity matrix; K is the Kalman gain; P - is the prior estimate of the error covariance matrix.

[0136] In step 1012, an extended Kalman filter algorithm is used to predict the battery SOC based on the optimized process noise covariance matrix and the measurement noise covariance matrix, the second state vector, and the updated error covariance matrix to obtain a second SOC estimate of the battery, which is used as the final battery SOC prediction value.

[0137] In summary, the electric vehicle battery SOC prediction method provided in this embodiment first establishes a second-order equivalent circuit model of the power battery and combines the extended Kalman filter algorithm (EKF) to convert the SOC estimation into a prediction and correction problem in the state space. The noise covariance matrices Q and R are optimized offline by the Harris Eagle optimization algorithm, and the optimized matrix values are applied to the EKF during the SOC prediction process to achieve accurate prediction of the SOC. The electric vehicle battery SOC prediction method provided in this embodiment reduces the estimation error caused by improper noise parameter settings in the traditional EKF algorithm, improves the accuracy of the SOC prediction, and can ensure the stability and accuracy of the SOC prediction under dynamic loads and various discharge conditions, which helps electric vehicles to be warned before the battery is exhausted, and provides an efficient SOC estimation solution for the battery management system of electric vehicles.

[0138] Example 2

[0139] Based on the same inventive concept as Example 1, this embodiment also provides an electric vehicle battery SOC prediction system. Since the principle of solving the problem by this system is similar to the aforementioned electric vehicle battery SOC prediction method, the implementation of this system can refer to the implementation of the electric vehicle battery SOC prediction method.

[0140] like Figure 2 As shown, the electric vehicle battery SOC prediction system includes:

[0141] The first building module 10 is used to build a circuit model of the battery using a second-order equivalent circuit model to determine the relationship between the open circuit voltage and the SOC of the battery.

[0142] The second building module 20 is used to build a state transfer equation and an observation equation according to the SOC and polarization voltage of the battery.

[0143] The initialization module 30 is used to initialize the estimated value of the battery SOC and the error covariance matrix.

[0144] The third building module 40 is used to build a process noise covariance matrix and a measurement noise covariance matrix.

[0145] The first SOC prediction module 50 is used to use the extended Kalman filter algorithm to predict the SOC of the battery based on the state transfer equation, the observation equation, the initial estimated value of the battery SOC, the initial error covariance matrix, the process noise covariance matrix and the measurement noise covariance matrix to obtain a first SOC estimated value of the battery.

[0146] The optimization module 60 is configured to optimize the process noise covariance matrix and the measurement noise covariance matrix according to the first SOC estimation value by using the Harris Eagle optimization algorithm.

[0147] The first determination module 70 is configured to determine a priori estimated value of the battery SOC according to the state transition equation as a first state vector.

[0148] The second determination module 80 is configured to determine a priori estimates of the error covariance matrix based on the state transfer equation and the optimized process noise covariance matrix.

[0149] The third determination module 90 is configured to determine the Kalman gain according to the optimized measurement noise covariance matrix and the prior estimation of the error covariance matrix.

[0150] The first updating module 100 is configured to update the first state vector according to the observed value of the battery SOC and the first SOC estimated value to obtain a second state vector.

[0151] The second updating module 110 is configured to update the error covariance matrix according to the Kalman gain, the observation equation and the prior estimation of the error covariance matrix.

[0152] The second SOC prediction module 120 is used to use the extended Kalman filter algorithm to predict the SOC of the battery based on the optimized process noise covariance matrix and the measurement noise covariance matrix, the second state vector and the updated error covariance matrix, to obtain a second SOC estimate of the battery as the final battery SOC prediction value.

[0153] For more specific working processes of the above modules, please refer to the corresponding content disclosed in Example 1, which will not be repeated here.

[0154] Example 3

[0155] This embodiment provides a computer device, including a processor and a memory; wherein, when the processor executes a computer program stored in the memory, the steps of the electric vehicle battery SOC prediction method described in Example 1 are implemented.

[0156] For more specific details about the above method, please refer to the corresponding content disclosed in Example 1, which will not be repeated here.

[0157] Example 4

[0158] This embodiment provides a computer-readable storage medium for storing a computer program; when the computer program is executed by a processor, the steps of the electric vehicle battery SOC prediction method described in Example 1 are implemented.

[0159] For more specific details about the above method, please refer to the corresponding content disclosed in Example 1, which will not be repeated here.

[0160] Example 5

[0161] This embodiment provides a computer program product, including computer executable instructions or a computer program. When the computer executable instructions or the computer program are executed by a processor, the steps of the electric vehicle battery SOC prediction method described in Example 1 are implemented.

[0162] For more specific details about the above method, please refer to the corresponding content disclosed in Example 1, which will not be repeated here.

[0163] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. References to the same or similar parts between the various embodiments will be sufficient. The systems, devices, storage media, and computer program products disclosed in the embodiments correspond to the methods disclosed in the embodiments, so their descriptions are relatively simplified. For relevant details, refer to the method descriptions.

[0164] Those skilled in the art will clearly understand that the techniques in the embodiments of the present invention can be implemented using software plus a necessary general-purpose hardware platform. Based on this understanding, the technical solutions in the embodiments of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention or certain portions of the embodiments.

[0165] In some embodiments, computer-executable instructions may be in the form of a program, software, software module, script, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0166] As an example, computer-executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinating files (e.g., files storing one or more modules, subroutines, or code portions).

[0167] By way of example, computer-executable instructions may be deployed to be executed on one electronic device, or on multiple electronic devices located at one site, or on multiple electronic devices distributed across multiple sites and interconnected by a communication network.

[0168] The present invention has been described in detail above with reference to specific embodiments and exemplary examples. However, these descriptions should not be construed as limiting the present invention. Those skilled in the art will appreciate that various equivalent substitutions, modifications, or improvements may be made to the technical solutions and implementations of the present invention without departing from the spirit and scope of the present invention, all of which fall within the scope of the present invention. The scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A method for predicting SOC of an electric vehicle battery, characterized in that: include: A second-order equivalent circuit model is used to construct a battery circuit model to determine the relationship between the battery's open circuit voltage and SOC; Construct the state transfer equation and observation equation according to the battery's SOC and polarization voltage; Initialize the estimated value and error covariance matrix of the battery SOC; Construct process noise covariance matrix and measurement noise covariance matrix; The extended Kalman filter algorithm is used to predict the battery SOC based on the state transfer equation, the observation equation, the initial estimated value of the battery SOC, the initial error covariance matrix, the process noise covariance matrix, and the measurement noise covariance matrix to obtain a first SOC estimate of the battery; The Harris Eagle optimization algorithm is used to optimize the process noise covariance matrix and the measurement noise covariance matrix according to the first SOC estimation value; Determine a priori estimated value of the battery SOC according to the state transition equation as a first state vector; Determine the prior estimate of the error covariance matrix based on the state transfer equation and the optimized process noise covariance matrix; Determining the Kalman gain based on the optimized measurement noise covariance matrix and the prior estimation of the error covariance matrix; Update the first state vector according to the observed value of the battery SOC and the first SOC estimated value to obtain a second state vector; Update the error covariance matrix based on the Kalman gain, the observation equation, and the prior estimate of the error covariance matrix; The extended Kalman filter algorithm is used to predict the battery SOC based on the optimized process noise covariance matrix and measurement noise covariance matrix, the second state vector and the updated error covariance matrix to obtain the second SOC estimate of the battery, which is used as the final battery SOC prediction value.

2. The electric vehicle battery SOC prediction method according to claim 1, characterized in that: The state transfer equation and observation equation are constructed according to the SOC and polarization voltage of the battery, including: Construct the expression of the state transition equation: Among them, SOC k′ is the battery state of charge at time k'; V p1,k′ is the first polarization voltage at time k'; V p2,k′ is the second polarization voltage at time k'; e is a natural constant; t is the sampling interval; R p1 is the resistance of the first resistor-capacitor pair in the second-order equivalent circuit model; C p1 is the capacitance of the first resistor-capacitor pair in the second-order equivalent circuit model; R p2 C is the resistance of the second resistor-capacitor pair in the second-order equivalent circuit model; p2 is the capacitance value of the capacitor in the second resistor-capacitor pair of the second-order equivalent circuit model; SOC k'-1 is the battery state of charge at time k'-1; V p1,k'-1 is the first polarization voltage at time k'-1; V p2,k'-1 is the second polarization voltage at time k'-1; Q u is the effective capacity of the battery; I d is the discharge current of the battery; Construct an expression for the observation equation: Among them, V d is the output voltage of the second-order equivalent circuit model; OCV is the open circuit voltage of the battery; Indicates the rate of change of OCV with SOC; R s is the resistance of the series resistor in the second-order equivalent circuit model.

3. The electric vehicle battery SOC prediction method according to claim 1, characterized in that: The method of using the Harris Eagle optimization algorithm and optimizing the process noise covariance matrix and the measurement noise covariance matrix according to the first SOC estimation value includes: The escape energy E is calculated according to the following formula: Among them, E initial is the initial escape energy; t' is the current iteration number of the Harris Hawk optimization algorithm; t' max The maximum number of iterations for the Harris Hawk optimization algorithm; f best is the global optimal fitness value found currently; f current is the fitness value of the current individual; Construct the expression of Levy flight distance LF(D): Where u is a random variable drawn from a normal distribution; μ is a random variable drawn from a Cauchy distribution; β is a parameter of the Levy distribution used to control the shape of the distribution; α is a weight factor that controls the step size; The Harris Eagle optimization algorithm updates the process noise covariance matrix and the measurement noise covariance matrix in each iteration by minimizing the mean square error (MSE) according to the following formula: Where N is the number of samples; SOC estimated (n) is the first SOC estimate of the nth iteration; SOC ture (n) is the true value of SOC at the nth iteration; After multiple iterations, the process noise covariance matrix and measurement noise covariance matrix corresponding to the minimum MSE are taken as the final process noise covariance matrix and measurement noise covariance matrix.

4. The electric vehicle battery SOC prediction method according to claim 1, characterized in that: The step of determining a priori estimated value of the battery SOC according to the state transition equation as a first state vector includes: The prior estimate of the battery SOC is calculated according to the following formula: Where A is the state transfer matrix used to describe the time variation of SOC and polarization voltage; is the a posteriori estimate of the battery SOC at time k'-1; B is the control matrix used to reflect the influence of current on the state; I d is the discharge current of the battery.

5. The electric vehicle battery SOC prediction method according to claim 1, characterized in that: The prior estimation of the error covariance matrix is determined based on the state transfer equation and the optimized process noise covariance matrix, including: The prior estimate P of the error covariance matrix is calculated according to the following formula - : P - =A·P k'-1 ·A T +Q; Where A is the state transfer matrix used to describe the time variation of SOC and polarization voltage; P k'-1 is the posterior estimate of the error covariance matrix at time k'-1; T represents the transpose of the matrix; Q is the optimized process noise covariance moment.

6. The electric vehicle battery SOC prediction method according to claim 1, characterized in that: The determining of the Kalman gain based on the optimized prior estimation of the measurement noise covariance matrix and the error covariance matrix includes: The Kalman gain K is calculated according to the following formula: K=P - ·H T (H·P - ·H T +R) -1 ; Among them, P - is the prior estimate of the error covariance matrix; H is the observation matrix; T represents the transpose of the matrix; R is the optimized measurement noise covariance matrix.

7. The electric vehicle battery SOC prediction method according to claim 1, characterized in that: The updating of the first state vector according to the observed value of the battery SOC and the first SOC estimated value to obtain the second state vector includes: Update the first state vector according to the following formula: in, is the second state vector; is the prior estimate of the battery SOC; K is the Kalman gain; Z is the observed value of the battery SOC; and H is the observation matrix.

8. The electric vehicle battery SOC prediction method according to claim 1, characterized in that: The updating of the error covariance matrix according to the Kalman gain, the observation equation and the prior estimation of the error covariance matrix comprises: Update the error covariance matrix according to the following formula: P=(I-KH)P - ; Where P is the updated error covariance matrix; I is the identity matrix; K is the Kalman gain; P - is the prior estimate of the error covariance matrix.

9. An electric vehicle battery SOC prediction system, characterized in that: include: The first building module is used to build a circuit model of the battery using a second-order equivalent circuit model to determine the relationship between the open circuit voltage and SOC of the battery; The second building module is used to construct a state transfer equation and an observation equation according to the SOC and polarization voltage of the battery; Initialization module, used to initialize the estimated value of battery SOC and error covariance matrix; The third building module is used to build the process noise covariance matrix and the measurement noise covariance matrix; A first SOC prediction module is configured to use an extended Kalman filter algorithm to predict the SOC of the battery based on a state transfer equation, an observation equation, an initial estimated value of the battery SOC, an initial error covariance matrix, a process noise covariance matrix, and a measurement noise covariance matrix to obtain a first SOC estimated value of the battery; an optimization module for optimizing a process noise covariance matrix and a measurement noise covariance matrix based on a first SOC estimate using a Harris Eagle optimization algorithm; a first determining module, configured to determine a priori estimated value of the battery SOC according to a state transition equation as a first state vector; a second determination module, configured to determine a priori estimates of an error covariance matrix based on a state transfer equation and an optimized process noise covariance matrix; a third determination module, configured to determine the Kalman gain based on the optimized measurement noise covariance matrix and the prior estimation of the error covariance matrix; A first updating module is configured to update the first state vector according to the observed value of the battery SOC and the first SOC estimated value to obtain a second state vector; A second updating module is used to update the error covariance matrix according to the Kalman gain, the observation equation and the prior estimate of the error covariance matrix; The second SOC prediction module is used to use the extended Kalman filter algorithm to predict the battery SOC based on the optimized process noise covariance matrix and measurement noise covariance matrix, the second state vector and the updated error covariance matrix, to obtain the second SOC estimation value of the battery, which is used as the final battery SOC prediction value.

10. A computer-readable storage medium, characterized in that Used to store a computer program; when the computer program is executed by a processor, the steps of the electric vehicle battery SOC prediction method according to any one of claims 1 to 8 are implemented.

Images