Method and apparatus for real-time estimation of battery state of charge

Abstract

The present disclosure provides a real-time estimation method and device for battery state of charge. The method comprises: obtaining state data in a running process of a target battery; determining an initial estimation value of the battery state of charge based on the state data; identifying a physical anchor point corresponding to a phase transition process of a battery material from the state data, the physical anchor point corresponding to a stable state of charge reference value; and correcting the initial estimation value based on the state of charge reference value to obtain a target state of charge when the physical anchor point is identified.

Description

Technical Field

[0001] This disclosure relates to the field of battery management technology, and more specifically to a method and apparatus for real-time estimation of battery state of charge. Background Technology

[0002] The State of Charge (SOC) of a lithium-ion battery is one of the core state parameters in a Battery Management System (BMS). A relatively accurate SOC value is a prerequisite for the safe, efficient, and reliable operation of the BMS, and real-time SOC estimation is crucial throughout the entire battery lifecycle. Therefore, improving the accuracy of real-time SOC estimation is extremely challenging. Summary of the Invention

[0003] In view of the above problems, this disclosure provides a method and apparatus for real-time estimation of battery state of charge.

[0004] According to a first aspect of this disclosure, a method for real-time estimation of battery state of charge is provided. The method includes: acquiring state data during the operation of a target battery; determining an initial estimate of the battery's state of charge based on the state data; identifying physical anchor points from the state data corresponding to phase transition processes of the battery materials, the physical anchor points corresponding to stable state of charge reference values; and, upon identification of a physical anchor point, correcting the initial estimate based on the state of charge reference value to obtain the target state of charge.

[0005] According to embodiments of this disclosure, identifying a physical anchor point from state data includes: acquiring capacity state data from the state data, the capacity state data indicating the relationship between battery capacity and voltage; and identifying a physical anchor point in response to determining that the difference between the capacity state data and the capacity state reference value corresponding to the physical anchor point is less than a first predetermined threshold.

[0006] According to embodiments of this disclosure, the method further includes: in response to determining that the difference between the capacity status data and the capacity status reference value corresponding to the physical anchor point is greater than or equal to a first predetermined threshold, and no physical anchor point is identified; determining the initial estimate as the target power status at the current moment.

[0007] According to embodiments of this disclosure, the method further includes: acquiring reference state data of a reference battery during cyclic aging, wherein the battery material of the reference battery is the same as that of the target battery; determining the anchor point charge state of the reference battery based on the reference state data; determining the anchor point charge state as a charge state reference value of the target battery; and determining the reference capacity state data corresponding to the anchor point charge state as a capacity state reference value corresponding to the physical anchor point of the target battery.

[0008] According to embodiments of this disclosure, the state of charge reference value includes one of the following: determined based on the phase change characteristics of the battery material; remains stable during the cyclic aging process of the target battery; has an offset within a preset temperature range less than a second predetermined threshold; and has multiple state of charge reference values, each corresponding to different capacity state data.

[0009] According to embodiments of this disclosure, when a physical anchor point is identified, an initial estimate is corrected based on a state of charge reference value to obtain a target state of charge. This includes: determining the battery capacity weight of the target battery when the physical anchor point is identified; weighting the initial estimate and the state of charge reference value based on the battery capacity weight to obtain the current state of charge of the target battery; and processing the current state of charge based on predetermined conditions to obtain the target state of charge of the target battery. The predetermined conditions restrict the target state of charge from decreasing over time during the discharge process of the target battery.

[0010] According to embodiments of this disclosure, processing the current state of charge based on predetermined conditions to obtain the target state of charge of the target battery includes: in response to determining that the target state of charge of the target battery at time t-1 is less than the current state of charge at time t, determining the target state of charge at time t-1 as the target state of charge of the target battery at time t; and in response to determining that the target state of charge of the target battery at time t-1 is greater than or equal to the current state of charge at time t, determining the current state of charge at time t as the target state of charge of the target battery.

[0011] According to embodiments of this disclosure, determining the battery capacity weight of a target battery when a physical anchor point is identified includes: processing capacity state data and capacity state reference values ​​to obtain the battery capacity weight of the target battery at the current moment.

[0012] According to embodiments of this disclosure, determining an initial estimate of battery capacity based on state data includes performing the following operations using a trained target model. The target model is used to establish a nonlinear mapping relationship between state data and battery capacity. The state data includes operational data related to battery aging during the operation of the target battery: extracting multidimensional feature data from the state data; fusing the multidimensional feature data to obtain fused features; wherein the fused features indicate operational features and aging features associated with the target capacity state; and evaluating the battery capacity using the fused features to obtain an initial estimate.

[0013] According to a second aspect of this disclosure, an apparatus for real-time estimation of battery state of charge is provided. The apparatus includes an acquisition module, a prediction module, an identification module, and a correction module. The acquisition module acquires state data during battery operation. The prediction module determines an initial estimate of the battery's state of charge based on the state data. The identification module identifies physical anchor points corresponding to phase transition processes of battery materials from the state data; these physical anchor points correspond to stable state of charge reference values. The correction module corrects the initial estimate based on the state of charge reference values ​​when a physical anchor point is identified, thereby obtaining a target state of charge.

[0014] According to embodiments of this disclosure, a real-time estimation method for battery state of charge (SOC) is provided. Stable SOC reference values ​​during the phase transition of battery materials during battery operation are mapped as physical anchor points. An initial estimate of the SOC is determined based on the battery's operating state data. Upon identification of the physical anchor point, the initial estimate is corrected based on the stable SOC reference value to obtain the target SOC. Since the physical anchor points are set according to the phase transition characteristics of the battery materials and are unaffected by the degree of battery aging, correcting the initial estimate based on the physical anchor points can reduce the error in the real-time estimated target SOC, thereby improving the stability of battery operation. Attached Figure Description

[0015] The foregoing contents, as well as other objects, features, and advantages of this disclosure, will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0016] Figure 1 This diagram illustrates the relationship between battery voltage and state of charge at different cycle numbers.

[0017] Figure 2 A partially enlarged view of the relationship between battery voltage and state of charge at different cycle numbers is shown schematically.

[0018] Figure 3 This diagram illustrates the relationship between battery voltage and state of charge at different temperatures.

[0019] Figure 4 An exemplary system architecture is illustrated in the diagram, showing an application of the real-time battery capacity estimation method of this disclosure.

[0020] Figure 5 A flowchart illustrating a real-time battery capacity estimation method according to an embodiment of the present disclosure is shown schematically.

[0021] Figure 6 A flowchart illustrating a power estimation method for a trained target model according to an embodiment of the present disclosure is shown schematically.

[0022] Figure 7This illustration schematically shows the voltage variation of a target battery with the number of cycles according to an embodiment of the present disclosure;

[0023] Figure 8 This illustration schematically shows the change in the state of charge of a target battery as a function of the number of cycles according to an embodiment of the present disclosure;

[0024] Figure 9 This illustration schematically shows the variation of the capacity state data of a target battery with the number of cycles according to an embodiment of the present disclosure;

[0025] Figure 10 This illustration schematically shows the variation of the area of ​​the capacity differential curve of a target battery with the number of cycles according to an embodiment of the present disclosure;

[0026] Figure 11 This illustration schematically shows the variation of the full-charge capacity of a target battery with the number of cycles according to an embodiment of the present disclosure;

[0027] Figure 12 This illustration schematically demonstrates the correlation analysis between state data in embodiments of the present disclosure;

[0028] Figure 13 This illustration shows the contribution of state data to power state prediction in an embodiment of the present disclosure;

[0029] Figure 14 A schematic diagram of the structure of a target model according to an embodiment of the present disclosure is shown.

[0030] Figure 15 A flowchart illustrating a method for calibrating physical anchor points according to an embodiment of the present disclosure is shown schematically.

[0031] Figure 16 This illustration schematically shows a diagram of interpolating the state of charge of a reference battery according to an embodiment of the present disclosure;

[0032] Figure 17 The coefficient of variation curve of a reference battery according to an embodiment of the present disclosure is illustrated schematically;

[0033] Figure 18 A schematic diagram illustrating the relationship between capacity status reference values ​​and cycle number according to embodiments of the present disclosure is shown.

[0034] Figure 19 A schematic diagram illustrating the relationship between capacity status reference values ​​and frequency according to embodiments of the present disclosure is shown.

[0035] Figure 20 The diagram illustrates the relationship between capacity differential and state of charge at different temperatures and cycle numbers according to embodiments of the present disclosure.

[0036] Figure 21A partially enlarged view of the relationship curves between capacity differential and state of charge at different temperatures and cycle numbers, according to embodiments of the present disclosure, is shown schematically.

[0037] Figure 22 A schematic diagram illustrating the comparison between the state of charge estimation curve of this disclosure embodiment and the state of charge estimation curve in related examples is shown.

[0038] Figure 23 The diagram schematically illustrates an error comparison between the state of charge estimation curve of this disclosure embodiment and the state of charge estimation curve in a related example;

[0039] Figure 24 This schematic diagram illustrates a structural block diagram of an apparatus for real-time estimation of battery state of charge according to an embodiment of the present disclosure;

[0040] Figure 25 A block diagram of an electronic device suitable for implementing a real-time battery power estimation method according to an embodiment of the present disclosure is shown schematically. Detailed Implementation

[0041] To make the objectives, technical solutions and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0042] The endpoints and any values ​​of the ranges disclosed in this disclosure are not limited to the precise ranges or values, and such ranges or values ​​should be understood to include values ​​close to such ranges or values. For numerical ranges, the endpoint values ​​of the various ranges, the endpoint values ​​of the various ranges and individual point values, and individual point values ​​can be combined with each other to obtain one or more new numerical ranges, which should be regarded as specifically disclosed in this disclosure.

[0043] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0044] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0045] In the description of this disclosure, it should be understood that the terms "longitudinal", "length", "circumferential", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this disclosure and simplifying the description, and do not indicate or imply that the subsystem or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this disclosure.

[0046] Similarly, to simplify this disclosure and aid in understanding one or more of the various aspects of the disclosure, in the above description of exemplary embodiments of the present disclosure, various features of the present disclosure are sometimes grouped together in a single embodiment, figure, or description thereof. The use of terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refers to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the present disclosure. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0047] The State of Charge (SOC) of a lithium-ion battery is one of the core state parameters in a Battery Management System (BMS). SOC can be characterized as the ratio of the battery's remaining capacity to its rated capacity, usually expressed as a percentage. A relatively accurate SOC value is a prerequisite for the safe, efficient, and reliable operation of the BMS, and real-time SOC estimation is crucial throughout all stages of the battery's lifecycle. Therefore, improving the accuracy of real-time SOC estimation is extremely challenging.

[0048] For example, silicon anodes for lithium-ion batteries have become one of the core candidate materials for next-generation high-energy-density batteries due to their high theoretical specific capacity (4200 mAh / g). However, silicon materials exhibit significant charge drift issues under different temperature environments (e.g., <10℃ or >40℃). This is because the volume expansion effect of silicon materials intensifies at lower or higher temperatures, leading to deterioration of lithium-ion diffusion kinetics and abnormal accumulation of polarization voltage. This significantly increases the error (typically exceeding 10%) of traditional SOC estimation methods (such as the ampere-hour integration method and open-circuit voltage method), thereby severely affecting the stable operation of the battery and the accuracy of range prediction.

[0049] In related examples, model-based methods (such as Kalman filtering) rely on relatively accurate equivalent circuit models, but the model parameters are easily mismatched due to the low-temperature phase transition of silicon. Machine learning methods are based on training with raw voltage and current data, which requires massive low-temperature datasets and lacks generalization ability for the unique non-ideal electrochemical properties of silicon. Specifically:

[0050] Method 1: The ampere-hour integration method is based on the law of charge conservation. It calculates the battery state of charge (SOC) by measuring the charging and discharging current in real time and integrating it over time, combined with the initial charge, as shown in formula (1).

[0051] SOC(t)=SOC(0)+(1 / Cn)×∫I(t)dt (1)

[0052] In formula (1), SOC(t) is the state of charge of the battery at time t, SOC(0) is the initial state of charge of the battery at time 0, Cn is the rated capacity of the battery, and I(t) is the real-time current. The accuracy of the SOC estimate obtained by the above method depends on the accuracy of the initial SOC and the rated capacity of the battery. Under low-temperature conditions, it is difficult to obtain an accurate prediction of the initial SOC of silicon anode batteries, which cannot reflect the actual changes in the internal electrochemical state of the battery, and the error in the state of charge of the battery will continue to accumulate.

[0053] Method 2: The open-circuit voltage method utilizes the inherent correlation between the battery's open-circuit voltage (OCV) and its state of charge (SOC). By measuring the open-circuit voltage after the battery has been left to rest, and then referring to a pre-calibrated OCV-SOC curve, the battery's SOC at the current moment can be obtained. This method requires the battery to be left to rest for an extended period (usually several hours) to eliminate polarization effects and obtain a relatively stable OCV. At low temperatures, the polarization effect of silicon anode batteries is significant and dissipates slowly, making it difficult to reach a stable state in a short time. This results in a large measurement error for OCV, failing to reflect changes in SOC in real time. Furthermore, the OCV-SOC curve is prone to shift at low temperatures, further increasing the estimation error.

[0054] Figure 1 This diagram illustrates the relationship between battery voltage and state of charge at different cycle numbers. Figure 2 A partially enlarged view of the relationship between battery voltage and state of charge at different cycle numbers is shown schematically. Figure 3 The diagram illustrates the relationship between battery voltage and state of charge at different temperatures. It should be noted that... Figure 1 Different cycle counts can characterize different aging states of the battery. Figure 2 The image shows a magnified view of the battery level in the range of 50% to 70%. Figure 3The temperature conditions shown are 0℃, 10℃ and 25℃, and the battery discharge rate is 0.2C.

[0055] like Figures 1-3 As shown, under the same battery voltage, the state of charge of the battery varies significantly due to differences in temperature and aging. Figure 2 For example, when the battery voltage is 3.95V, the battery cycles 100 times at 25℃ (e.g.) Figure 2 When the blue curve in the diagram indicates that the battery's state of charge is 60.5%, the battery has undergone 600 cycles at 40°C (as shown in the diagram). Figure 2 (The gray curve in the diagram illustrates this). The battery's state of charge is 57.3%. The state of charge estimated by the above method will introduce an error of more than 3%, and the error will increase with differences in temperature and aging.

[0056] Therefore, the state of charge estimation methods in the relevant examples are difficult to achieve in real time, or have insufficient adaptability to silicon anode batteries and silicon-carbon anode batteries, or have difficulty handling the effects of aging degree and temperature differences.

[0057] In view of this, embodiments of the present disclosure provide a real-time battery power estimation method that can combine physical laws and battery operating data to estimate battery power. This method is independent of the degree of battery aging and can improve the real-time estimation accuracy in the low SOC region. It is particularly suitable for power management of silicon-carbon anode batteries.

[0058] Figure 4 An exemplary system architecture is schematically illustrated in which the real-time battery capacity estimation method of this disclosure can be applied. It should be noted that... Figure 4 The examples shown are merely examples of system architectures that can be applied to the embodiments of this disclosure, in order to help those skilled in the art understand the technical content of this disclosure, but do not mean that the embodiments of this disclosure cannot be used in other devices, systems, environments or scenarios.

[0059] like Figure 4 As shown, the system architecture of the real-time battery power estimation method disclosed herein may include a data acquisition module, a dQ / dV calculation module, a feature extraction module, an AI prediction module, and a physical anchor point recognition module. Figure 4 The module includes a convergence point detection module, a physical correction module, a monotonicity constraint module, and an output module.

[0060] The data acquisition module uses voltage sensors, current sensors, and temperature sensors to acquire battery status data during operation, such as voltage data, current data, and temperature data.

[0061] The dQ / dV calculation module can calculate the differential value of the battery capacity based on the state data, and the physical anchor point identification module can identify the physical anchor point based on the differential value of the capacity.

[0062] The feature extraction module can extract feature data related to the degree of battery aging based on the battery's state data and the capacity differential value calculated by the dQ / dV calculation module.

[0063] AI prediction modules can utilize features such as MLP neural networks to predict an initial estimate of the battery's current charge level based on feature data related to the degree of battery aging.

[0064] The physical correction module and the monotonicity constraint module can correct the initial estimate and impose monotonicity constraints using the corresponding state of charge reference value after the physical anchor point is identified by the physical anchor point identification module. The output module then outputs the target state of charge (SOC). corrected .

[0065] Figure 5 A flowchart illustrating a real-time battery power estimation method according to an embodiment of the present disclosure is shown.

[0066] like Figure 5 As shown, the estimation method of this embodiment may include operations S510 to S540.

[0067] The S510 is used to acquire status data during battery operation.

[0068] In some embodiments, the battery operation process may include a battery discharge process or a battery charging process. The state data during battery operation may include voltage data, current data, temperature data, etc. The acquisition frequency of the state data may be 1Hz to 10Hz, for example, 1Hz, 2Hz, 3Hz, 4Hz, 5Hz, 6Hz, 7Hz, 8Hz, 9Hz, 10Hz, etc.

[0069] In operation S520, an initial estimate of battery capacity is determined based on status data.

[0070] In some embodiments, the initial estimate can be predicted using a pre-trained model. The pre-trained model can include neural network models, such as multilayer perceptron neural network models.

[0071] In operation S530, physical anchor points corresponding to the phase transition process of battery materials are identified from the state data. These physical anchor points correspond to stable state of charge reference values.

[0072] In some embodiments, the physical anchor point corresponds to a stable state of charge reference value during the phase transition process of the battery material.

[0073] For example, the phase transition boundary where the O3 phase completely transforms into the O1 phase during the delithiation process of lithium cobalt oxide (LiCoO2, LCO) cathode material, or the phase transition boundary where the silicon material completes delithiation in silicon-carbon (Si-C) anode, etc. The physical anchor point can be specifically set according to the positive and negative electrode materials in the battery; the physical anchor points of other electrode materials are not described in this disclosure.

[0074] When operating S540, upon identifying a physical anchor point, the initial estimate is corrected based on the power status reference value to obtain the target power status.

[0075] According to embodiments of this disclosure, a real-time estimation method for battery state of charge (SOC) is provided. Stable SOC reference values ​​during the phase transition of battery materials during battery operation are mapped as physical anchor points. An initial estimate of the SOC is determined based on the battery's operating state data. Upon identification of the physical anchor point, the initial estimate is corrected based on the stable SOC reference value to obtain the target SOC. Since the physical anchor points are set according to the phase transition characteristics of the battery materials and are unaffected by the degree of battery aging, correcting the initial estimate based on the physical anchor points can reduce the error in the real-time estimated target SOC, thereby improving the stability of battery operation.

[0076] In some embodiments, during the real-time acquisition of battery status data during operation, capacity status data can be obtained from the status data. Capacity status data can indicate the relationship between battery capacity and voltage changes, such as capacity derivatives.

[0077] For example, the capacity differential value can be calculated using a sliding window method based on voltage and current data, as shown in formula (2).

[0078] dQ / dV = (Q n+1 —Q n ) / (V n+1 —V n (2)

[0079] In formula (2), dQ / dV represents the capacity differential value, with units of mAh / V. n+1 Q represents the cumulative capacity of the (n+1)th measurement point. n V represents the cumulative capacity at the nth measurement point. n+1 V represents the voltage at the (n+1)th measurement point. n This represents the voltage at the nth measurement point. The cumulative capacity can be obtained by integrating the current data.

[0080] In some embodiments, if the difference between the acquired capacity status data and the capacity status reference value corresponding to the physical anchor point is less than a first predetermined threshold, it can be determined that a physical anchor point has been identified, as shown in formula (3).

[0081] (3)

[0082] In formula (3), The state-of-capacity data for the target battery can be the capacity differential value; This is a reference value for the state of capacity of the target battery; This is the first predetermined threshold.

[0083] For example, the difference between the capacity status data and the capacity status reference value can represent the absolute value of the difference. The first predetermined threshold can be set to 0~50mAh / V, such as 0, 10mAh / V, 20mAh / V, 25mAh / V, 30mAh / V, 40mAh / V, 50mAh / V, etc. For example, when the first predetermined threshold is 50mAh / V and the capacity status reference value is -1875mAh / V, a physical anchor point is identified when the capacity status data is -1850mAh / V.

[0084] According to the embodiments of this disclosure, the capacity status reference value corresponding to the physical anchor point is set to an interval. When the capacity status data of the battery falls into the interval, the battery can be regarded as having identified the physical anchor point. This can avoid the problems of weak anti-interference ability, poor robustness, and insufficient adaptability caused by the single value of the capacity status reference value corresponding to the physical anchor point in practical applications, and improve the accuracy of power estimation.

[0085] In some embodiments, if the difference between the acquired capacity status data and the capacity status reference value corresponding to the physical anchor point is greater than or equal to a first predetermined threshold, it can be determined that no physical anchor point has been identified.

[0086] For example, when the first predetermined threshold is 50mAh / V and the capacity status reference value is -1875mAh / V, if the capacity status data is -1820mAh / V, it is determined that no physical anchor point has been identified.

[0087] In a specific embodiment, if it is determined that no physical anchor point has been identified, the initial estimate of the battery charge can be determined as the target charge state at the current moment. For example, if the initial estimate of the battery charge is determined to be 13% based on the battery state data, and no physical anchor point has been identified, the target charge state of the battery at the current moment can be determined to be 13%.

[0088] According to embodiments of this disclosure, accurate identification of physical anchor points depends on specific phase transition processes. In the absence of battery physical anchor point identification conditions, specifying the initial predicted value as the target charge state can ensure the stability of the estimation method, improve the continuity and stability of real-time estimation, and take into account both the accuracy and robustness of the above estimation method.

[0089] In some embodiments, the initial estimate of battery charge can be predicted using a pre-trained target model. The pre-trained target model can include neural network models, such as a multilayer perceptron (MLP) neural network model, used to establish a nonlinear mapping between state data and battery charge. MLP neural network models excel at static regression, their multilayer structure can fit arbitrary nonlinear mappings, and they are computationally fast.

[0090] Figure 6 A flowchart illustrating a power estimation method for a trained target model according to an embodiment of the present disclosure is shown.

[0091] like Figure 6 As shown, the power estimation method for the target model may include operations S610 to S630.

[0092] In operation S610, multidimensional feature data is extracted from the status data.

[0093] In some embodiments, the state data includes operational data related to battery aging during the operation of the target battery, such as voltage V, capacity derivative dQ / dV, and full charge capacity (FCC) retention rate FCC. ret Voltage offset V shift Normalized differential value of capacity dQ / dV norm Cycle number of cycles normalized value norm wait.

[0094] When operating the S620, feature fusion is performed on multidimensional feature data to obtain fused features.

[0095] In some embodiments, the fused features indicate operating and aging features associated with the target electrical state.

[0096] During operation of S630, the battery capacity of the fused features is evaluated to obtain an initial estimate.

[0097] The following will combine Figures 7-14 The method for estimating the charge of the trained target model is further described.

[0098] Figure 7 This illustration schematically shows the voltage variation of a target battery with the number of cycles according to an embodiment of the present disclosure; Figure 8 This illustration schematically shows the change in the state of charge of a target battery as a function of the number of cycles according to an embodiment of the present disclosure; Figure 9 This illustration schematically shows the variation of the capacity state data of a target battery with the number of cycles according to an embodiment of the present disclosure; Figure 10This illustration schematically shows the variation of the area of ​​the capacity differential curve of a target battery with the number of cycles according to an embodiment of the present disclosure; Figure 11 The illustration schematically shows the variation of the full-charge capacity of a target battery according to an embodiment of the present disclosure with the number of battery cycles. It should be noted that... Figures 7-11 The horizontal axis represents the number of battery cycles, and the vertical axis represents voltage, state of charge, capacity derivative, area under the capacity derivative curve, and full charge capacity, respectively. Figure 12 This illustration schematically demonstrates the correlation analysis between state data in embodiments of the present disclosure; Figure 13 The illustration schematically shows the contribution of state data to power state prediction in an embodiment of this disclosure. It should be noted that... Figure 12 and Figure 13 The status data includes the current voltage V, the capacity differential dQ / dV, and the full charge capacity (FCC) retention rate FCC. ret Voltage offset V shift Normalized differential value of capacity dQ / dV norm Cycle number of cycles normalized value norm .

[0099] like Figures 7-13 As shown, correlation analysis was performed on a large amount of state data during battery operation to obtain the variation pattern of state data with the number of battery cycles, i.e., the correlation pattern with the degree of battery aging. Before performing the correlation analysis, outliers in the state data can be removed to obtain valid data.

[0100] Based on the correlation analysis results above, multidimensional feature data can be extracted from the state data. Multidimensional feature data can include an M-dimensional feature vector X = [x1, x2, ..., x...]. M This includes the current voltage V, the capacity differential dQ / dV, and the full charge capacity (FCC) retention rate FCC. ret Voltage offset V shift Normalized differential value of capacity dQ / dV norm Cycle number of cycles normalized value norm The extraction of multidimensional feature data can be determined based on the type of target battery and application requirements, but this disclosure is not limited thereto.

[0101] Figure 14 A schematic diagram of the structure of a target model according to an embodiment of the present disclosure is shown.

[0102] like Figure 14 As shown, multidimensional feature data is input into the pre-trained target model. The target model can include an input layer, multiple hidden layers, and an output layer. The input layer can have M neurons to receive M-dimensional feature vectors. Figure 14 The example only shows a neural network structure with 6 neurons, such as V, dQ / dV, and FCC.ret V shift ,dQdV norm Cycle norm This disclosure does not impose specific limitations. The hidden layers can be L fully connected layers, each containing H neurons. The activation function can be ReLU. For example, hidden layer 1 can contain 128 neurons; hidden layer 2 can contain 64 neurons; and hidden layer 3 can contain 32 neurons. The output of each layer = activation function × (output of the previous layer × weight matrix + bias vector). The number of hidden layers can be set according to specific needs, and this disclosure does not impose specific limitations. The output layer can contain one neuron, used to output the initial estimated value SOC. raw .

[0103] By using multiple hidden layers, the limited expressive power of a single hidden layer for complex nonlinear relationships can be addressed. The first hidden layer is a low-level feature combination; the six original feature dimensions are insufficient to express complex nonlinear mappings. Extending the 6-dimensional feature vector to 128 dimensions creates a richer feature combination space. Examples include V×dQ / dV and Cycle. norm ×FCC ret Implicit features are used. Higher-level hidden layers can abstract high-level features to remove redundant information, better capture complex mapping relationships, and retain the features most relevant to battery state. The output layer then fuses all abstract features into an initial estimate. The entire process requires only matrix operations and is easy to deploy.

[0104] During the training of the target model, training state data of a reference battery under multi-cycle aging tests and the corresponding power state label sequences can be obtained. For example, the training state data can include state data from 700 cycles. Training state data from cycles 100 to 600, along with the power state label sequences, can be used as training data (approximately 500,000 samples); training state data from cycle 700, along with the power state label sequences, can be used as test data (approximately 6,000 samples). The training state data is input into the initial model, which outputs a sequence of power state evaluation values. Based on the loss value between the power state evaluation value sequence and the power state label sequence, the model parameters of the initial model are adjusted to obtain the trained target model described above.

[0105] In a specific implementation, the optimizer for the target model can be the Adam (Adaptive Moment Estimation) optimizer, with a learning rate ranging from 0.001. The batch size for training can be 256, and the number of training epochs can be 100. The loss function can be the Mean-Square Error (MSE) function, etc.

[0106] Figure 15A flowchart illustrating a method for calibrating physical anchor points according to an embodiment of the present disclosure is shown schematically.

[0107] like Figure 15 As shown, the physical anchor point calibration method in this embodiment may include operations S1510~S1530.

[0108] In operation S1510, reference state data of the reference battery during the cycle aging process is obtained.

[0109] In some embodiments, the reference battery can be a battery with the same battery materials as the target battery. Reference state data can be multi-cycle aging test data of the reference battery under standard conditions. Standard conditions may include the test temperature, discharge rate, etc., for the cycle aging test. The test temperature can be set to 25°C, and the discharge rate can be set to 0.5C, etc. Multi-cycle aging test data can include the correspondence between the battery's capacity derivative and state of charge during multiple discharge cycles.

[0110] In operation S1520, the anchor point charge state of the reference battery is determined based on the reference state data.

[0111] For ease of explanation, it can be combined with Figure 16 and Figure 17 The above-mentioned method for calibrating physical anchor points is further described.

[0112] Figure 16 This illustration schematically shows a diagram of interpolating the state of charge of a reference battery according to an embodiment of the present disclosure; Figure 17 The coefficient of variation curve of a reference battery according to an embodiment of the present disclosure is illustrated schematically.

[0113] like Figure 16 As shown, the correspondence between the battery's capacity differential and state of charge (SOC) can be selected within a predetermined range in the multi-cycle aging test data. The predetermined range can be a low SOC region with a SOC of less than 25%. Within this predetermined range, high-resolution interpolation processing can be performed on the battery's SOC during each discharge cycle. The interpolation resolution can be set according to actual needs, for example, to 0.1% SOC.

[0114] like Figure 17 As shown, the coefficient of variation of each SOC interpolation point can be calculated within a predetermined range, as shown in formula (3).

[0115] (4)

[0116] In formula (4), CV(SOC) i Let be the coefficient of variation for the i-th SOC interpolation point. Let be the standard deviation of the capacity derivative of the i-th SOC interpolation point in each cycle. Let be the mean of the capacity derivative of the i-th SOC interpolation point in each cycle.

[0117] Subsequently, local minima in the coefficient of variation curve can be identified as physical anchor points. (Continue to refer to...) Figure 17 The coefficients of variation for the local minimum points in the curve can be 1.48%, 1.40%, and 3.34%, as shown in Table 1.

[0118] Table 1

[0119] SOC (%) CV (%) dQ / dV mean dQ / dV standard deviation 12.00 1.40 -1875.10 24.83 4.20 1.48 -992.01 14.12 22.30 3.34 -4431.50 147.94

[0120] Depending on the specific application of the battery, a point with a coefficient of variation of less than 2% can be selected as a physical anchor point. The corresponding anchor point state of charge can be 4.2% and 12.0%, respectively. The anchor point state of charge of the reference battery is determined by processing the state of charge.

[0121] In operation S1530, the anchor point's charge status is determined as the target battery's charge status reference value, and the reference capacity status data corresponding to the anchor point's charge status is determined as the capacity status reference value corresponding to the target battery's physical anchor point.

[0122] In a specific embodiment, the anchor point state of charge of the reference battery can be determined as the reference value for the state of charge of the target battery. For example, the reference value for the state of charge of the target battery could be 4.2%, 12.0%, etc. The reference capacity state data corresponding to the anchor point state of charge can be determined as the reference capacity state value corresponding to the physical anchor point of the target battery. The reference capacity state data can be the average of the capacity derivative of the anchor point state of charge in each cycle.

[0123] Figure 18 A schematic diagram illustrating the relationship between capacity status reference values ​​and cycle number according to embodiments of the present disclosure is shown. Figure 19 A schematic diagram illustrating the relationship between capacity status reference values ​​and frequency according to embodiments of the present disclosure is provided. It should be noted that... Figure 18 The horizontal axis represents the number of cycles, and the vertical axis represents the capacity status reference value. Figure 19 The horizontal axis represents the capacity status reference value, and the vertical axis represents the frequency. Among these, Figure 19 The capacity state reference value μ and the Gaussian kernel width parameter σ are shown in the figure.

[0124] like Figure 18 and Figure 19 As shown, the capacity state reference value corresponding to the physical anchor point does not change with the number of cycles, and the capacity state reference value corresponding to the physical anchor point is a narrow distribution-high convergence state.

[0125] Taking a reference battery with lithium cobalt oxide as the positive electrode material and silicon-carbon as the negative electrode material as an example, the electrochemical physical mechanism of the physical anchor point of the reference battery can be explained as follows.

[0126] At an anchor charge state of 12%, lithium cobalt oxide is at the phase transition boundary of complete delithiation. Lithium cobalt oxide in the delithiation reaction can be represented as Li x CoO2, where x is the lithium content. Li x The phase transformation of CoO2 under different lithium contents x is shown in Table 2.

[0127] Table 2

[0128] Lithium content x Crystal phase Battery State of Charge (SOC) 1.0>x O3 phase (hexagonal) Fully charged x≈0.75 O3 phase → O3' phase Order-disorder transition x≈0.5 Monoclinic phase appearance SOC≈50% x<0.5 H1-3 phases Low SOC region x≈0.12 O1 phase boundary SOC≈12%

[0129] When x ≈ 0.12, lithium cobalt oxide is close to a fully delithiated state, and the crystal phase is in the boundary region of the O3 phase → O1 phase transition. Here, the slope of the voltage curve changes drastically, producing a stable capacity differential characteristic peak. Since the phase transition is determined by the crystal energy, it is a thermodynamic intrinsic property of the material and is independent of the number of cyclic aging cycles.

[0130] When the state of charge at the anchor point is between 4% and 5%, the silicon component in the silicon-carbon composite anode is undergoing delithiation. The lithium intercalation phase transition of silicon is shown in Table 3.

[0131] Table 3

[0132] Phase transition Voltage (vs Li) illustrate Si→LiSi ~0.4V Initial lithium intercalation <![CDATA[→Li 12 Si7]]> ~0.3V intermediate phase <![CDATA[→Li 15 Si4]]> ~0.1V lithium-rich phase <![CDATA[→Li 22 Si5]]> ~0.05V final phase

[0133] When the battery is deeply discharged (SOC < 10%), carbon materials such as graphite have basically completed delithiation (LiC6 → C), and the lithium content is mainly derived from silicon materials. 22 The Si5→Si delithiation process contributes to battery capacity. The delithiation voltage plateau of silicon materials is very stable. Even if silicon particles pulverize due to volume expansion, the phase transition voltage is still determined by the thermodynamic properties of the Li-Si alloy and can remain essentially unchanged.

[0134] According to embodiments of this disclosure, the calibrated physical anchor point is determined based on the phase transition process of the battery material. The phase transition is determined by the crystal structure, and the phase transition voltage depends on the crystal free energy. Therefore, the phase transition is a thermodynamic intrinsic property of the battery material and is independent of the number of cycles. Furthermore, the phase transition occurs at a fixed lithium content x, rather than at the battery's absolute capacity, and is unaffected by the degree of battery aging. Simultaneously, even if the silicon particles in the negative electrode pulverize due to battery aging, the thermodynamic properties of the Li-Si phase transition remain unchanged.

[0135] Figure 20 The diagram illustrates the relationship between capacity differential and state of charge at different temperatures and cycle numbers according to embodiments of the present disclosure. Figure 21A partially enlarged diagram illustrating the relationship between capacity differential and state of charge curves under different temperatures and cycle numbers according to embodiments of the present disclosure is shown. It should be noted that... Figure 21 The range of the locally magnified charge state is 0~15%.

[0136] like Figure 20 and Figure 21 As shown, the physical anchor point of the battery varies with different cycle numbers. Figure 20 and Figure 21 The diagram shows convergence point 1 and convergence point 2. The state of charge (SCC) reference values ​​corresponding to these convergence points can remain stable during the cyclic aging process of the target battery. The target battery can have multiple SCC reference values, each corresponding to different capacity state data.

[0137] Continue to refer to Figure 20 and Figure 21 Since the phase-change voltage is less sensitive to temperature, the deviation of the state of charge reference value within the preset temperature range can be less than a second predetermined threshold. For example, the preset temperature range can be set to -10℃ to 45℃, and the second predetermined threshold can be set to 0.5% SOC.

[0138] Furthermore, the aforementioned physical anchor point is triggered every time the battery passes through the SOC region during discharge, without the need to set additional test conditions.

[0139] Based on this, the state of charge reference value is determined based on the phase change characteristics of the battery material and is not affected by the aging degree of the battery or the operating temperature. Therefore, the accuracy of the target state of charge corrected based on the state of charge reference value can be improved.

[0140] Next, upon identifying the physical anchor point, the initial estimate can be corrected based on the state of charge reference value to obtain the target state of charge of the target battery.

[0141] First, the battery capacity weight of the target battery when the physical anchor point is identified can be determined.

[0142] In some embodiments, the battery capacity weight of the target battery at the current moment can be obtained by processing the capacity state data and the capacity state reference value. For example, the battery capacity weight can be a confidence weight, as shown in formula (5).

[0143] w=exp(-|dQ / dV-dQ / dV anchor |² / (2σ²)) (5)

[0144] In formula (5), w is the battery capacity weight of the target battery; dQ / dV is the capacity state data of the target battery, which can be the capacity differential value; dQ / dV anchorThis is a reference value for the state of capacity of the target battery; |dQ / dV - dQ / dV anchor | represents the dynamic distance between the capacity status data and the capacity status reference value; σ is the Gaussian kernel width parameter, which can range from 10mAh / V to 100mAh / V.

[0145] For example, the current capacity differential value of the target battery is dQ / dV = -1850mAh / V, and the capacity state reference value corresponding to the physical anchor point is dQ / dV. anchor The value is -1875 mAh / V, the dynamic distance between the two is 25 mAh / V, the Gaussian kernel width parameter σ is 50 mAh / V, and the battery capacity weight w is 0.88.

[0146] Next, the initial estimate and the state of charge reference value can be weighted based on the battery capacity weight to obtain the current state of charge of the target battery.

[0147] In some embodiments, the current state of charge of the target battery is obtained by weighting the initial estimate and the state of charge reference value, as shown in formula (6).

[0148] SOC corrected =w×SOC anchor +(1-w)×SOC raw (6)

[0149] In formula (6), SOC corrected The target battery's current state of charge (SOC) is represented by w, where w is the battery capacity weight. anchor The State of Charge (SOC) is the reference value for the target battery's state of charge. raw This is an initial estimate of the battery capacity.

[0150] For example, assume an initial estimate of the battery capacity at SOC. raw The target battery physical anchor point corresponds to a State of Charge (SOC) of 13.5%. anchor Given a weighting of 12% and a battery capacity weight w of 0.88, the current state of charge (SOC) of the target battery is determined. corrected It is 12.18%.

[0151] Finally, the current battery state can be processed based on predetermined conditions to obtain the target battery state.

[0152] In some embodiments, predetermined conditions can limit the decrease of the target state of charge over time during the discharge of the target battery. During the discharge of the target battery, the battery state of charge should follow a monotonic decreasing trend over time.

[0153] In some embodiments, if it is determined that the target charge state of the target battery at time t-1 is less than the current charge state at time t, the target charge state at time t-1 is determined as the target charge state of the target battery at time t.

[0154] For example, if SOC corrected If SOC{t-1} is greater than or equal to SOC{t-1}, then SOC{t} is forcibly set to SOC{t-1}. Here, SOC{t} represents the target state of charge of the target battery at time t, and SOC{t-1} represents the target state of charge of the target battery at time t-1. corrected This refers to the current state of charge of the target battery. For example, the current state of charge (SOC). corrected If the target energy state SOC{t-1} at time t-1 is 12.18%, then the target energy state SOC{t} at time t is 12.0%.

[0155] In other embodiments, when the target charge state of the target battery is determined to be greater than or equal to the current charge state at time t-1, the current charge state at time t is determined as the target charge state of the target battery.

[0156] For example, if SOC corrected If SOC{t-1} ≤ SOC, then SOC{t} = SOC corrected Where SOC{t} represents the target state of charge of the target battery at time t, and SOC{t-1} represents the target state of charge of the target battery at time t-1. corrected This refers to the current state of charge (SOC) of the target battery. For example, the current state of charge (SOC). corrected If the target energy state SOC{t-1} at time t-1 is 12.18%, then the target energy state SOC{t} at time t is 12.18%.

[0157] According to embodiments of this disclosure, correcting the initial estimate based on the state-of-charge reference value corresponding to the physical anchor point can alleviate the shortcomings of traditional machine learning methods in terms of insufficient generalization for battery power estimation. The state-of-charge reference value is not constrained by battery aging or temperature, thereby reducing errors in battery power estimation and improving the accuracy of power assessment. Furthermore, applying a monotonicity constraint to the target state-of-charge can further enhance the stability of the prediction.

[0158] Figure 22 A schematic diagram illustrating the comparison between the state of charge estimation curve of this disclosure embodiment and the state of charge estimation curve in related examples is shown. Figure 23 The diagram schematically illustrates an error comparison between the state of charge estimation curve of an embodiment of this disclosure and the state of charge estimation curve in a related example.

[0159] like Figure 22 and Figure 23 As shown, the root mean square error (RMSE) of the state of charge (SOC) estimate in this embodiment is 1.98%, while the RMS error of the SOC estimate obtained by the ampere-hour integration method in related examples is 3.43%. Compared to the methods in related examples, this embodiment significantly reduces the error of the SOC estimate in the low SOC region, reducing the error from 7% to 1.5%, a reduction of up to 78%, making it particularly suitable for power management of silicon-carbon anode batteries.

[0160] According to embodiments of this disclosure, the above method can accurately capture multi-step lithium intercalation reactions and effectively analyze the electrochemical behavior of silicon-carbon anode batteries without calibrating battery temperature and aging conditions. In low-temperature scenarios, by utilizing the thermodynamic stability of physical anchor points, the battery capacity differential value is used to improve the accuracy of online capacity assessment. This method is adaptable to complex operating conditions, has low computational resource requirements, and can run on conventional BMS hardware, achieving accurate calibration across all scenarios, such as tablets, smartphones, and laptops.

[0161] In practical applications, it can also be used for battery health diagnosis and silicon content estimation. A shift in the physical anchor point indicates a change in the battery material structure, which can be used to assist in detecting abnormal battery aging or other safety hazards. In silicon-carbon / lithium cobalt oxide batteries, the peak intensity of the capacity differential at 5% state of charge is related to the silicon content, and can be used to infer the actual silicon doping ratio, differences between different battery systems, and differences between batteries from different manufacturers.

[0162] Based on the above-described train operation control method, this disclosure also provides a device for real-time estimation of battery state of charge. The following will be combined with... Figure 24 The device is described in detail.

[0163] Figure 24 A schematic block diagram of a device for real-time estimation of battery state of charge according to an embodiment of the present disclosure is shown.

[0164] like Figure 24 As shown, the device 2400 for real-time estimation of battery power status in this embodiment may include an acquisition module 2410, a prediction module 2420, an identification module 2430, and a correction module 2440.

[0165] The acquisition module 2410 is used to acquire state data during battery operation. In one embodiment, the acquisition module 2410 can be used to perform the operation S510 described above, which will not be repeated here.

[0166] The prediction module 2420 is used to determine an initial estimate of the battery charge based on the state data. In one embodiment, the prediction module 2420 can be used to perform the operation S520 described above, which will not be repeated here.

[0167] The identification module 2430 is used to identify physical anchor points corresponding to the phase transition process of battery materials from the state data. These physical anchor points correspond to stable state of charge reference values. In one embodiment, the identification module 2430 can be used to perform the operation S530 described above, which will not be repeated here.

[0168] The correction module 2440 is used to correct the initial estimate based on the power status reference value when a physical anchor point is identified, so as to obtain the target power status. In one embodiment, the correction module 2440 can be used to perform the operation S540 described above, which will not be repeated here.

[0169] According to embodiments of this disclosure, any plurality of modules among the acquisition module 2410, prediction module 2420, identification module 2430, and correction module 2440 may be combined into one module, or any one of these modules may be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules may be combined with at least part of the functionality of other modules and implemented in one module. According to embodiments of this disclosure, at least one of the acquisition module 2410, prediction module 2420, identification module 2430, and correction module 2440 may be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or any other reasonable means of integrating or packaging the circuitry, or implemented in any one of software, hardware, and firmware methods, or in a suitable combination of any of these methods. Alternatively, at least one of the acquisition module 2410, prediction module 2420, identification module 2430 and correction module 2440 may be implemented at least partially as a computer program module, which can perform corresponding functions when the computer program module is run.

[0170] Figure 25 A block diagram of an electronic device suitable for implementing a real-time battery power estimation method according to an embodiment of the present disclosure is shown schematically. Figure 25 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0171] like Figure 25As shown, an electronic device 2500 according to an embodiment of the present disclosure includes a processor 2501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 2502 or a program loaded from a storage portion 2508 into a random access memory (RAM) 2503. The processor 2501 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 2501 may also include onboard memory for caching purposes. The processor 2501 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0172] RAM 2503 stores various programs and data required for the operation of electronic device 2500. Processor 2501, ROM 2502, and RAM 2503 are interconnected via bus 2504. Processor 2501 performs various operations of the method flow according to embodiments of the present disclosure by executing programs in ROM 2502 and / or RAM 2503. It should be noted that the programs may also be stored in one or more memories other than ROM 2502 and RAM 2503. Processor 2501 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in said one or more memories.

[0173] According to embodiments of this disclosure, the electronic device 2500 may further include an input / output (I / O) interface 2505, which is also connected to a bus 2504. The electronic device 2500 may also include one or more of the following components connected to the input / output (I / O) interface 2505: an input section 2506 including a keyboard, mouse, etc.; an output section 2507 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 2508 including a hard disk, etc.; and a communication section 2509 including a network interface card such as a LAN card, modem, etc. The communication section 2509 performs communication processing via a network such as the Internet. A drive 2510 is also connected to the input / output (I / O) interface 2505 as needed. A removable medium 2511, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 2510 as needed so that computer programs read from it can be installed into the storage section 2508 as needed.

[0174] According to embodiments of this disclosure, the method flow according to embodiments of this disclosure can be implemented as a computer software program. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 2509, and / or installed from removable medium 2511. When the computer program is executed by processor 2501, it performs the functions defined in the system of embodiments of this disclosure. According to embodiments of this disclosure, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0175] This disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs that, when executed, implement the method according to the embodiments of this disclosure.

[0176] According to embodiments of this disclosure, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0177] For example, according to embodiments of this disclosure, a computer-readable storage medium may include the ROM 2502 and / or RAM 2503 described above and / or one or more memories other than ROM 2502 and RAM 2503.

[0178] Embodiments of this disclosure also include a computer program product comprising a computer program containing program code for performing the methods provided in the embodiments of this disclosure. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the train operation control method provided in the embodiments of this disclosure.

[0179] When the computer program is executed by the processor 2501, it performs the functions defined in the system / apparatus of this disclosure embodiments. According to embodiments of this disclosure, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0180] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 2509, and / or installed from the removable medium 2511. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0181] According to embodiments of this disclosure, program code for executing the computer programs provided in embodiments of this disclosure can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can execute entirely on a user's computing device, partially on a user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0182] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0183] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A real-time estimation method for battery state of charge, characterized in that, include: Acquire the status data of the target battery during its operation; An initial estimate of the battery charge is determined based on the aforementioned status data; Identify physical anchor points from the state data that correspond to the phase transition process of the battery material, the physical anchor points corresponding to stable state of charge reference values; Upon identification of the physical anchor point, the initial estimate is corrected based on the power status reference value to obtain the target power status.

2. The estimation method according to claim 1, characterized in that, The step of identifying physical anchor points from the state data includes: Obtain capacity status data from the status data, the capacity status data indicating the relationship between battery capacity and voltage; In response to determining that the difference between the capacity status data and the capacity status reference value corresponding to the physical anchor point is less than a first predetermined threshold, the physical anchor point is identified.

3. The estimation method according to claim 2, characterized in that, The method further includes: In response to determining that the difference between the capacity status data and the capacity status reference value corresponding to the physical anchor point is greater than or equal to the first predetermined threshold, the physical anchor point is not identified. The initial estimate is determined to be the target power state at the current moment.

4. The estimation method according to any one of claims 1-3, characterized in that, The method further includes: Obtain reference state data of a reference battery during the cycle aging process, wherein the battery material of the reference battery is the same as that of the target battery; The anchor point charge status of the reference battery is determined based on the reference state data; The power status of the anchor point is determined as the power status reference value of the target battery, and the reference capacity status data corresponding to the power status of the anchor point is determined as the capacity status reference value corresponding to the physical anchor point of the target battery.

5. The estimation method according to claim 2, characterized in that, The battery status reference value includes one of the following: This is determined based on the phase transition characteristics of the battery material; The target battery remains stable during cyclic aging. The offset within the preset temperature range is less than the second predetermined threshold; It has multiple power status reference values, each power status reference value corresponding to different capacity status data.

6. The estimation method according to claim 2, characterized in that, The step of correcting the initial estimate based on the power status reference value when the physical anchor point is identified to obtain the target power status includes: Determine the battery capacity weight of the target battery when the physical anchor point is identified; Based on the battery capacity weight, the initial estimate and the state of charge reference value are weighted to obtain the current state of charge of the target battery; The current state of charge is processed based on predetermined conditions to obtain the target state of charge of the target battery; wherein the predetermined conditions restrict the target state of charge from decreasing over time during the discharge process of the target battery.

7. The estimation method according to claim 6, characterized in that, The step of processing the current battery state based on predetermined conditions to obtain the target battery state includes: In response to determining that the target charge state of the target battery at time t-1 is less than the current charge state at time t, the target charge state at time t-1 is determined as the target charge state of the target battery at time t. In response to determining that the target charge state of the target battery at time t-1 is greater than or equal to the current charge state at time t, the current charge state at time t is determined as the target charge state of the target battery.

8. The estimation method according to claim 6, characterized in that, Determining the battery capacity weight of the target battery when the physical anchor point is identified includes: The capacity status data and the capacity status reference value are processed to obtain the battery capacity weight of the target battery at the current time.

9. The estimation method according to claim 1, characterized in that, Determining the initial estimate of battery capacity based on the state data includes: The following operations are performed using a trained target model, which establishes a non-linear mapping relationship between the state data and the battery charge level, wherein the state data includes operational data related to battery aging during the operation of the target battery: Extract multidimensional feature data from the state data; The multidimensional feature data is fused to obtain fused features; wherein the fused features indicate the operating features and aging features associated with the target power state. The battery capacity is evaluated based on the fused features to obtain the initial estimate.

10. A device for real-time estimation of battery state of charge, characterized in that, include: The acquisition module is used to acquire status data during battery operation. The prediction module is used to determine an initial estimate of the battery charge based on the state data; The identification module is used to identify physical anchor points corresponding to the phase change process of battery materials from the state data, and the physical anchor points correspond to stable state of charge reference values. The correction module is used to correct the initial estimate based on the power status reference value when the physical anchor point is identified, so as to obtain the target power status.

Images