Method, device and equipment for estimating state of charge of battery based on hysteresis battery model

Abstract

The invention provides a method, a device and equipment for estimating the state of charge of a battery based on a hysteresis battery model. The method comprises the following steps: acquiring an initial charge state of an energy storage battery; determining a numerical interval where the initial charge state is located; a target parameter set matched with the numerical value interval is determined in multiple parameter sets, the target parameter set is used for recording values of all parameters in a hysteresis battery model, and the hysteresis battery model is used for simulating the hysteresis effect of the energy storage battery and compensating for errors generated by the hysteresis effect; and determining a target charge state of the energy storage battery according to the hysteresis battery model and the target parameter set. According to the method, the hysteresis effect of the energy storage battery is simulated and the error generated by the hysteresis effect is compensated by designing the hysteresis battery model, and the target parameter set most suitable for the hysteresis battery model is acquired according to the numerical interval of the initial charge state, so that the charge state of the energy storage battery can be accurately calculated.

Description

Technical Field

[0001] This application relates to the field of energy storage battery technology, and in particular to a method, apparatus and equipment for estimating the state of charge of a battery based on a hysteresis battery model. Background Technology

[0002] As the core of an energy storage system, the accurate estimation of the State of Charge (SOC) of the battery is crucial for the system's safety and efficient operation. However, SOC estimation for energy storage batteries faces several technical challenges. First, there is a complex nonlinear relationship between the battery's SOC and parameters such as current and voltage, making it difficult for traditional estimation methods to provide high-precision estimates under different operating conditions. Second, as battery usage time increases, factors such as temperature changes and aging cause changes in the battery's electrochemical characteristics, leading to the gradual failure of estimation methods based on simple models. Finally, SOC estimation also needs to comprehensively consider the battery's charge and discharge history and dynamic changes in internal impedance, further increasing the difficulty of real-time estimation. Therefore, improving the accuracy and robustness of SOC estimation has become a key challenge in current energy storage battery management technology.

[0003] The State of Charge (SOC) of a battery cannot be directly measured by sensors and must rely on model-based indirect estimation methods. A key technical challenge in accurately estimating SOC in energy storage batteries lies in the complex hysteresis effect between the Open-Circuit Voltage (OCV) and the SOC. This hysteresis phenomenon stems from various physicochemical factors, including thermodynamics, entropy effects, mechanical stress, and microstructural changes, resulting in a non-linear relationship between OCV and SOC. Especially in the flatter regions of the hysteresis curve, small voltage measurement errors are amplified, significantly impacting the accuracy of SOC estimation. Therefore, existing estimation methods generally have limitations in addressing the hysteresis problem. For example, the traditional average loop method has a large estimation error, while some improved models that consider the hysteresis effect still lack sufficient error suppression capabilities under conditions such as small hysteresis ranges or frequent charge-discharge cycles. Furthermore, even combining advanced dynamic estimation methods such as Kalman filtering or neural networks has not effectively solved the problems caused by the hysteresis effect. Summary of the Invention

[0004] This application provides a method, apparatus, and device for estimating the state of charge (SOC) of a battery based on a hysteresis battery model. This method can accurately calculate the SOC of an energy storage battery, overcoming the problem in the prior art where the SOC of an energy storage battery cannot be accurately calculated due to the hysteresis effect.

[0005] This application provides a method for estimating the state of charge of an energy storage battery based on a hysteresis battery model, comprising the following steps: Obtain the initial state of charge of the energy storage battery; A target parameter set is determined from multiple parameter sets to match the numerical range of the initial state of charge. The target parameter set includes the values ​​of each parameter in the hysteresis battery model, which is used to simulate the hysteresis effect of the energy storage battery and compensate for the error caused by the hysteresis effect. The target state of charge of the energy storage battery is determined based on the hysteresis battery model and the target parameter set.

[0006] According to the energy storage battery state-of-charge estimation method based on a hysteresis battery model provided in this application, determining the target state-of-charge of the energy storage battery based on the hysteresis battery model and the target parameter set includes: Based on the hysteresis battery model and the target parameter set, the initial open-circuit voltage of the energy storage battery is determined; The initial open-circuit voltage is input into the Presac model to obtain the target state of charge of the energy storage battery; The Presack model is used to determine the target state of charge based on the initial open-circuit voltage and the correspondence between the open-circuit voltage and the state of charge of the energy storage battery under hysteresis.

[0007] According to the energy storage battery state-of-charge estimation method based on a hysteresis battery model provided in this application, the step of inputting the initial open-circuit voltage into the Presack model to obtain the target state of charge of the energy storage battery includes: The initial open-circuit voltage is input into the Presack model to obtain the first state of charge. The first state of charge is determined by the Presack model based on the initial open-circuit voltage, its internal hysteresis state, and a first weight. The hysteresis state is a quantitative description of the current hysteresis effect of the energy storage battery, and the first weight represents the degree of influence of each region in the mathematical model of the energy storage battery on the state of charge. The first weight is updated based on the first state of charge to obtain the second weight; The initial open-circuit voltage is re-inputted into the Presack model to obtain a second state of charge. The second state of charge is determined by the Presack model based on the initial open-circuit voltage, its internal hysteresis state, and the second weight. The second state of charge is the target state of charge.

[0008] According to the energy storage battery state-of-charge estimation method based on the hysteresis battery model provided in this application, the step of updating the first weights based on the first state of charge to obtain the second weights includes: Based on the first state of charge and the hysteresis battery model, the theoretical output current of the energy storage battery is determined; The current error is determined based on the theoretical output current and the actual output current corresponding to the energy storage battery. The first weight is updated based on the current error to obtain the second weight.

[0009] According to the energy storage battery state-of-charge estimation method based on a hysteresis battery model provided in this application, the step of updating the first weight based on the current error to obtain the second weight includes: Based on the current error, the first weight is updated using the following formula to obtain the second weight; in, Indicates the first Each sampling time, Indicates the first Each sampling time, Indicates the gain coefficient. Indicates current error. Indicates weight, This indicates the updated weights. This indicates a hysteresis state.

[0010] According to the energy storage battery state-of-charge estimation method based on a hysteresis battery model provided in this application, determining the initial open-circuit voltage of the energy storage battery based on the hysteresis battery model and the target parameter set includes: Based on the values ​​of each parameter in the target parameter set, assign values ​​to each parameter in the hysteresis battery model; Based on the assigned hysteresis battery model, the initial open-circuit voltage of the energy storage battery is determined.

[0011] According to the energy storage battery state-of-charge estimation method based on a hysteresis battery model provided in this application, the hysteresis battery model includes a voltage source, an ohmic internal resistance, a RC network, and a hysteresis module. The RC network includes a capacitor and a resistor. The hysteresis battery model is represented as follows: in, Indicates the first One sampling point, Indicates the attenuation factor. Indicates the sampling interval. This indicates the voltage of the hysteresis module. Represents current. Indicates the maximum hysteresis voltage. This represents the capacitance in an RC network. This represents the resistance in an RC network. Indicates capacitance voltage, Represents the time constant. = , Indicates the internal resistance of the ohm. This represents the open-circuit voltage of the voltage source corresponding to the SOC. This represents the total terminal voltage between the positive and negative terminals of the energy storage battery.

[0012] According to the energy storage battery state-of-charge estimation method based on the hysteresis battery model provided in this application, the multiple parameter sets are determined in the following manner: The numerical range of the state of charge of the energy storage battery is divided into multiple numerical intervals; In each of the numerical intervals, for the values ​​of each parameter in the hysteresis battery model, at least one parameter optimization is performed using the particle swarm optimization algorithm to obtain the optimal parameter set corresponding to each of the numerical intervals. The optimal parameter set corresponding to each of the aforementioned numerical intervals is determined as the plurality of parameter sets. This application also provides a device for estimating the state of charge of an energy storage battery based on a hysteresis battery model, comprising the following modules: The acquisition module is used to acquire the initial state of charge of the energy storage battery; The first determining module is used to determine a target parameter set that matches the numerical range of the initial state of charge from multiple parameter sets. The target parameter set includes the values ​​of each parameter in the hysteresis battery model, which is used to simulate the hysteresis effect of the energy storage battery and compensate for the error caused by the hysteresis effect. The second determining module is used to determine the target state of charge of the energy storage battery based on the hysteresis battery model and the target parameter set.

[0013] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the energy storage battery state-of-charge estimation method based on the hysteresis battery model as described above.

[0014] This application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the energy storage battery state-of-charge estimation method based on the hysteresis battery model as described above.

[0015] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the energy storage battery state-of-charge estimation method based on the hysteresis battery model as described above.

[0016] This application provides a method, apparatus, and device for estimating the state of charge (SOC) of a battery based on a hysteresis battery model. The method of this application first obtains the initial SOC of the energy storage battery; then, it determines the numerical range within which the initial SOC lies, and identifies a target parameter set that matches this range from multiple parameter sets; finally, it determines the target SOC of the energy storage battery based on the hysteresis battery model and the target parameter set. The first aspect of this application designs a hysteresis battery model. Since the hysteresis battery model can simulate the hysteresis effect of the energy storage battery and compensate for the errors caused by the hysteresis effect, using the hysteresis battery model can improve the accuracy of the calculated SOC. The second aspect of this application obtains the target parameter set most suitable for the hysteresis battery model based on the numerical range within which the initial SOC lies, further improving the accuracy of the calculation results when the hysteresis battery model uses the target parameter set to calculate the SOC. Therefore, this application can effectively overcome the problem in the prior art where the SOC of energy storage batteries cannot be accurately calculated due to the hysteresis effect. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating a battery state-of-charge estimation method based on a hysteresis battery model, as shown in one embodiment of this application.

[0019] Figure 2 This is a schematic diagram of an OCV-SOC curve shown in one embodiment of this application.

[0020] Figure 3 This is a schematic diagram of a hysteresis curve shown in one embodiment of this application.

[0021] Figure 4 This is a circuit diagram of a hysteresis battery model shown in one embodiment of this application.

[0022] Figure 5 This is a schematic diagram of a charge / discharge input variation curve shown in one embodiment of this application.

[0023] Figure 6 This is a schematic diagram of a Preisach planar stepped memory curve provided in an embodiment of this application.

[0024] Figure 7This is a schematic diagram of a discrete Preisach operator scheme provided in an example of this application.

[0025] Figure 8 This is a complete flowchart illustrating a method for estimating the state of charge of an energy storage battery according to an embodiment of this application.

[0026] Figure 9 This is a structural block diagram of a battery state-of-charge estimation device based on a hysteresis battery model, as shown in one embodiment of this application.

[0027] Figure 10 This is a schematic diagram of the physical structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0029] The method provided in this application is implemented by a battery state of charge estimation system. After the initial state of charge of the energy storage battery is input into the battery state of charge estimation system, the system can accurately calculate the state of charge of the energy storage battery at each subsequent sampling time.

[0030] Figure 1 This is a flowchart illustrating a battery state-of-charge estimation method based on a hysteresis battery model, as shown in one embodiment of this application. (Refer to...) Figure 1 The method of this application may include: Step 101: Obtain the initial state of charge of the energy storage battery.

[0031] In this embodiment, the initial state of charge (SOC) of the energy storage battery refers to the first SOC of the energy storage system received by the battery SOC estimation system after startup. The battery SOC estimation system needs this initial SOC as a starting point to calculate the accurate SOC of the energy storage battery at subsequent sampling times.

[0032] The initial state of charge (SPC) can be the most recently recorded SPC read when the energy storage battery is turned on, or it can be a SPC calculated using various algorithms. This application does not impose specific restrictions on the method of obtaining the initial SPC.

[0033] Step 102: Determine the target parameter set that matches the numerical range of the initial state of charge from multiple parameter sets. The target parameter set is used to record the values ​​of each parameter in the hysteresis battery model. The hysteresis battery model is used to simulate the hysteresis effect of the energy storage battery and to compensate for the error caused by the hysteresis effect.

[0034] In this embodiment, the state of charge of the energy storage battery can be pre-divided into multiple different numerical ranges. For example, when the step size is 5%, the multiple numerical ranges obtained can be 95%-100%, 90%-95%, etc. 0%-5%, and when the initial state of charge is 96%, the corresponding value range is 95%-100%.

[0035] In actual implementation, the numerical range can also be divided in other ways, and this embodiment does not limit this.

[0036] In this embodiment, an optimal set of parameters is pre-determined for each numerical interval. The optimal set of parameters corresponding to a certain numerical interval X records the optimal values ​​of each parameter in the hysteresis battery model when the state of charge is within the numerical interval X.

[0037] After determining the numerical range of the initial state of charge, step 102 can further determine the optimal set of parameters corresponding to that numerical range, i.e., the target set of parameters.

[0038] Among them, the hysteresis battery model can simulate the hysteresis effect of energy storage batteries and compensate for the error caused by the hysteresis effect when calculating the state of charge.

[0039] Before performing step 101 of this application, some preliminary work is required, which will be described in detail below.

[0040] First, data initialization and testing.

[0041] A series of standardized experimental tests were conducted on the cells of the energy storage battery to obtain data on the correspondence between the open-circuit voltage (OCV) and the state of charge (SOC) during the charging and discharging processes, ultimately yielding the OCV-SOC curve and hysteresis curve. The OCV-SOC curve is shown below. Figure 2 As shown, the hysteresis curve is as follows Figure 3 As shown. Figure 2 This is a schematic diagram of an OCV-SOC curve shown in one embodiment of this application. Figure 3 This is a schematic diagram of a hysteresis curve shown in one embodiment of this application.

[0042] exist Figure 2In the graph, the horizontal axis represents the battery's state of charge (SOC), ranging from 0 to 1 (i.e., 0% to 100%), and the vertical axis represents the open-circuit voltage (OCV), measured in volts (V). Figure 2 It includes charging OCV curves and discharging OCV curves. The charging OCV curve, represented by the solid line at the top, shows the OCV value at different SOC points during battery charging. The discharging OCV curve, represented by the dashed line at the bottom, shows the OCV value at different SOC points during battery discharging. At the same SOC point, the OCV during charging is always higher than the OCV during discharging. The closed region formed between the two curves is called a hysteresis loop.

[0043] exist Figure 3 In the graph, the horizontal axis is SOC and the vertical axis is OCV / V. Figure 3 The image shows the superposition of multiple OCV-SOC curves obtained through experimental testing, which together form a series of hysteresis loops. Based on... Figure 3 Through repeated cyclic testing, it was found that the basic shape and range of the hysteresis loop are relatively fixed, that is, the hysteresis phenomenon has repeatability and stability, which provides a data foundation for the subsequent use of the Preisach model to describe and predict the hysteresis phenomenon.

[0044] Second, a first-order RC hysteresis battery model is constructed.

[0045] This application adds a hysteresis module to the traditional first-order RC battery model, ultimately constructing a first-order RC hysteresis battery model, the circuit diagram of which is shown below. Figure 4 As shown. The first-order RC hysteresis battery model is the hysteresis battery model mentioned above. The first-order RC hysteresis battery model of this application consists of four parts: an ideal voltage source, an ohmic internal resistance, an RC network (RC network), and a hysteresis module. Figure 4 This is a circuit diagram of a hysteresis battery model shown in one embodiment of this application.

[0046] An ideal voltage source represents the battery's open-circuit voltage (OCV), which is the battery's most ideal and core voltage. This voltage value is directly related to the battery's remaining charge (SOC).

[0047] This represents the battery's internal resistance in ohms. When current flows through it, a voltage drop is immediately generated across this resistor, simulating the most direct and rapid voltage loss inside the battery.

[0048] RC networks consist of resistors and capacitor It is composed of parallel components, and the voltage across its terminals is The RC network represents the battery's polarization effect. It's a slow-response module that simulates the slow, gradual change in battery voltage as current changes. This is the voltage loss caused by the polarization effect.

[0049] A hysteresis module is a special voltage module specifically designed to compensate for and simulate the hysteresis phenomenon in batteries (i.e., the charging voltage is higher than the discharging voltage for the same amount of charge). The voltage across the hysteresis module is... .Voltage It is a dynamic value that is adjusted according to factors such as current, thus adding a correction term to the final output voltage to make the model more realistic.

[0050] The voltage effects of the four modules mentioned above are combined to ultimately form the battery's output terminal voltage. .

[0051] Step 103: Determine the target state of charge of the energy storage battery based on the hysteresis battery model and the target parameter set.

[0052] In this embodiment, the parameters in the hysteresis battery model are assigned values ​​using the values ​​of each parameter in the target parameter set. Then, the state of charge of the energy storage battery, i.e. the target state of charge, can be accurately calculated using the assigned hysteresis battery model.

[0053] The method of this application first obtains the initial state of charge (SOC) of the energy storage battery. Then, it determines the numerical range within which the initial SOC lies and identifies a target parameter set that matches this range from multiple parameter sets. Finally, it determines the target SOC of the energy storage battery based on the hysteresis battery model and the target parameter set. The first aspect of this application designs a hysteresis battery model. Since the hysteresis battery model can simulate the hysteresis effect of the energy storage battery and compensate for the errors caused by the hysteresis effect, using the hysteresis battery model can improve the accuracy of the final calculated SOC. The second aspect of this application obtains the most suitable target parameter set for the hysteresis battery model based on the numerical range within which the initial SOC lies, further improving the accuracy of the calculation results when the hysteresis battery model uses the target parameter set to calculate the SOC. Therefore, this application can effectively overcome the problem in the prior art where the SOC of the energy storage battery cannot be accurately calculated due to the hysteresis effect.

[0054] In conjunction with the above embodiments, in one implementation, step 103 may include: Step 1031: Determine the initial open-circuit voltage of the energy storage battery based on the hysteresis battery model and the target parameter set.

[0055] When performing step 1031, the parameters in the hysteresis battery model can be assigned values ​​according to the values ​​of each parameter in the target parameter set; then, the initial open-circuit voltage of the energy storage battery can be determined according to the assigned values ​​of the hysteresis battery model.

[0056] In this embodiment, the hysteresis battery model is represented as follows: (1) (2) (3) in, Indicates the first One sampling point, Indicates the attenuation factor. Indicates the sampling interval. This indicates the voltage of the hysteresis module. Represents current. Indicates the maximum hysteresis voltage. This represents the capacitance in an RC network. This represents the resistance in an RC network. Indicates capacitance voltage, Represents the time constant. = , Indicates the internal resistance of the ohm. This represents the open-circuit voltage of the voltage source corresponding to the SOC. This represents the total terminal voltage between the positive and negative terminals of the energy storage battery.

[0057] Formula (1) describes Figure 4 Voltage across the hysteresis module How does it change over time? Among them, This is a decay term, indicating that the existing hysteresis voltage will gradually disappear; The term "establishment" indicates that the current is establishing a new value, tending towards its maximum value. The hysteresis voltage. As can be seen from formula (1), the hysteresis voltage at the next moment is a dynamic equilibrium result determined by the decay of the old voltage and the establishment of the new voltage.

[0058] Formula (2) describes Figure 4 middle Voltage at both ends of a parallel network How does it change over time? Indicates capacitance The voltage; This is the natural discharge term, representing the capacitance. The voltage on will pass through Natural decay; The term "current charging" indicates that a new voltage is established on the module when current flows through it. Therefore, the polarization voltage at the next moment is determined by both the voltage remaining after the capacitor's natural discharge and the voltage newly charged by the current.

[0059] Formula (3) describes the total terminal voltage between the positive and negative terminals of the energy storage battery. The final output voltage of the energy storage battery = ideal voltage - two main voltage losses (ohmic loss + polarization loss) + a special hysteresis compensation.

[0060] Step 1032: Input the initial open-circuit voltage into the Presack model to obtain the target state of charge of the energy storage battery; wherein, the Presack model is used to determine the target state of charge based on the initial open-circuit voltage and the correspondence between the open-circuit voltage and the state of charge of the energy storage battery under the hysteresis effect.

[0061] In this embodiment, to accurately describe and quantify the OCV-SOC hysteresis effect in the energy storage battery, the Presack model is employed. This model possesses core memory capabilities, recording the historical changes of the input signal (i.e., SOC), that is, recording the correspondence between the open-circuit voltage and the state of charge (SOC) of the energy storage battery under the hysteresis effect. Therefore, after inputting the initial open-circuit voltage into the Presack model, the Presack model can accurately determine the SOC of the energy storage battery based on the initial open-circuit voltage and the correspondence between the open-circuit voltage and the SOC under the hysteresis effect.

[0062] In one implementation, step 1031 may include: The initial open-circuit voltage is input into the Presack model to obtain the first state of charge. The first state of charge is determined by the Presack model based on the initial open-circuit voltage, its internal hysteresis state, and the first weight. The hysteresis state is a quantitative description of the current hysteresis effect of the energy storage battery, and the first weight represents the degree of influence of each region in the mathematical model of the energy storage battery on the state of charge. The first weight is updated based on the first state of charge to obtain the second weight; The initial open-circuit voltage is re-inputted into the Presack model to obtain the second state of charge. The second state of charge is determined by the Presack model based on the initial open-circuit voltage, its internal hysteresis state, and the second weight. The second state of charge is the target state of charge.

[0063] The second weight, obtained by updating the first weight based on the first state of charge, may include: Based on the first state of charge and the hysteresis battery model, the theoretical output current of the energy storage battery is determined. Obtain the actual output current corresponding to the energy storage battery; Determine the current error based on the theoretical output current and the actual output current; The first weight is updated based on the current error to obtain the second weight.

[0064] In this embodiment, based on the current error, the first weight can be updated using the following formula to obtain the second weight; (4) in, Indicates the first Each sampling time, Indicates the first Each sampling time, Indicates the gain coefficient. Indicates current error. Indicates weight, This indicates the updated weights. This indicates a hysteresis state.

[0065] Next, the process of estimating the state of charge of the energy storage battery in real time based on the Preisach model will be described in detail. This estimation operation is performed at each time step (sampling time). Execution. Within a time step. The following steps need to be performed: Step 1: Input the initial open-circuit voltage OCV and calculate the initial SOC.

[0066] First, an estimated OCV (Open Circuit Voltage) is calculated using a first-order RC hysteresis battery model. This OCV value is then used as input to the Preisach model. The Preisach model is a model that can remember historical inputs. Based on the input OCV value and the internal hysteresis state vector and weight vector, it calculates an initial SOC (State of Charge).

[0067] Specifically, the current moment calculated by the first-order RC hysteresis battery model The initial open-circuit voltage (OCV) is input into the Preisach model, and the Preisach model is combined with the current time step. The hysteresis state and the previous moment The weights are used to calculate an uncorrected, initial SOC, called the initial SOC. The formula is as follows: (5) in, For the previous moment The weight vector. For the current moment The hysteresis state vector, which is updated according to the history of SOC changes, reflects the battery's memory effect. Indicates the use of time up to the previous moment. The model information for the current moment The initial prediction is made based on the SOC.

[0068] Step 2: Calculate the current error.

[0069] Calculate the current error based on the initial SOC output from the previous step. ), and calculate the current time in reverse. The theoretical current value is calculated and compared with the actual current value measured by the battery management system (BMS) to obtain the difference between the two, i.e., the current error. .

[0070] Specifically, using a first-order RC hysteresis battery model, based on the initial SOC ( This involves working backwards to calculate the theoretically required output or input current value of the energy storage battery at that moment, i.e., estimating the current value, which is expressed as... Next, the actual current value of the energy storage battery will be measured in real time from the battery sensor. , and the estimated current value Subtracting the two values ​​yields the difference, which is the current error. The formula is as follows: (6) Step 3: Update the weights.

[0071] Using current error, a model parameter feedback correction algorithm is used to adjust the weight vector of the Preisach model. Adjustments and updates are needed. The update formula is: in, This is a learning rate or gain coefficient. For a moment The hysteresis state vector.

[0072] Specifically, the weight vector is adjusted based on the magnitude and direction of the current error. If the error is positive, it indicates that the model's estimation deviates from reality, and the weights are adjusted in one direction; if it is negative, they are adjusted in the other direction. The above formula represents adding a correction amount determined by both the error and the current state to the old weights.

[0073] Finally, step three yields a new, revised weight vector. .

[0074] Step 4: Correct the SOC result and proceed to the next loop.

[0075] The updated weight vector from the previous step Substituting the values ​​back into the Preisach model, a new, corrected SOC estimate is obtained. Then, the current time... Add a time step For the next moment Prepare for the estimation.

[0076] Specifically, the updated weight vector Substitute the Preisach model back into the calculation and perform the calculation again. The calculation formula is as follows: (7) in, For the current moment The hysteresis state vector. Because this calculation uses error-corrected weights, the result is closer to the true value than the initial SOC in step one. Finally, the current time step is obtained. Final, revised SOC estimate .this The value is the final output of this iteration, i.e., the target state of charge. Simultaneously, the updated weight vector... and the current hysteresis state It will be saved as the next sampling time. The state at the previous moment during calculation, followed by the system time step. And repeat the entire loop.

[0077] Step 5: Loop and End.

[0078] The system determines whether the estimation task needs to be terminated. If the task is not terminated, the process returns to step one, reads the initial open-circuit voltage at the new moment, and starts a new cycle of initial calculation, error feedback, weight update, and result correction. If the task is terminated, the final result is output and the process is terminated.

[0079] In this embodiment, a first-order RC hysteresis battery model is combined with an adaptive Presack model. Through complementary advantages and closed-loop adaptive correction mechanisms, the accuracy of the state of charge (SOC) estimation of the energy storage battery is significantly improved. On the one hand, the first-order RC hysteresis battery model excels at describing the dynamic electrochemical processes of the battery, can simulate the polarization effect caused by changes in load current in real time, and performs preliminary dynamic compensation for the hysteresis voltage, thereby outputting a preliminarily corrected open-circuit voltage (OCV), solving the problem that traditional methods rely solely on the static OCV-SOC curve and ignore the battery's dynamic response. On the other hand, the Presack model receives dynamic OCV values ​​and utilizes its historical memory capability to handle the complex nonlinear mapping relationship between OCV and SOC caused by the hysteresis effect. This estimation method allows for a refined description of both the battery's dynamic and static hysteresis characteristics, thus ensuring the accuracy of the final calculated state of charge of the energy storage battery.

[0080] Secondly, this application does not simply cascade two different models, but introduces a feedback loop based on real physical quantities (measured current). By comparing the theoretical current corresponding to the initial SOC estimated by the model with the actual current measured by the sensor, a real-time error signal is obtained. This error signal is used to continuously update and correct the weight vector in the Presack model online through a model parameter feedback correction algorithm. This mechanism endows the model with the ability to learn and calibrate in real time, which not only effectively eliminates the accumulation of initial SOC error, but also dynamically adapts to model parameter drift caused by factors such as battery aging and temperature changes, ensuring high accuracy even under long-term and complex operating conditions.

[0081] In conjunction with the above embodiments, in one implementation, the multiple parameter sets in step 102 are determined in the following manner: The numerical range of the state of charge of energy storage batteries is divided into multiple numerical intervals; In each numerical interval, for the values ​​of each parameter in the hysteresis battery model, the particle swarm optimization algorithm is used to perform at least one parameter optimization to obtain the optimal parameter set corresponding to each numerical interval. The optimal parameter set corresponding to each numerical interval is determined as multiple parameter sets.

[0082] Before performing step 101, model parameter identification is also required. The goal of this step is to find the most accurate values ​​for all unknown parameters in the established first-order RC hysteresis battery model, thereby enabling the model to reproduce the battery's true electrochemical behavior to the greatest extent possible. Model parameter identification may include the following steps: 1) Determine the identification targets and evaluation criteria.

[0083] The parameters to be identified are defined as a set containing all the key factors describing the battery's ohmic, polarization, and hysteresis characteristics; these parameters are organized into a vector. .

[0084] in, Internal resistance for charging (ohms) The discharge internal resistance is ohmic. For polarization internal resistance, It is a time constant. H is the attenuation factor, and H is the maximum hysteresis voltage.

[0085] The goal of parameter identification is to minimize the error between the model's output voltage and the experimentally measured voltage. To quantify this error, this application uses the root mean square error (RMSE) as the core metric for evaluating model accuracy. The identification algorithm aims to find the set of parameters that minimizes the RMSE value. The formula for calculating RMSE is as follows: in, The theoretical terminal voltage calculated based on a first-order RC hysteresis battery model. .

[0086] 2) Execute the PSO interval identification algorithm.

[0087] After defining the target and criteria, the identification process begins. This process employs a modified Particle Swarm Optimization (PSO) interval identification algorithm and includes the following steps: A: SOC Range Division. To improve the accuracy of parameters under different states of charge, the entire SOC range (0%~100%) is first divided into multiple independent numerical ranges. For example, it can be divided into 20 numerical ranges, such as 100%-95%, 95%-90%...0%-5%. Subsequent parameter identification will be performed independently within each numerical range.

[0088] B: Iterative optimization within the interval. For a specific SOC numerical interval, perform the following operations: Start the Particle Swarm Optimization (PSO) algorithm to optimize the parameters on the data in that interval. This optimization process will be repeated at least once (set according to actual needs), for example, 10 times.

[0089] C: Optimal Parameter Selection. After at least one optimization for a numerical range, at least one set of candidate parameters will be obtained. The set of model parameters that minimizes both the model value and the experimental error value is then selected as the optimal parameter set for that SOC numerical range.

[0090] 3) Obtain the final parameter set.

[0091] The iterative optimization and optimal parameter selection process within the aforementioned intervals is repeated for each SOC numerical interval. After identifying all numerical intervals, the final result is not a fixed set of parameters, but a series of optimal parameter sets corresponding to different SOC numerical intervals. In subsequent SOC estimation, the model will call the optimal parameter set for the corresponding numerical interval for calculation based on the current SOC, thereby achieving higher-precision dynamic simulation.

[0092] To accurately describe the OCV-SOC hysteresis effect in energy storage batteries, this application adopts the Preisach model, which has core memory properties and can record the historical changes of the input signal.

[0093] Theoretically, the Preisach model treats macroscopic hysteresis phenomena as a large number of microscopic, fundamental hysteresis operators. The result of superposition. Its continuous mathematical expression is a double integral: (8) in, and Represent Open-circuit voltage and state of charge at any given time. Represents the weighting function. The x-axis is... The vertical axis represents the change in SOC. On the plane, the output value of the hysteresis factor (hysteresis operator) It changes between 1 and -1.

[0094] According to the Preisach model, the Preisach plane divides sub-regions by uniformly dividing each axis into n parts, thus the number of grids on the plane is... It can be calculated as Then each square in the Preisach plane It can be represented as: (9) In the Preisach model definition, the weight function on each square is treated as a constant, thus discretizing the model. The integral, substituted and summed, yields: (10) (11) In the formula, and Let be the hysteresis factor value and weight function on any grid. Divide the Preisach triangle into two equal parts, then for any and , and The value is constant: (12) (13) At this point, a new variable hysteresis state is introduced. : (14) In summary, the discrete Preisach model can be expressed as: (15) in, It can be determined by the input SOC. It represents the output at the same point on the hysteresis curve, while the weight function is obtained by solving the function offline using a computer. After determining the weight vector, the Preisach model can be used to describe the hysteresis curve, i.e., when the SOC is at... When the time interval changes, the effect of the memory curve on the square in the Preisach triangle forms a new hysteresis state vector, which in turn combines the two vectors. and Multiplying yields the output OCV. If the error between the experimental output and the model output is large, adjust the number of model partitions. This process continues until the output accuracy meets expectations. This process is called model training.

[0095] In the Preisach model, the discretely partitioned planar regions form the physical and data structure foundation connecting the weights and hysteresis states. Specifically, the weight vector... Each element corresponds to a region, which, as a relatively static parameter, quantifies the inherent contribution of the microscopic hysteresis characteristics represented by that region to the macroscopic voltage of the battery. Correspondingly, the hysteresis state vector... The elements also correspond one-to-one with the regions, but it is a dynamic variable whose value (usually 1) is... or The hysteresis state is determined in real-time by the battery's historical charging and discharging paths, reflecting whether each region is activated at the current moment. Therefore, the divided regions are a map, the weights are the importance level of each region on the map, and the hysteresis state is the real-time status of each region. (In the process of...) During calculation, the voltage output value that reflects the complete historical information of the battery is obtained by integrating the real-time status of all regions with their corresponding importance levels.

[0096] The Preisach model has memory properties, and its output depends not only on the current input (SOC), but also on the historical path and historical extreme points of the input signal. Figure 5 The charging and discharging input variation curves in the embodiment are shown. Fluctuations over time represent the charging and discharging process of the battery; local maxima (denoted as...) ) and local minima (denoted as These are recorded by the model and form the basis of the memory trajectory. Figure 5 This is a schematic diagram of a charge / discharge input variation curve shown in one embodiment of this application.

[0097] like Figure 6 As shown, the aforementioned extreme points are mapped onto the Preisach triangle plane, forming a stepped memory curve that represents the current internal hysteresis state of the model. The geometry of this curve on the plane is uniquely determined by the input history and is a visual representation of the system's hysteresis behavior. Figure 6This is a schematic diagram of a Preisach plane stepped memory curve provided in one embodiment of this application, illustrating the internal state of the Preisach model, i.e., visualizing the Preisach plane, used to record the historical changes of the input signal (SOC), specifically demonstrating the model's memory function. Within this plane, The axis represents the reversal point (historical peak) during the rise of the input signal, while The axis represents the reversal point (historical trough) in the downward process. The line is a solid diagonal line that passes through the origin, and... shaft and The axes together define the Preisach triangle, defining the effective working area of ​​the model.

[0098] Figure 6 The core is a stepped curve composed of dashed lines, which acts as a memory carrier. Its shape and position are determined by the historical peak values ​​of the input signal. Valley value Uniquely determined. The current input value (SOC) and its horizontal dashed line indicate the latest state of the stepped memory curve. This stepped curve not only divides the different states of countless microscopic hysteresis operators on the plane, but also defines the current state of the overall macroscopic hysteresis characteristics of the battery, playing a crucial demarcation role.

[0099] Because the Preisach triangle contains a large number of hysteresis factors, the corresponding weight functions are difficult to obtain directly. Theoretically, the Preisach model consists of countless microscopic hysteresis operators and their weight functions, and directly solving for these continuously distributed weight functions is extremely difficult. In practical applications, the Preisach model is discretized using numerical methods. The specific steps are as follows: the Preisach triangle is divided into multiple square regions, thereby simplifying the continuous model into a finite-dimensional model. Figure 7 This process was demonstrated, namely, the process of... Figure 6 The continuous Preisach plane is meshed to form a finite number of small square regions, each representing a discrete Preisach operator with assumed constant weights. This method simplifies the originally infinite-dimensional continuous model into a finite-dimensional model that can be processed on a computer, facilitating computation and implementation. Figure 7 This is a schematic diagram of a discrete Preisach operator scheme provided in an example of this application.

[0100] When the SOC is at time When changes occur, the effect of the memory curve on the square in the Preisach triangle will generate a new hysteresis state vector. That is, when SOC changes, Figure 6 The memory curve will shift accordingly, and the change in the curve's position will alter its relationship with... Figure 7The relative relationships of the various square grids are determined, and the result of this change is quantized into a new hysteresis state vector. By using two vectors and Multiply to get the output. The dynamically changing hysteresis state vector With a pre-calibrated static weight vector representing the weights of each grid cell By performing multiplication, the current OCV output value can be obtained. If the error between the experimental results and the model output is large, the number of partitions in the model can be adjusted. Continue this process until the output accuracy meets the expected requirements. If the model accuracy is insufficient, the number of partitions can be increased. (that is, let) Figure 7 (A denser mesh is used in the model) to improve the model's accuracy, and the initial SOC is calculated based on this OCV value.

[0101] The formula for calculating the initial SOC using the OCV value estimated through the hysteresis equivalent circuit model is as follows: (16) Figure 7 The discretization process for transforming a continuous Preisach theoretical model into a finite-dimensional numerical model is illustrated to meet engineering application requirements. In the figure, shaft and The axes form the Preisach plane coordinate system. and These represent the lower and upper limits of the state of charge (SOC), respectively, and are perpendicular to the diagonal. The effective computational domain of the Preisach triangle is defined by these elements. This triangular region is divided into several square grid cells (e.g., A1, A2, etc.), each cell representing a discretized Preisach operator. Through this division, theoretically an infinite number of microscopic hysteresis factors are simplified to a finite number of discrete cells (15 in this example). Weight vector Hysteresis state vector Each element corresponds to a fixed weight and a real-time activation state of each grid cell. This discretization method transforms the continuous integration operations in the original model into a vector inner product form, significantly reducing computational complexity and is key to achieving efficient numerical solutions.

[0102] This application discloses a method for estimating the State of Charge (SOC) of an energy storage battery based on an equivalent circuit model and an adaptive Preisach model. The entire process is as follows: Figure 8 As shown. Figure 8 This is a complete flowchart illustrating a method for estimating the state of charge (SOC) of an energy storage battery according to an embodiment of this application. Figure 8As shown, the method of this application includes: 1) obtaining charge / discharge OCV-SOC curves and hysteresis curves by performing various tests on the cells of the energy storage battery; 2) adding a hysteresis module to the first-order RC battery model to construct a first-order RC hysteresis battery model, compensating for the original model's neglect of hysteresis; 3) using a particle swarm optimization (PSO) interval identification algorithm to identify parameters of the first-order RC hysteresis battery model; 4) using the OCV value estimated by the hysteresis equivalent circuit model as the input of the Preisach model, dividing the Preisach triangle into several squares to determine the weight function, and calculating the hysteresis state matrix and weights; 5) calculating the initial SOC using the OCV value estimated by the hysteresis equivalent circuit model; 6) adjusting the weight vector by the difference between the current estimated value and the measured current value, thereby correcting the SOC estimate. The original weight parameters are corrected and updated using a model parameter feedback correction algorithm, and the updated model parameters are input into the Preisach model again to obtain a new SOC estimate; 7) continuously updating the weights and hysteresis states, and iteratively estimating the SOC at the next sampling time.

[0103] This application addresses the challenge of accurately estimating the State of Charge (SOC) of energy storage batteries. By adding a hysteresis module to the equivalent circuit model and introducing a discrete Preisach model based on that hysteresis model, combined with a fine division of the OCV-SOC curve of the energy storage battery, the SOC estimate is corrected. Furthermore, to balance accuracy and computational complexity, the number of divisions in the Preisach triangle is adjusted, effectively improving the SOC estimation accuracy and optimizing computational efficiency.

[0104] The energy storage battery state of charge estimation device based on the hysteresis battery model provided in this application is described below. The energy storage battery state of charge estimation device based on the hysteresis battery model described below can be referred to in correspondence with the energy storage battery state of charge estimation method based on the hysteresis battery model described above. Figure 9 This is a structural block diagram of a battery state-of-charge estimation device based on a hysteresis battery model, as shown in one embodiment of this application. (Refer to...) Figure 9 The energy storage battery state-of-charge estimation device 900 based on the hysteresis battery model of this application may include: The acquisition module 901 is used to acquire the initial state of charge of the energy storage battery; The first determining module 902 is used to determine a target parameter set that matches the numerical range of the initial state of charge from multiple parameter sets. The target parameter set includes the values ​​of each parameter in the hysteresis battery model. The hysteresis battery model is used to simulate the hysteresis effect of the energy storage battery and to compensate for the error caused by the hysteresis effect. The second determining module 903 is used to determine the target state of charge of the energy storage battery based on the hysteresis battery model and the target parameter set.

[0105] According to the energy storage battery state of charge estimation device 900 based on a hysteresis battery model provided in this application, the second determining module 903 is specifically used to: determine the initial open-circuit voltage of the energy storage battery according to the hysteresis battery model and the target parameter set; input the initial open-circuit voltage into the Presack model to obtain the target state of charge of the energy storage battery; wherein, the Presack model is used to determine the target state of charge based on the initial open-circuit voltage and the correspondence between the open-circuit voltage and the state of charge of the energy storage battery under the hysteresis effect.

[0106] According to the energy storage battery state of charge estimation device 900 based on a hysteresis battery model provided in this application, the second determining module 903 is specifically used for: inputting the initial open-circuit voltage into the Presack model to obtain a first state of charge, wherein the first state of charge is determined by the Presack model based on the initial open-circuit voltage, its own internal hysteresis state, and a first weight, wherein the hysteresis state is a quantitative description of the current hysteresis effect of the energy storage battery, and the first weight represents the degree of influence of each region in the mathematical model of the energy storage battery on the state of charge; updating the first weight based on the first state of charge to obtain a second weight; and re-inputting the initial open-circuit voltage into the Presack model to obtain a second state of charge, wherein the second state of charge is determined by the Presack model based on the initial open-circuit voltage, its own internal hysteresis state, and the second weight, and the second state of charge is the target state of charge.

[0107] According to the energy storage battery state of charge estimation device 900 based on the hysteresis battery model provided in this application, the second determining module 903 is specifically used to: determine the theoretical output current of the energy storage battery based on the first state of charge and the hysteresis battery model; determine the current error based on the theoretical output current and the actual output current corresponding to the energy storage battery; and update the first weight based on the current error to obtain the second weight.

[0108] According to the energy storage battery state of charge estimation device 900 based on the hysteresis battery model provided in this application, the second determining module 903 is specifically used to: update the first weight according to the current error using the following formula to obtain the second weight; in, Indicates the first Each sampling time, Indicates the first Each sampling time, Indicates the gain coefficient. Indicates current error. Indicates weight, This indicates the updated weights. This indicates a hysteresis state.

[0109] According to the energy storage battery state of charge estimation device 900 based on the hysteresis battery model provided in this application, the second determining module 903 is specifically used to: assign values ​​to each parameter in the hysteresis battery model according to the values ​​of each parameter in the target parameter set; and determine the initial open-circuit voltage of the energy storage battery according to the assigned hysteresis battery model.

[0110] According to the energy storage battery state-of-charge estimation device 900 based on a hysteresis battery model provided in this application, the hysteresis battery model includes a voltage source, an ohmic internal resistance, a RC network, and a hysteresis module. The RC network includes a capacitor and a resistor. The hysteresis battery model is represented as follows: in, Indicates the first One sampling point, Indicates the attenuation factor. Indicates the sampling interval. This indicates the voltage of the hysteresis module. Represents current. Indicates the maximum hysteresis voltage. This represents the capacitance in an RC network. This represents the resistance in an RC network. Indicates capacitance voltage, Represents the time constant. = , Indicates the internal resistance of the ohm. This represents the open-circuit voltage of the voltage source corresponding to the SOC. This represents the total terminal voltage between the positive and negative terminals of the energy storage battery.

[0111] According to the energy storage battery state-of-charge estimation device 900 based on a hysteresis battery model provided in this application, multiple parameter sets are determined in the following manner: The numerical range of the state of charge of energy storage batteries is divided into multiple numerical intervals; In each numerical interval, for the values ​​of each parameter in the hysteresis battery model, the particle swarm optimization algorithm is used to perform at least one parameter optimization to obtain the optimal parameter set corresponding to each numerical interval. The optimal parameter set corresponding to each numerical interval is determined as multiple parameter sets.

[0112] Figure 10 This is a schematic diagram of the physical structure of an electronic device according to an embodiment of this application, as shown below. Figure 10As shown, the electronic device may include a processor 1010, a communications interface 1020, a memory 1030, and a communication bus 1040, wherein the processor 1010, communications interface 1020, and memory 1030 communicate with each other via the communication bus 1040. The processor 1010 can call logical instructions in the memory 1030 to execute a method for estimating the state of charge (SOC) of an energy storage battery based on a hysteresis battery model. This method includes: obtaining the initial SOC of the energy storage battery; determining a target parameter set from multiple parameter sets that matches the numerical range of the initial SOC, wherein the target parameter set includes the values ​​of each parameter in the hysteresis battery model, which is used to simulate the hysteresis effect of the energy storage battery and compensate for errors caused by the hysteresis effect; and determining the target SOC of the energy storage battery based on the hysteresis battery model and the target parameter set.

[0113] Furthermore, the logical instructions in the aforementioned memory 1030 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute a method for estimating the state of charge of an energy storage battery based on a hysteresis battery model provided by the above methods. The method includes: obtaining the initial state of charge of the energy storage battery; determining a target parameter set that matches the numerical range of the initial state of charge from a plurality of parameter sets, wherein the target parameter set includes the values ​​of each parameter in the hysteresis battery model, the hysteresis battery model being used to simulate the hysteresis effect of the energy storage battery and compensate for the error caused by the hysteresis effect; and determining the target state of charge of the energy storage battery according to the hysteresis battery model and the target parameter set.

[0115] In another aspect, this application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for estimating the state of charge (SOC) of an energy storage battery based on a hysteresis battery model, as provided by the methods described above. The method includes: obtaining an initial SOC of the energy storage battery; determining a target parameter set from a plurality of parameter sets that matches the numerical range of the initial SOC, the target parameter set including the values ​​of each parameter in a hysteresis battery model, the hysteresis battery model being used to simulate the hysteresis effect of the energy storage battery and compensate for errors caused by the hysteresis effect; and determining a target SOC of the energy storage battery based on the hysteresis battery model and the target parameter set.

[0116] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0117] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part 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 computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method of estimating state of charge of a battery based on a hysteresis battery model, characterized by, The method comprises: acquiring an initial state of charge of an energy storage battery; determining, from a plurality of parameter sets, a target parameter set matching a numerical interval in which the initial state of charge is located, the target parameter set comprising values of respective parameters in a hysteresis battery model used to simulate a hysteresis effect of the energy storage battery and compensate for errors caused by the hysteresis effect; determining a target state of charge of the energy storage battery according to the hysteresis battery model and the target parameter set.

2. The method of claim 1, wherein, The determining of the target state of charge of the energy storage battery according to the hysteresis battery model and the target parameter set comprises: determining an initial open-circuit voltage of the energy storage battery according to the hysteresis battery model and the target parameter set; inputting the initial open-circuit voltage into a Pecht model to obtain the target state of charge of the energy storage battery; wherein the Pecht model is used to determine the target state of charge according to the initial open-circuit voltage and based on a correspondence between open-circuit voltages and states of charge of the energy storage battery under the hysteresis effect.

3. The method of claim 2, wherein, The inputting of the initial open-circuit voltage into the Pecht model to obtain the target state of charge of the energy storage battery comprises: inputting the initial open-circuit voltage into the Pecht model to obtain a first state of charge, the first state of charge being determined by the Pecht model according to the initial open-circuit voltage, a hysteresis state inside the Pecht model and a first weight, wherein the hysteresis state is a quantitative description of a current hysteresis effect of the energy storage battery, and the first weight represents an influence degree of respective regions in a mathematical model of the energy storage battery on the state of charge; updating the first weight according to the first state of charge to obtain a second weight; re-inputting the initial open-circuit voltage into the Pecht model to obtain a second state of charge, the second state of charge being determined by the Pecht model according to the initial open-circuit voltage, the hysteresis state inside the Pecht model and the second weight, and the second state of charge being the target state of charge.

4. The method of claim 3, wherein, The updating of the first weight according to the first state of charge to obtain a second weight comprises: determining a theoretical output current of the energy storage battery according to the first state of charge and the hysteresis battery model; determining a current error according to the theoretical output current and a real output current corresponding to the energy storage battery; updating the first weight according to the current error to obtain the second weight.

5. The method of claim 4, wherein, The updating of the first weight according to the current error to obtain the second weight comprises: updating the first weight according to the current error to obtain the second weight by the following formula: wherein, represents the sampled time, represents the sampled time, represents a gain coefficient, represents a current error, represents a weight, represents an updated weight, represents a hysteresis state.

6. The method of claim 2, wherein, The determining of the initial open-circuit voltage of the energy storage battery according to the hysteresis battery model and the target parameter set comprises: assigning values of respective parameters in the hysteresis battery model according to values of the respective parameters in the target parameter set; determining the initial open-circuit voltage of the energy storage battery according to the hysteresis battery model after the assignment.

7. The method of claim 6, wherein, The hysteresis battery model comprises a voltage source, an ohmic internal resistance, a resistance-capacitance network and a hysteresis module, the resistance-capacitance network comprises a capacitor and a resistor, and the hysteresis battery model is expressed as follows: wherein, represents the nth sample point, represents the attenuation factor, represents the sampling interval, represents the voltage of the hysteresis module, represents the current, represents the maximum hysteresis voltage, represents the capacitance in the RC network, represents the resistance in the RC network, represents the voltage of the capacitance , represents the time constant, = , represents the ohmic internal resistance, represents the open circuit voltage of the voltage source corresponding to the SOC, represents the total terminal voltage between the positive and negative electrodes of the energy storage battery.

8. The method according to any one of claims 1 to 7, characterized in that, The plurality of parameter sets are determined in the following manner: The numerical range of the state of charge of the energy storage battery is divided into a plurality of numerical intervals; For the value of each parameter in the hysteresis battery model, at least one parameter optimization is performed on each numerical interval by using a particle swarm optimization algorithm to obtain an optimal parameter set corresponding to each numerical interval; The optimal parameter set corresponding to each numerical interval is determined as the plurality of parameter sets.

9. A battery state of charge estimation device based on a hysteresis cell model, characterized by, Comprise: An acquisition module is configured to acquire an initial state of charge of an energy storage battery; A first determination module is configured to determine, from a plurality of parameter sets, a target parameter set matching a numerical interval in which the initial state of charge is located, the target parameter set comprising values of each parameter in a hysteresis battery model, the hysteresis battery model being used to simulate a hysteresis effect of the energy storage battery and compensate for an error caused by the hysteresis effect; A second determination module is configured to determine, according to the hysteresis battery model and the target parameter set, a target state of charge of the energy storage battery.

10. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to implement the battery state of charge estimation method based on the hysteresis battery model according to any one of claims 1 to 8.

Images